‘We’re stubborn enough to create our own world’: how programme directors frame higher education quality in interdependence

ABSTRACT Little is known about how the complex notion of higher education quality is understood and (strategically) handled by a specific group of key university actors: directors of educational programmes. A framing analysis of in-depth interviews was conducted to explore how bachelor-programme directors in Dutch social science departments understand and enact quality, while maintaining multiple commitments. The analysis revealed that directors share a non-problematic, understanding of quality as realising a good educational programme, programme. They enact different quality frames while upholding their programme and position but face issues in practice. Balancing different goals and interests is a recurrent strategy. The directors’ room for manoeuvre to enact their quality views, however, is position-dependent. Whereas some directors can play it out in any direction, others experience responsibility without power. Quality’s plasticity provides the flexibility to maintain the idea of improvement, even in limiting circumstances, while preventing structural changes at a more fundamental level.


Introduction
The concept of higher education quality is both airy and concretised in practice (Giroux, 2006). The notion played an evocative role in change processes during the European 'quality revolution' in the 1980s and 1990s, as it was easily accepted as a fashionable management concept. Policymakers and institutional managers recognised themselves in its different manifestations and its vague appeal enabled the rearrangement of steering relations between the Dutch government and institutions (Giroux, 2006;Stensaker, 2007;Weenink et al., 2018).
A vague notion can, however, become problematic when articulated in practice. Quality's translation into practice was resisted by academics from below, who opposed its formal meanings, and understandings such as 'improvement' and 'transforming the learner' were countered by articulations such as 'lack of trust', 'burden' and a culture of 'getting by'. These articulations were not stable or uniform and varied in resistance to, and adoption and adaptation of, formal quality frameworks (Newton, 2002;Overberg, 2019). Moreover, the same person could deploy different understandings and enactments depending on whether that person was teaching, researching or managing, or in other interrelations and contexts (Harvey & Green, 1993;Seidl, 2006).
Discursive studies have patterned quality discourses and note that quality is cloaked, tends to melt into its context and lacks conceptual power (Vidovich, 2001;Harvey & Newton, 2007). Following a dialectic perspective, Morley (2004) argued that quality is not easily opposed communicatively, as it has become a metanarrative that discursively carries with it the threat and trace of 'the other'. To oppose quality is to become the opposite of its goodness, demanding espousal of its negative sides. It is in this sense a modular and plastic word, a word without meaning (Poerksen, 1995;van der Laan, 2001).
Practice scholars, however, note that notions such as quality, care and goodness are layered and strikingly complex. People draw upon multiple meaning structures like quality assurance schemes and teaching experiences to assess and improve them in practice (Mol, 2010;Vettori, 2018). Opposing quality perspectives against one another is just one of many ways to constitute it communicatively. Although it may be difficult to reconcile competitive quality notions with the tacit teaching process, it can be apposite to engage academics in marketing practices to increase student numbers. The quality perspective adopted depends on the specific situation and context. Such constituent practices are inherently dynamic and processual and concern not only how quality is understood but also how it is performed (Mol, 2010). As Wittgenstein (1953) noted, the flexibility and multiplicity of language not only sustain existing views and latent meaning structures but also enable people to use them as a toolbox and commodify their meaning in specific, situated contexts. This also applies if people can only vaguely indicate what they mean, for example when saying 'stand roughly there', while indicating a certain spot. Playing language games creates new meanings and directions.
Contextual shifts in professionalisation, research, education and teaching régimes are constantly reconfigured and mediated with fellow academics (Leisyte & Dee, 2014). People apply several communicative strategies in dealing with dynamic and often contradictory institutional logics and environmental complexity. These include, for example, bridging towards other perspectives and bonding with like-minded people (Smets et al., 2015). It is not known, however, what such complexities mean for how quality is performed. Things do not always transpire as expected and people find themselves entangled in their own rules and practices while playing language games. People often utter several ambivalent and contrasting perspectives in one sentence. It is this entanglement of different rules and perspectives that this study aims to elucidate and understand (Wittgenstein, 1953).
Presuming that quality is difficult to contest as well as constituted and improved communicatively, it is key to assess how the notion is enacted by academics in engagement with varying actors and contexts. How quality understandings and enactments vary in engagement with contextual dynamics and the room that academics have to optimise it are, however, understudied. Context-oriented studies such as Westerheijden andKohoutek (2013), Blanco Ramírez (2014) and Overberg (2019) focus on how quality is implemented or 'translated' and how external quality assurance régimes, policies and institutionalised environments shape actor perceptions and actions, rather than starting with what academics themselves find salient and select as relevant quality notions for their specific situations and contexts (Cardoso et al., 2017;Elken & Stensaker, 2018). Even studies, as for example Elken and Stensaker (2018) and Vettori (2018), which focus on sensemaking processes, situated perspectives and barriers to improvement, barely address how various repertoires of meaning structures and conceptualisations are performed in different practices.
This article elucidates how quality is understood and enacted by academics in interrelation with various dynamic and open-ended formations and complexities, and what this means for its optimisation. Twenty-four directors of bachelor programmes in social science departments of six Dutch research universities were interviewed about achieving quality. Academic middle managers often combine their managerial role with a position as lecturer and researcher. They navigate different commitments and their work has been framed on the boundaries between managerialism and collegiality (Clegg & McAuley, 2005). As they adapt and adopt policies and other inputs in the situated context, studying their quality enactments and understandings enables an assessment of how the notion is communicatively constituted in complex, dynamic interdependencies. The research question is: how do directors of bachelor programmes in Dutch research universities understand and enact higher education quality, while interacting with dynamic actors and contexts?
This study presumes that whether it is considered problematic and articulated as such depends on the specific contexts and processes. The following sub-questions address these processes: how do the directors understand and enact quality? What tensions do they identify and what actors and contexts are involved and considered relevant? How do they deal with these tension fields as they unfold?

Combining framing analysis and figurational analysis
Framing analysis and human-figuration analysis are combined from a languagecentred practice perspective to examine how quality is performed and what directors find salient in complex situations and processes. Framing analysis serves as an analytical perspective and method to assess what is going while making sense of situations and issues. Figurational analysis draws attention to people's space to enact specific quality perspectives in interdependence with other people.
Framing creates meanings in interaction while relating previous experiences and cognitions to dynamic, situational contexts. Frames are implicit theories of a situation and framing is a language-driven ordering process through which people select and label the relevant features of the situation, structure these into an understandable whole and behave accordingly (Goffman, 1974;van Herzele & Aarts, 2013). Framing analysis exposes quality's interrelations with webs of power, as people actively construct frames that fit their interests, feelings, convictions and backgrounds to achieve specific goals. What gets framed are usually the issues at stake, actors' identities and relationships and the process itself. Intersubjective processes of meaning creation draw on previous experiences and understandings. Whether these are activated as people negotiate the meaning(s) of their actions depends on the situated context and the inter-relational dynamics (Goffman, 1974;Aarts & van Woerkum, 2006).
Framing analysis provides both a theoretical lens and a method to assess how people deal with issues in complex environments (Dewulf et al., 2009). Rein and Schön (1993) have applied it methodologically to assess problem setting in intractable policy controversies (Schön & Rein, 1994). van Hulst and Yanow (2016) built upon their and Goffman's (1974) work by looking at what people find meaningful and how they frame a way forward in the event of tensions. What people select, name and categorise as relevant from their complex environment is key to a dynamic, processual understanding of framing as it develops in sensemaking processes. The exclamation 'that's not quality!' for example selects and names a specific situation but also categorises its features as undesirable and pushes actions in another direction. It is crucial to identify what directors select, name and categorise as salient, to understand when and how quality understandings differ and change.
The figurational approach developed by sociologist Nobert Elias is used to further assess quality in interdependence with its contexts, as it prioritises selections and connections in relational processes. A figuration is a constellation of mutually oriented and dependent people, with shifting asymmetrical power balances: a nexus of human interdependencies (Elias, 1994;van Iterson, 2009;van Krieken, 2001). These figurations create meaning in practice and shape society, as they restrict and enable what directors can do. Power therefore develops within the relationships as people are mutually dependent; both the lecturer and the student have control over each other as they are both needed to realise good teaching. Such interdependencies are at least bipolar, but usually multi-polar (Elias, 1978). Figurations are in this sense interdependency networks (van Iterson, 2009). For example, the way in which governmental quality policies are translated into institutional practices influences directors' room to enact specific quality frames. What programme directors want to achieve, however, can be more relevant to how their figuration develops, as also how they relate to colleagues or how the programme is positioned within its research domains.
The boundaries of these figurations are not pre-given as they may change according to changing contexts and also form larger figurations, nesting within one anoth,er. The interdependency chains have become so long, interwoven and complex that it becomes impossible for people to second-guess the actions of others. It is likely, for example, that directors do not take the views of members of parliament into account, although these parliamentarians might engage with departmental deans. Directors and parliamentarians are then indirectly related. The consequences of this interweaving of interests and actions of different groups are that none of these groups can pursue entirely their own interests. The social order is the unintended, slowly emerging result of people's actions (Elias, 1978(Elias, , 1983van Iterson, 2009: Kuipers, 2018.

Methodology
Figurational analysis and framing analysis were used to explore the patterns of quality performance in specific settings.

Sampling strategy
A two-step (purposeful) maximum variation sampling strategy was applied to select the interviewees (Patton, 2002); first, at institution level: six of the 14 publicly funded research universities, varying in geographical location, size and profile (generic as well as technical universities) were selected. Then, all 37 directors of bachelor-level social science programmes in these six universities * The programme director role in Dutch social science departments is often fulfilled by associate professors, as reflected in the selection. The sex ratio reflects the distribution of men and women in academic positions in the Netherlands.
were approached, of whom 24 participated. Table 1 shows the directors' distribution across positions, including sex ratio. The aim was to vary maximally on the programme size and field dimensions but also on academic position and sex distribution. Between three and six directors from each selected university were interviewed. The formal positions and responsibilities of the interviewees varied: some directors coordinated a single bachelor programme, whereas others managed employees and programmes from undergraduate to PhD level.

Interview procedure
The interviews (average duration 1.5 hours) were open and minimally structured to allow the directors to share their perceptions, experiences and strategies, after first introducing themselves and describing their work as a director, their responsibilities and their position within the organisation. To answer the research questions, several key topics intentionally recurred in each interview: their concept of quality, situations that involved quality, how they assessed quality, the issues and dilemmas that they experienced in realising quality and how they dealt with these. Finally, they were asked about policy measures that would greatly improve their programme's quality. Throughout the interviews, the interviewees were asked to whom they related regarding quality and which documents and policies they considered relevant in specific situations.

Analysis
All interviews were recorded, transcribed verbatim and analysed using Atlas-TI. An interpretivist grounded-theory approach was taken to elucidate the quality framings and theorise upon the dynamics in human figurations. Charmaz (2014) interactional and constructivist grounded theory approach was followed, as it combines well with the interpretive framing methodology and changing figurations (Gioia et al., 2013;Schaffer, 2016). Analytical rigour was achieved by constant comparison in the initial coding phase as well as in the focused coding phase of how the notion was framed in human interdependence (Charmaz, 2014;Gioia et al., 2013;Schaffer, 2016). The interviews were treated as sites of director-interviewer interaction. Following van Hulst and Yanow (2016), the issues, notions, identities and relationships in their environment that the interviewed directors selected, named and categorised as salient were systematically identified. Differences and similarities between the interviewees were, therefore, constantly compared. For instance, the initial coding phase elucidated that the directors consistently demarcated higher education quality as the realisation of a coherent programme. It also indicated that they situated the main interdependencies and issues within the institutions and that academic hierarchies mattered for how the notion was understood and enacted.
In the focused phase, the strategies that they deployed to deal with the various intra-organisational issues and what they considered as hindering or enabling the realisation of educational quality were further compared across the figurations. Specific attention was paid to similarities and differences across academic positions and what they said they could do in interrelation with their colleagues and other academic actors was further compared. Furthermore, how these findings related to other aspects, such as educational models and policies or academic fields, was assessed, as also whether such aspects were considered specifically enabling or restrictive.

Findings
This section follows the line of reasoning laid out in the sub-questions, addressing first how the higher education quality notion was framed and understood by bachelor-programme directors by looking at how quality was selected, named and categorised in the interviews; and second the tension fields identified and how these were handled across divergent figurations.

A good educational programme
The interviews revealed that the directors (D) shared an understanding of higher education quality as a close synonym of good education. It was recurrently articulated as different from other qualities and the directors specifically selected and named educational practice as important and distinctive. In response to the interviewer's (I) focus on quality: D: Because now it seems like you're investing a lot in 'what is quality', and so. The quality of cheese is something different than the quality of water, and the quality of education. The word 'quality' itself seems to me totally uninteresting to study, because you want to know something about education. And of course, put it in the right context: what is it in the Netherlands . . . what are those people's practices? And that while we are now working with a totally developed system of quality standards on which we assess one another. And that is just there. And I find that the liveliest that there is, and what the discussion is about. (1) It was difficult to discern whether the directors were talking about good education or education quality. Comparison of the sparse word use across interviews, however, revealed that education quality was interwoven with quality assurance. Directors operationalised it and had 'the liveliest discussions' on how to assess one another with the quality standards system.
The directors considered it their prime responsibility to achieve a qualitatively good educational programme and considered this a concrete, non-problematic goal. D: You're saying; 'quality is elusive'. Well, I can tell you about that. I don't think it is elusive at all! I: No, how can it be caught? D: Well, if I now-I haven't done that systematically, I haven't read it, so maybe, then I should come up with a different story-but if you just ask me as an educational director: 'what do you consider educational quality?' Well, that's actually really simple. What do I want people to learn in four years? . . . So, this has elements related to practice. You can analyse it and report about it. And after four years, for me you're an academic. I: Yes. D: I mean, tell me if I'm mistaken, but I don't find that complicated at all. (7) Quality was understood as setting and meeting objectives for an academic programme and relating these to student attainment. The directors listed elements to realise this, such as the quality of lecturers and courses, coherence in tracks and the curriculum, effective learning strategies and labour-market preparation. Some interviews consisted mainly of elaborations of such elements. This goal frame of realising a good educational programme situated educational quality within the institutions and connected the essential lecturer-student interactions with other programme elements.

Aligning programme elements
The continuous alignment with one another of the programme's goals, means and assessment was considered key. D: These three things must really be connected . . .. And if you can intertwine these well, then you have a somewhat greater assurance that the quality of education is at least guaranteed to the students. (9) Guaranteeing quality education through continuous alignment was essential for all directors and several mentioned 'constructive alignment' as the underpinning educational perspective. The alignment process concerned discussions and choices at course level with colleagues such as lecturers and course coordinators. D: We were having a discussion about 'what do we want to achieve with the defence of the master thesis?' Because if the students get feedback, they will certainly learn from it. But that is a different skill than having to defend something yourself. (15) Whereas discussions with lecturers concerned courses, the directors named the programme level as their main locus for quality Realising coherent learning paths was key and they wanted to ensure that all students were able to learn what they should and that all elements were taught.  (7) Aspects such as theoretical knowledge and analytical and writing skills related to the assessment of students' attainment; the knowledge, insight and skills that they should attain in a coherent programme that prepared them for the labour market and society. Many directors were proud of their coherent programme with solid learning tracks. This fragment also exemplifies how the quality of the programme was named and categorised as different from the programme quality at universities of applied sciences, yet comparable to other academic programmes.
In summary, the interviews elicited a widely shared quality frame that understands educational quality as a close synonym of good education and directors confined it to the consistent alignment of a coherent programme in relationship with various stakeholders. This way of thinking is 'in their veins' and most directors do not consider this alignment process very problematic. However, the interviews revealed also that directors face issues when they want to achieve quality improvement in practice and that these issues are very much related to the development and maintenance of constructive relationships within their institutions.

Quality and its tensions
D: Yes, well, I think that we pretty much agree with each other on 'what quality actually is', to put it like that. When we think we have delivered a good student. But given our resources, how can we achieve that in the best way? Well, I find that a quest. My feeling is also that the frameworks are still getting more and more narrow. (8) This fragment reveals that tensions related to how a good programme could be achieved and improved given limited resources, in the context of perceived narrowing quality frameworks. Various issues were associated with realising programme quality, ranging from the language taught to a lack of teacher professionalisation. Although they involved societal questions and the adaptation of external quality demands, these issues were situated and dealt with within the organisation.
There were two interwoven domains where tensions were recurrently perceived and situated. First, there were other goals and qualities within the university. These tensions concerned the complex distribution of budgets, as well as interdependence with other programmes and research. Second, there were specific tensions regarding the situated teaching process.

Balancing budgets and programmes
D: We had thought it all out well in years of budget cuts, and reduced it. How can we solve it? With for example online things, or . . .. And, then, then they're saying: 'the contact hours have to be raised'. Hoppa! In two years, you have to raise all those contact hours. The directors found it difficult to change policies, as they had to make arrangements and renegotiate with a multitude of actors, while already finding the current budgets too limited. Governmental budgetary restrictions were heavily criticised and 'more money!' was on the tip of their tongues when they were asked what would really improve educational quality. Money is time and more hours enable intensive classes and contacts, which apparently contribute to higher quality.
The directors did not engage with the national government though and dealt with budgetary issues within their institution. Several considered it their task to handle the budget as efficiently as possible.
D: It is of course very easy to say, 'we'll throw in more lecturers, and more time'. That will surely improve the quality. But we have only a limited amount of money, so I have to balance the two. (4) The quality of different courses and elements were weighed against one another. 'If you want to supervise this individual thesis, that goes at the expense of . . . it's a real optimisation problem!' (8). Another director noted, 'we are not distributing money, we are distributing losses' (22).
These optimisation processes transcended their direct influence and the directors depended on other institutional actors' budget allocations, rules and practices. A small university, for example, centralised the distribution of educational resources and programmes 'bought' courses from, and 'offered' courses to, other programmes and student groups. This enabled an extensive programme, even when it drew a limited number of students. It also, however, threatened analytical depth, as courses served students from different programmes. 'Quality for whom?' (17).

Balancing research and education
All directors believed that a good programme could not come without academic research skills and knowledge. Lecturers were often researchers and the tension field was framed as 'just time' (6). Intensifying education was, however, found to be achieved at the expense of research time and often valued less. 'You are of course talking about "educational burden" and "research time". That makes a big difference!' (14). Education and research were often articulated hierarchically in relation to each other and the interviewees stated that lecturers' careers still depended on research performance, whereas teaching was believed to be more essential.
Tensions between research and education were reflected in the directors' powers within figurations. Several were engaged in hiring new staff. D: If I leave this to the professor who is responsible for the research programme, there will be all research-hotshots hired, who mainly want to do research. And if I look at the quality of my programme, I want someone who is intrinsically motivated, preferably for education. Who also has the skills. So, we have to ask for attention on that, and that is the kind of game being played. (9) Foregrounding educational quality is a game with limited possibilities, 'you cannot do this all the time' (1). Several directors noted that there had to be a good balance in the team. The interdependence between research and education was articulated hierarchically, yet was complementary and evolving slowly. Various directors noted that the research-education balance had changed but that an academic career built on teaching remained an exception.

Issues concerning the situated educational process
Directors depended on their relationship with lecturers, who had to deliver good courses. Most believed that decisions should be made in conjunction with those who have to deal with them in practice and that lecturers and course coordinators should be trusted in their professional autonomy. However, the interviews also revealed tensions and contrasting frames. 'One cynical lecturer can ruin a whole course' (22). The directors were held responsible for educational quality but had a limited view on what happened in practice and lecturers maintained their own views and multiple commitments.
Tensions with budgets, other programmes and research concerned the situated educational process but were played out in hierarchical relationships, as the directors depended on managers' decisions about resources and on professors about learning assignments. Whereas directors at professor and associate professor level were 'playing the game' in both the managerial and the situated context, others noted that 'we are very much in the position that we have responsibility without power ' (14).

Dealing with tensions
Tensions concerning budgets, research, other goals and situated practices occurred across all figurations. The interviews show that the directors deployed similar strategies to deal with these, for example, bracketing their work into smaller pieces and balancing different interests. Such strategies were similar to generic strategies to deal with competing but complementary goals within complex organisations (Smets et al., 2015). The directors also shared their work and responsibilities with actors, such as programme and course coordinators, support staff or education and examination committees. These academic interdependencies were, however, often articulated hierarchically and the directors' room for manoeuvre to enact specific quality understandings depended on their figurational position. They switched between protective and changeoriented strategies regarding situated educational practices.

Protective and change-oriented strategies
Directors were protective when external demands, such as policy changes, seemed to hinder the programme's situated processes. They were selective in what they adapted from policies and identified with their academic staff while maintaining their quality views. D: Well, you can dig in your heels, and say, 'We will not do it'. Then you'll have a problem, also with the University Board. So, you can better say; 'Oh, this is what they are demanding from us. Let us see what we find valuable and necessary, and that is how we arrange it.' I: Yes. D: And then we're stubborn enough to create our own world. And that's what you do, and then suddenly you're the university's best practice, if you do it like that. (19) The directors created their own world with their colleagues involved in the programme and maintained their quality frames, while selecting what they considered valuable and necessary. Conflict was thereby avoided, as they did not contravene the interests and quality views of higher management and policymakers. Nonetheless, quality practices did change. This protective strategy was also deployed when directors felt that quality policies went against their own views and tacit situated practices but did not want to offend higher management by defying their policies. D: What speaks for this programme is when I explain it: 'Guys, it is not my idea'. That makes it totally different. And, 'Don't shoot the messenger!' So, let's use this to do it in the best possible way, but stick with our inner drive. (7) This professor bonded with the staff and positioned himself as protecting core educational practices. These practices were, however, affected by the criticised policies. 'It can all be done. The point is that I don't think that a rubric makes education much better' (7). The compromising paradoxically strengthened the detested bureaucratisation and rationalisation of educational practices.
The protection of educational practices concerned the situated teaching process and the lecturers' academic autonomy to shape their lessons in interaction with students, as well as their own practices. These interests were not always the same, however, and the directors also aimed to change teaching practices to improve the programme.
Curriculum changes were widely considered useful to keep the programme up to date with changing societal and organisational demands and there were constant incremental changes. Experienced directors noted, furthermore, that there was always at least one major change process going on. These processes were initiated by different actors but the directors tried to be selective in what they did and when they did it. Opening a new campus, for example, was a good opportunity for further expansion and differentiation. In other instances, changes were considered necessary to survive.
Such change-oriented strategies had different quality frames than those of protective strategies. One professor commented that the programme was 'going down the drain' and had to be repositioned to make it more attractive for students. He compared and valued the programme against other Dutch programmes in the research domain. 'I know the landscape very well!' Improvements were made to be more distinctive, 'We are now educating for the future' (22). This quality perspective is competitive and can be identified as both entrepreneurial and consumer-protective, as it serves students as consumers (Vettori, 2018). It was considered appropriate in the context of attracting more students and supported by institutional management. It also strengthened the coherence of the programme in practice. Not all directors enacted an outward, competitive view, however, and the analysis suggests that it was enacted mainly by professors and associate professors.

Different spaces for quality
Directors' academic position was not the sole element determining their figurational space to enact quality understandings. Various aspects such as programme size and allocation rules and practices mattered: large programmes, such as psychology, could operate relatively independently of other programmes and higher management. Moreover, directors in all positions had little leeway when institutional relationships were seriously troubled. In one institute, the director and the institutional management felt that they could not act upon the programme as they wanted to, as renowned research groups were protecting their education and specialisations against change. He argued, however, that he did achieve quality improvement. 'Well, I caught up the things where I could make a difference' (21). The situation changed only after it entered the public domain.
The constant comparison of different aspects across different figurations suggests that the directors' hierarchical academic position was important for their quality repertoire and how they played it in all directions. It seemed to matter for how they dealt with interrelated issues concerning budgets, other educational and research qualities and teaching practices.
A salient distinction between the directors' positions was whether and how they engaged with other actors and contexts within the university. Full professors, for example, were in close contact with other professors and higher departmental and institutional management, whereas administrative directors engaged with other directors and students. The differences also concerned the shape of these relationships and their effects on relationships on which all directors relied to bring quality into practice, especially among the educational staff. These differences recurred across the different domains and it did not seem to matter much what the specific topic of education and research was.

Full professors.
Directors with the position of full professor switched between protective and change-oriented strategies and considered which perspective was apposite. Policies that they considered harmful in practice were discussed with the management team and departmental management. They also ensured that they had managerial support to change the programme. As one director noted, 'I have the authority to put things under pressure. That is accepted'. Several professors had a say in the distribution of the budget and staff reviews. They also garnered staff support and another professor noted: 'You can be authoritative, but it sets rancour. In the end it works against you' (7).
One communicative strategy to connect their management strategy with educational practices and to create engagement was to identify and act from their researcher or lecturer role. 'I find it important to still have the teaching experience. To keep understanding what is happening there, what the primary process is' (7). They proffered their own experience in interaction with lecturers and engaged them in changing teaching practices towards what they considered best.
The strategy of speaking from the situated perspective was also deployed to protect the programme against hindering demands from other actors. The professors noted a strengthening of the administrative position within academia and felt held accountable by administrators who used instruments such as the National Student Evaluation. 'They are doing all those things therewith. But I really don't need that! I just need a panel session, a good team' (24). Such inquiries were framed as conflicting with their situated practices, including their interpretation of student evaluations to improve the teaching process.
Associate professors. The associate professors' strategies showed similarities with those of full professors. Some noted that they had the authority to take the final decision and that it would not differ if they were professors. Most, however, identified more strongly as managers of the specific educational domain and they were less involved in decisions regarding budgets and academic staff.
Associate professors temporarily devoted much time to programme management and several had taken the same educational leadership course on complex educational issues and programme organisation. They focused on improving different elements and conducted research or initiated projects on, for example, teacher professionalisation or the effects of distance learning on student attainment. They used these experiential and analytical skills strategically to create engagement. 'Many people saw that we made a good analysis of the situation and how you can shape it in the future' (1).
Associate professors' perceived powers to enact their views varied, however. Several directors could garner ample support amongst managers and lecturers by combining their analytical framings with the strategy of making issues manageable. They sought, for example, the right time to start a project or discuss issues at management level. Other directors, though, perceived more limitations and the interviews in one department were divergent. Whereas some easily convinced departmental management of their strategy to empower lecturers, others felt that their proposals were not heard and that they could not connect the situated practices with managerial views.

Assistant professors.
The few assistant-professor-level directors usually worked together with an educational director who had more managerial responsibilities. Their role was mostly coordination; they looked after quality evaluations and ensured that staff members were heard. One director described herself as the 'primus inter pares' amongst the educational staff. They were less involved in the management game, though, and their interactions and quality enactments concerned primarily the situated perspective.

Administrative directors.
Directors with an administrative position framed themselves as 'lubricant oil', stressing that they aligned different perspectives, for example by consulting different research groups ('blood groups') and students in preparing curriculum decisions. Their approach was processual and they bridged different quality perspectives but did not always feel themselves heard. Those without a teaching background were especially limited in their quality repertoire and connective work. One administrative director felt framed by academics as 'part of the bureaucracy'. The administrative directors depended on procedures and investments in personal relationships to have impact. They were reluctant to speak out and another one recounted that professors did not even realise that they had different concerns. Some female directors with the position of assistant professor or administrative director noted that the power differences were gendered.

Student evaluations
Student evaluations played a particular role in how quality was enacted, especially concerning tensions with educational practices. All directors used them as part of the instrumental repertoire to monitor the quality of courses and teaching. They were 'sailing on the evaluation figures' (6) and one director remarked that 'it is in my interest to know what is going on' (1). When discussing student evaluations, directors noted that such evaluations were biased, did not necessarily measure quality as student attainment, or provided a limited view on practices. They found them limited but a valuable addition to what was heard and seen.
Evaluations were used to monitor and discuss the quality of teaching, courses and lecturers but also their own programme, with actors such as the educational committee. Several directors noted that a tremendous value was put on them and were critical about abstract, internal uses to improve and assess their programme. D: Well, what is measured now, study success, dropout-rates, average duration, that is a bit what is available at national level. You will have to do it with that . . .. You could approach it differently and for example include student ordeals. For example, the NSE [National Student Evaluation], or the Elsevier-survey, or rankings, but I would be careful with that . . .. Because if you use them to steer, you will also be held accountable for that. And that was just not the intent. (9)

Conclusion and discussion
To understand how higher education quality is performed in interdependence with environmental complexity, this study elucidates how directors of Dutch social science bachelor programmes enacted and understood it, while maintaining different commitments.
Framing analysis and figurational analysis were combined to analyse indepth interviews and to explore the interrelationships and quality perspectives that programme directors considered relevant in their specific situations and contexts. It can be concluded that the directors shared a non-problematic understanding of quality as realising a good educational programme. We found also that quality was enacted in such a way that the directors could deal with tensions emanating from changing societal and organisational demands. These tensions were experienced and acted upon within the organisation and concerned quality's interdependence with limited budgets and other intra-organisational qualities and goals, including the situated teaching process. Directors' strategies such as bracketing quality into manageable elements and compromising and balancing different interests showed similarities with generic strategies that Smets et al. (2015) identified to deal with complexities as expressed within organisations.
The results also suggest that the directors' room for manoeuvre to enact their quality frames depended on their specific, power-ridden figurations within academia and differed per position. Interdependencies with research and other educational qualities were often articulated hierarchically: whereas some directors could deploy their quality frames in any direction, others felt that they had responsibility without power. It is striking that the directors' relationship with lecturers and their influence on the teaching process were affected by their relationship with (other) full professors, even though professors experienced a strengthening of academic bureaucracy. All in all, the enactment of educational quality in complex interdependencies pushed directors to develop strategies in interaction with their near academic environment, whereby traditional hierarchies constrained their room for manoeuvre.
Quality is a broad notion. To uphold their programme as well as their own position, directors put their efforts into those aspects that they could change if they could not improve what they wanted to. Quality's multiplicity and plasticity provided the flexibility to maintain the notion of quality improvement, even though it changed what they considered quality.
The study supports Vettori's (2018) analysis that people seemingly share a view of educational quality, while drawing upon different underlying meaning structures. It also shows, however, that what might be individually seen as a quality view was inherently interrelated with what was considered apposite for the directors' specific position in their hierarchically ordered academic figuration. Quality's plasticity therefore suited their interactional, situated process for dealing with tensions and contradictory demands. As the directors resolved their issues within situated figurations, there were no countervailing powers against external quality demands and decreasing budgets. Even protective strategies could not stop situated educational practices from changing in unwanted directions.
The study draws attention to how quality is played out in academic hierarchical figurations. Although directors' room for manoeuvre is related to their formal rank, it is worthwhile to further investigate how it relates to how the academic order is being shaped. Gender differences, but also the tendency to value research over education, do seem to play a role here and also to affect the directors'room for manoeuvre in relation to other figurational actors, specifically their near colleagues. Such aspects are, however, also dynamic, multiple and multifaceted; and it would be interesting to study how the valuation of higher education quality relates to other valuation processes in academia.
The present study is restricted to Dutch research universities, with a differently organised hierarchy than universities of applied sciences. Further comparative research would provide both a broader and a deeper perspective on quality enactments and understandings in divergent higher education systems. Finally, it should be noted that the interviews were conducted before the COVID-19 crisis. Current research is needed to assess how changing teaching practices and communications may affect the enactment of quality understandings.

Disclosure statement
No potential conflict of interest was reported by the author(s).