Who takes part in participation? Challenges to empowering student voice in music teacher education

ABSTRACT In this study, we employ a participatory action research framework to investigate how preservice music teachers can take part in developing their own education. The main focus is on how two music teacher educators at two institutions in Norway work to create a space for student voice and participation. The study’s data are analysed and presented through a combination of self-study methodology and narrative analysis conducted by two of the authors, followed by reflections by the two other authors. In these analytical steps, we identify four teacher roles in the form of metaphors: ‘the impatient manager’, ‘the conflicted gatekeeper’, ‘the balancing artist’ and ‘the reluctant host’. These roles are further discussed in light of theoretical perspectives on student voice and participation. The results reveal challenges in the use of participatory action research and in making changes to music teacher education.


Introduction
It is a challenge for teacher education to continually change and update itself to remain relevant to students and society. According to Elliott and Silverman (2020), in some ways, music teacher educators must 'read the future' and consider what today's teachers require in order to meet the needs of tomorrow's students (81). However, research on music teacher education has shown a strong resistance to change due to influences from the long traditions of music conservatories (Aróstegui 2011;Bowman 2007;Saetre 2014). Scholars have addressed the need for stronger teacher and student agency in music education in order to move beyond the circle of repeatedly reproducing the same practices (Allsup and Westerlund 2012;Georgii-Hemming and Westvall 2010;Powell 2019). Students' prior experiences with music and music education are a resource that should be acknowledged and included in music teacher education, and participation will facilitate knowledge construction (Henley 2017;Kenny 2017). In this article, we discuss music teacher educators' role in facilitating student participation and voice. 1 Centring students' prior experiences in the teacher education curriculum encourages them to begin constructing a personal orientation to their future work as music educators. As Campbell, Thompson, and Barrett (2012) argue, it is 'through deep analysis of past educational experiences that our future music teachers will begin to conceptualise for themselves the complexity of teaching as well as explicitly "own" their personal development as future music educators' (76). In addition, preservice teachers bring to their education personal metaphors of teaching and teachers, as well as assumptions about the profession, grounded in their own personal experiences in music (Thompson and Campbell 2003). Often, these metaphors and assumptions go unrecognised by the students themselves and can impact the ways in which they approach teaching and learning in the university classroom and beyond. Conway et al. (2020) encourage music teacher educators to create an atmosphere in which 'future teachers are actively pushed to develop an agentic, forward-thinking teacher identity' (913)(914). This is no easy task, as we must 'step outside of our comfort zones, take chances, and develop musical and teaching skills that are unfamiliar to us ' (911). This involves being willing to sometimes give up control and feel vulnerable. These challenges and others were experienced by the music teacher educators in our study.
Our guiding research question is: How do music teacher educators perceive their impact on student participation in a participatory action research setting?
Asking this question also involves exploring what was at stake for the music teacher educators involved and the vulnerability implied in giving up control. Although participatory action research (Pant 2014) is the framework for the study's data collection, we have, as a result of focusing on teacher educators, employed a self-study methodology (Vanassche and Kelchtermans 2015) to work with the material. The data from the two study locations are presented as narratives (Polkinghorne 1995) written by the researchers/teacher educators who conducted the sessions, one from each institution. The narratives are followed by short reflections by the two other researchers, in which they identify teacher educator roles in the form of metaphors. Lastly, we discuss our findings in light of the concepts of student voice and participation. There are several pathways to becoming a music teacher in Norway, such as a one-year postgraduate teacher training course, following a degree in performance or musicology, the new fiveyear teacher education in practical and aesthetic subjects (including music), or, as in this case, choosing music as an elective subject within generalist teacher education. The circumstances of the data collection were quite different at the two institutions. The sessions 2 at the first institution were conducted by Bendik within the regular teaching schedule and on a course on the history of Western classical music, in which all of the students in the course took part. By contrast, the sessions at the second institution were conducted by Silje as an extracurricular offering, with a group of student volunteers. Choosing these two different contexts was both because of convenience and because the two different contexts would provide us with richer material that could shed light on the impact of the structural conditions on the teacher educator role. Western classical music is a genre that is steeped in ideology and values and strongly connected to the conservatory tradition. Combined with a curriculum and a final exam, this course is an example of a context that is fairly determined and in dire need of rethinking to be in line with what a modern music teacher needs. Juxtaposing this case with the more empty canvas of the sessions at the second institution sheds light on the conditions for student participation and the role of the teacher educator as a facilitator of student participation under various circumstances.

Theory
To situate this study, we begin with a review of the literature on the two concepts at the heart of this research: student voice and participation.

Conceptualising student voice
Educational participatory action research is predicated on the conviction that the voices of students are essential to the conversation of how, what and why we teach. Although the discourse of student voice has grown over the past few decades, particularly in P-12 settings, a clear conceptualisation of student voice is lacking (Charteris and Smardon 2019;Seale 2010). Scholarly works sometimes offer 'under-theorized, unproblematic accounts of student voice' in educational research (Spruce 2015, 285), thus adding to the lack of clarity about the nature and purposes of student voice. Broadly stated, student voice work proposes to 'empower students to engage in dialogue and action' (Pearce and Wood 2019, 8). Yet, the emerging discourses of student voice present multiple, and sometimes conflicting, student positionings and ideological purposes (Charteris and Smardon 2019). Seale (2010) outlines three key expressions of student voice in the educational setting: listening to and valuing students' views of their learning experiences, positioning students as equal partners in the evaluation of teaching and learning and, ultimately, utilising their views to enact change. As such, student voice in educational settings can offer 'a rupture of the ordinary' (Fielding 2004, 296) and an opportunity to step outside of the performative culture of schooling (Bragg 2007;Keddie 2016;Chua 2009). Pearce and Wood (2019) go further and suggest that student voice is grounded in developing 'tools for transgression' (21) that afford students a unique sense of ownership over their educational experiences. Student voice work at the P-12 level has grown; however, discussions of student voice in higher education contexts present a more conflicting picture. Some scholars question whether student voice work in higher education will necessarily lead to action and institutional change, given the oppositional nature of structural institutional forces (Harvey 2001;Shah and Nair 2006;Seale 2010). Sperling and Appleman (2011) propose that 'voice' is the product of two interacting sources: the specific context (social order and hierarchical relations) and 'the individual person speaking or writing here and now, projecting onto the words his or her own slant and thereby adding to the cultural and historical possibilities of those words' (74). Although this understanding suggests the individual's ability to exercise voice in opposition to institutional forces, it also suggests the difficulty of students achieving a sense of empowerment in a setting that has prescribed hierarchical identities and structural policies that exist beyond the immediate classroom environment (Taylor and Robinson 2009). This can make the promise of empowerment and institutional transformation elusive. Furthermore, students working in an institutional setting may-out of necessity, convenience or comfort-simply 'reproduce the dominant voices' of that space (Pearce and Wood 2019).
All these considerations and challenges were evident in this study.
'Participation' in a participatory action research methodology The notion of participation in an educational setting is intimately connected to the aforementioned ideas of student voice. However, similar to student voice, what constitutes participation often changes with context and intent. Educational scholars identify participation in a spectrum of five stages that range from merely keeping students informed to actually creating opportunities for learners to 'set agendas for changes [in which] self-organisation and responsibility for management is given to the learners' (Després and Dubé 2020, 3). In participatory action research in an educational setting, the notion of participation requires a renegotiation of the roles of both students and teachers-a comprehensive rethinking of our ways of being together in our educational spaces. For this reason, it offers a unique view of what it means to be 'in dialogue' (Freire 1970, 87). From the students' perspective, participation in participatory action research work speaks to an opportunity to impact the content and processes of learning by offering 'their own meaningful thematics and generative themes' (Freire 1970, 87), that is, the ideas and concerns that grow out of the students' lived experiences as young people and future music educators. As such, this notion of participation situates the student as a key player in the design and execution of learning in a collaborative partnership with the teacher (Somekh 2006). Participatory action research work seeks to shift power hierarchies inherent in the common teacher-student dyad and allows for new expressions of 'equality, democracy, and reciprocity' (Herr and Anderson 2005;Huisman 2008). The result is the creation of a space of 'co-learning, participation, and transformation' (Greenwood, Whyte, and Harkavy 1993). An explicit aim of participation in participatory action research work is to empower students 'through the process of constructing and using their own knowledge' to act in and upon their immediate world (Reason 1998, 71).
How does this understanding of participation find expression in the changed roles and experiences of the teacher? First, it situates the teacher as a partner with her students. No longer the sole voice of pedagogy, the teacher is tasked with finding new ways of being in the educational space that allow a multiplicity of voices to impact the pedagogical experience. This inevitably means that the design and content of learning will be constructed, for the most part, collaboratively and 'in process' rather than being based on teacher-determined methods and procedures. The key tool for the teacher, therefore, is her ability to engage in 'radically open listening' (Haynes and Murris 2011, 286): being able to recognise the generative themes offered by her students and to accompany the students in an exploration of these themes and their implications. This can create a sense of uncertainty, ambiguity and disequilibrium for the teacher. Participation is messy. Yet, as Ellsworth (2005) notes, it is this openness to the uncertainty of the pedagogical process that can afford glimpses of 'pedagogical anomalies' (5)-that is, radically new ways of understanding and executing the shared goal of teaching and learning.
Given these considerations, it is clear that engaging in participatory action research in an institutional setting can be fraught with contradictions, paradoxes and challenges. How does this type of participation find expression when certain 'curricular goals' (material to be covered, exams to be administered) remain present? How do students and teachers negotiate an open agenda when policies related to 'official knowledge' in degree programmes are ever-present?
These and other contradictions, paradoxes and challenges to participation are at the core of this article.

Methods
This article is the result of a sequence of methods. The framework for the data collection was participatory action research, as a result of the broader research project of which this study is part. (Other results are presented elsewhere; Onsrud et al. 2022.) As the emphasis in this article is on the teacher educators' role in facilitating participation and student voice, we employed self-study methodology and narrative analysis to analyse and present our findings.

Participatory action research as a framework
The study began as a participatory action research study among preservice music teachers at the bachelor's level in five-year generalist teacher education at two universities in Norway from January to April 2020, with the goal of enhancing student participation in education. Each portion of the study consisted of six sessions spaced out to take place approximately twice a month. Due to the COVID-19 pandemic, the last sessions had to be conducted online. All sessions were documented on video recordings and transcribed. The preservice music teachers documented their individual reflections in written logs after each session. We, as researchers, also wrote logs and held reflection and evaluation meetings with the participants between sessions. These meetings were also recorded.
General teacher education in Norway is twofold: one programme focuses on students in grades 1-7 and the other on those in grades 5-10. The study was conducted in groups from both gradelevel programmes in both institutions involved, resulting in a total of 24 recorded sessions. We (Silje and Bendik) followed the action research spiral of planning, acting, reflecting and evaluating before planning new sessions based on the experiences of the previous cycle (Pant 2014). While Silje and Bendik conducted the research sessions with preservice music teachers, Hanne and Judy served as critical research partners in what is sometimes referred to as critical friendship. 3 Their role was to maintain a certain distance in the critical reflection on the progress of the cycles and to provide their feedback in meetings with the researchers between the sessions. According to Pant (2014, 4), participatory action research reflects an equal collaboration between researcher(s) and other participants. In our study, there was variation in the degree to which planning was collaborative. Some of the sessions were planned by the researchers, based on reflection sessions with the preservice music teachers or their logs, while other sessions were planned solely by the preservice music teachers without the involvement of the researcher.
The safe and legal collection, storage and sharing of data were approved by the Norwegian Centre for Research Data. 4 All data were anonymised. The participants signed an agreement and were informed of their rights to know how their contribution was being used and to withdraw from the study at any time.

Self-study and narrative analysis
In this article, our focus is on the researching educator's role in facilitating preservice music teachers' active participation in the content and procedures of their education. We consider this a form of self-study (Vanassche and Kelchtermans 2015) to draw attention to the researcher's selfreflexivity. Self-studies are often combined with action research, narrative inquiry and/or autoethnography (Fletcher 2020;Tidwell and Jónsdóttir 2020), and in addition there are examples of participatory action research being combined with narrative analysis (Holmgren 2022). We employed narrative analysis, or what Polkinghorne calls narrative configuration, in which Silje and Bendik present their experiences of the sessions and '[organise] the data into a coherent developmental account' (Polkinghorne 1995, 15). Kitchen (2020) emphasises self-study as an empowering way of examining and learning about practice while simultaneously developing opportunities for exploring scholarship in and through teaching. Despite the use of the word 'self' in self-study, the methodology favours some form of collaboration (Bullock 2020, 270). For example, Mena and Russell (2017) stress the use of critical friends to help review data, challenge assumptions and suggest additional perspectives. In particular, they regard critical friendship as an important way of explaining why a self-study might be considered trustworthy, as a self-study that includes critical friendship involves perspectives other than those of the researching teachers, resulting in a form of triangulation that hinders bias. In this study, Hanne and Judy took part in the discussions between the interventions and provided their perspectives on the preliminary results and how the action research should proceed. Their short reflections on the narratives serve to expand on and triangulate the narratives.
Narrative configuration is not merely transcription of thoughts and actions, 'it is a means of making sense and showing the significance of them in the context of the denouement' (Polkinghorne 1995, 19). In the preliminary analysis phase, we (Silje and Bendik) read the transcriptions and logs in light of the research question, watched the video material and looked for meaningful events where we, as teacher educators, had an impact on student voice and participation. In particular, we looked for conflict, both as antagonism in the classroom and as inner conflict that we felt as researchers and teacher educators during the sessions. For this reason, the narratives are also based on and express our emotions, thoughts and interpretations (Chase 2005).
In the second phase, these events were related to each other and combined to form a story, meaning that the narratives are both the process and product of the analysis. As life is not neat and formulaic (Caine, Estefan, and Jean Clandinin 2013), the narratives will not necessarily confine to the standards of a 'good story', but we have attempted to turn the events into some form of plot structure that moves towards a denouement, using the five elements of plot structure Ollerenshaw and Creswell (2002) identified: setting, character, problem(s) faced, actions taken and resolution. The narratives also include quotes from the preservice music teachers, taken either from their individual logs or from the dialogues recorded during the events themselves or the reflection sessions afterwards. The narratives do not cover the entirety of what was uncovered during the sessions, but they are indicative of how the teacher educators influenced student participation in this study.

Narrative 1
Bendik's story: My research took place in a course on the history of Western classical music. Western classical music is not mentioned specifically in the curriculum for Norwegian teacher education, but the students are required to have knowledge about a broad palette of musical styles. At this particular institution, there is an obligatory written exam in the history of Western classical music. There were several restrictions as a result of the setting, such as a learning standard, a defined curriculum and a final exam. These restrictions obviously narrowed the space for change and participation, at least compared to Silje's study. The preservice music teacher group for grades 1-7 (Group 1) consisted of 6 preservice music teachers, whereas there were 18 in the group for grades 5-10 (Group 2). There were other differences as well, which we will return to below, but both groups included participants with prior knowledge of music history and those who were complete novices. The examples shared here come from the first two sessions on campus and help illustrate how choices were made and how participation was initiated.
The first sessions were quite similar in both groups. First, I gave the students information about the course and the research project. I then invited them to participate by presenting them with opportunities for how we could proceed, and informed them of which elements of the course could be challenged, such as content, teaching methods, the course's status as an isolated subject and forms of assessment. The course requirements and a few relevant competence goals from the school curriculum were also shared. In general, many of the students were positive about taking part in the development of something new, but there were also a few students who were critical and preferred regular lectures. After the introductory session, the two groups chose to proceed in different ways.

Group 1
As this group was quite small, there was room for everyone's voice to be heard. Interestingly, those who spoke the most, thus more or less representing 'the class voice', were those who had little prior subject area knowledge and said they felt unsure of their own abilities. When we discussed how to proceed, the discussions often took the form of negotiations in which we agreed on what to do next. During the second session, we sang a mediaeval song, which turned into a nice session in which we discussed both pedagogical and historical matters. At the end of this session, as we planned the next session, one student suggested that we work with a storyline, a teaching method that involves telling stories.

BENDIK
Yes. Shall we say that in the next two lessons, we will do something like that, which you can use for Baroque music? Does that sound okay? This excerpt is indicative of the exchanges I had with the participants during the course; only a few voices are audible here, but more students took part, although two or three were particularly active. I had no experience working with the storyline method but assumed that the preservice music teachers did. As a result, I probably gave them too much responsibility to carry out the next sessionmore responsibility than they themselves realised they had been given. While the in-class planning went reasonably well, the results in the following session were less successful. The preservice music teachers found it difficult to read up on the topic by themselves and my knowledge of storyline was insufficient, so we ended up not doing storyline at all but rather working with the material in classroom discussions. Some of the students in this group also expressed that they wanted me to teach instead.

Group 2
At the end of the first session in Group 2, several decisions of importance for the rest of the course were made quickly, but not necessarily with extensive participation. In contrast to Group 1, the strongest voices in this group came from students with previous knowledge of music history. This meant that they also had experienced what music history teaching typically looks like. Even though I offered them various alternatives for challenging the subject, just as in Group 1, the sessions ended up being quite traditional music history lessons. Some important decisions were made quickly, and at the moment, it felt participatory. BENDIK Should we work chronologically? Would you like to do that, or should we break it up a bit? I need some feedback.
STUDENT 1 I think, in my mind, it would be clearer if it were a bit chronological. But that's just me.

BENDIK
Yes. We can start with that. Because then we can, for instance, take an era that I will assign to someone. Now, you will teach the others about this. … We can divide ourselves into groups. Let's say you have an assignment and work on that music. Or would you like to go about it differently? STUDENT 2 I feel like a group is a really safe sort of way to do it, and we can do this together. This short exchange was the catalyst for the design of the next two sessions, and at that moment, I was largely unaware of how leading my questions actually were. The idea of working chronologically, dividing the class up into smaller groups and letting them teach came from me, with a few students constituting 'the class voice'. In the following sessions, the student groups were each given an era on which to present, and I mainly observed. Their presentations were quite traditional-a combination of lectures using PowerPoint and music examples, as well as some very good practical tasks. In the following session, I suggested that the remaining groups give feedback to the group after its presentation. This turned into a form of assessment, with myself as examiner, as I also gave feedback. Given that I was supplying the groups with supplementary information after the presentations, I must have appeared as an expert teacher, maintaining the teacher-student dyad. Another contributing factor to this less-than-ideal participation was impatience on my part. I rushed decisions, possibly because of the curriculum but also because the participatory aspect appeared to be accounted for in my mind.
After working with both groups, I was left with the feeling of not having succeeded. I realised that I had come to these sessions with my own idea of what participation should look like, and the results did not fit my unspoken preconceptions. An important part of this resulted from my impression that the preservice music teachers were reluctant to think freely and without constraint. In addition, the upcoming exam seemed to play a significant role. The preservice music teachers were anxious about an exam they knew little about, other than that it would be based on the course and on content knowledge that to them seemed vague and unsettled. It is likely that the impending exam also affected my impatience, as I felt responsible for their learning. Due to the pandemic, the last three sessions were online. As I was experiencing a great deal of time pressure, I succumbed to giving traditional lectures. The participants, especially those in Group 1, seemed happy with this, but my own feeling of not succeeding grew more pronounced.

'The impatient manager'
Hanne's reflection: After reading Bendik's narrative, I imagine him here as a supplier of ideas and as an effective and even slightly-using his own word-impatient manager. He constructively supplies the students with ideas for how they can participate and he manages the implications of the short student responses effectively and in detail. Another metaphor that appears to my mind is seeing participatory action research as a ball game in which Bendik controls the ball during a large part of the game.
It is interesting that Bendik considers the participatory aspect to 'be accounted for', as he has organised the students' teaching. It appears to him to be in line with participatory action research and 'at the moment it felt participatory', as he writes. Similarly, we will see that Silje finds that positive feedback from the students leads her to think she is succeeding in doing participatory action research. Thus, both researchers experienced some features associated with participation in their sessions: approval and enthusiasm, as well as simply the appearance of student activity. Ellsworth (1989) criticises strategies such as 'student empowerment' and 'student voice' for giving the illusion of equality while leaving the authoritarian nature of the student-teacher dyad intact (306). Thus, the good intentions of making spaces for participation through participatory action research could actually contribute to hiding notions of privilege and silencing some student voices (301)(302)(303)(304)323). This is probably a danger when participatory action research is conducted within institutional frames. However, with this study, our aim was to do exactly the opposite-to disclose these mechanisms and persevere in the strive to enhance student participation and student voice in teacher education in spite of these barriers.

'The conflicted gatekeeper'
Judy's reflection: Opening spaces for student voice in an institutional setting is fraught with obstacles that require a unique set of navigational skills. Bendik and his students, even before they begin, were up against the constraints of course expectations, learning standards, curriculum goals and a final exam. It is therefore not surprising that the students who became the 'class voice' in Group 2 were those with the strongest background experience-those who knew 'what music history teaching look(s) like'. Sensing these external expectations, the students advocated for a traditional approach, even when given the option to do things differently. We do what we know, especially when we know the stakes are high. This begs the question: Can students' voices ever really be amplified within an institutional setting?
Yet, throughout this episode, Bendik struggles to do just that-to offer options, to ask for students' suggestions, and to amplify students' voices. He is placed in the untenable position of trying to relinquish control and help students feel confident in influencing their own learning while keeping his third eye on the prescribed goals of the course. For this reason, I see him as a 'conflicted gatekeeper'. He tries, with his students, to chip away at the institutional wall that he himself is charged with holding up.
In the end, this reality and his heartfelt desire to empower his students leads Bendik to conflate student decision-making (Group 2) or student total control (Group 1: storytelling episode) with participation, both of which fail to produce the type of 'rupture of the ordinary' (Fielding 2004, 296) that epitomises a participatory action research methodology. We are left with the question of how we and our students can recognise and encourage acts of participation-rupture and 'transgression' (Pearce and Wood 2019, 21)-if we have never experienced them before.

Narrative 2
Silje's story: I was given access to a group of preservice music teachers to conduct a participatory action research project as a supplement to their ordinary classes, aiming to involve them in the development of their education. Instead of using their usual classrooms, I invited the students into a room with sofas and a fireplace. I was hoping the room would create a more relaxed atmosphere for dialogue, so that the students could share experiences from their worlds and articulate their thoughts and needs. We started the first session by telling each other our musical life stories to get to know each other better. I asked the students to share their music education experiences from their own school time as examples of what they considered good or bad music teaching, as well as experiences that might have had an impact on their own decision to become music teachers. Inspired by Freire (1970) and Boal (1993Boal ( [1974), I invited the students to make tableaus of the situations they had shared with each other, which became our starting point for a dialogue about music teacher education. I also asked them to keep a log of their thoughts and reflections for the duration of the project.
The feedback on this first session was quite positive. In their logs, the students wrote: 'I think I can learn a lot from this way of working', 'It's different from what we have done before', 'It was nice to be in a different environment than usual. It made it easier to come forward with one's own opinions and perspectives', 'I'm looking forward to being part of this'. This positive feedback made me think I was succeeding, even though I was in control throughout and one could question whether the nature of the students' participation was that much different from ordinary teaching.
At the end of the first session, I asked the participants to write in their logs about the challenges they had envisioned in their future music teaching. The main challenges they wrote about were: 1) digitalisation, 2) the overabundance of potential music repertoire, 3) adapting music teaching for diverse groups and 4) advocacy for the music subject in school. During our second session, I put these issues up on the whiteboard and asked the students to dig deeper into what these issues consist of and how they might address them so as to feel more prepared for their future teaching. The students suggested different tools and sources to inform their work on these issues. These included imagined cases from the music classroom, online sources of music teaching material, relevant literature, educational documents and workshops of different kinds. Then, I tried to arrange who would take responsibility for the different parts and when to schedule each issue and activity. Due to my own discomfort with the chaos during the brainstorming, I probably rushed a bit faster than necessary to get things organised, perhaps even hindering some students from speaking up and being heard.
After sharing my experience of these first two sessions with my research team, I was challenged to give up some control and let the students take over more of the planning and content of upcoming meetings. However, while I tried to follow this advice in the third and fourth sessions, some of the students took only partial responsibility for the progress and expected me to take the lead. Others took full responsibility for planning and leading sessions. I found myself in a situation where, on the one hand, I had to be careful not to maintain the typical teacher-student dyad, while, on the other, I had to avoid taking on too much of an observer role so that I could be a participant on equal terms with the students. The following quotes from the preservice music teachers' logs show some of their own thoughts after four meetings: 'Creating tasks for the sessions gives us more responsibility and we have to prepare. But it also gives us the freedom to choose themes we find relevant and important to discuss', 'I think the cases [made for discussion] were very good and relevant. I really liked it that my classmates had made the cases and that they were so close to real-life situations'.
In the middle of March 2020, after our first four meetings, conditions beyond our control began to seriously affect the continuation of the study. Due to the COVID-19 pandemic, our sessions, as well as all other teaching, went online. At first, I thought our use of Zoom went surprisingly well. According to the participants' logs, they also found the use of the Zoom application to be successful. Some even found it more effective than live sessions because it made us focus better and we were able to finish our sessions sooner. After a couple of digital meetings, however, I realised that fewer students were actually participating in the conversations. Some did not turn on their cameras, and some did not show up at all. Doing participatory action research online generated some new challenges to navigate and for which we had to find tools.
After completing all the planned sessions for which the preservice music teachers had been tasked with deciding on the content and organising the activities, we still had one date open for a final session. I suggested that the participants create dream scenarios for their education, which we could discuss before having a final session to reflect on and evaluate the whole study experience. The participants supported my suggestion and none of them objected. They presented some interesting dream scenarios and spiced up with some critiques of the music programme. I provided supportive comments on some of the ideas they brought to the table and asked questions to clarify some of their other points.

'The balancing artist'
Hanne's reflection: The picture Silje's story calls to my mind is that of a balancing artist trying to balance between leading and not leading, between participating and not participating too much, between being an observer and being a teacher. Further, to continue with the metaphor of the ball game, it appears as if Silje, in contrast to Bendik, tries to handle the ball without touching it. These metaphors illustrate the uncertainty, ambiguity, disequilibrium and 'messiness' that characterise student voice work, especially the teacher role in this work. Yet, it is exactly this openness to uncertainty that may afford glimpses of 'pedagogical anomalies' (Ellsworth 2005, 5) that create spaces for participation in education. Still, the situation described in Silje's narrative raises a number of questions. Is it at all possible for Silje, with all her knowledge and preconceptions, to start a process of rethinking music teacher education on equal terms with the students, without expecting the session to have certain outcomes? Is it possible to do participatory action research in an institutional setting by just wanting to be open-minded and engage in 'radically open listening' (Haynes and Murris 2011, 286)?

'The reluctant host'
Judy's reflection: Silje's setting is quite different from Bendik's classroom. She and her participants have the option to construct an experience that is outside of institutional framing, given that it happens outside of any prescribed course. So, from the first session, Silje and her students accept their space as a non-traditional, alternative one where new things can happen and new ideas can be floated and explored. By designing non-traditional learning activities-musical life stories, tableaus and logs-Silje puts a focus on student voice and experience. In doing so, she encourages her students to think outside the institutional box, as well as to take ownership of the processes and outcomes of their work together. The experience becomes exploratory rather than prescriptive. The students offer their own 'meaningful thematics and generative themes' (Freire 1970, 87).
Nevertheless, Silje remains haunted by doubts regarding her own positionality as the 'host' of this new space. She struggles with how to position herself as a member of the group, whether her students want or need more guidance and structure and how to guide them along their own self-designed exploratory path. Silje's 'discomfort' (Burdick and Sandlin 2010, 119)-a hallmark of participatory action research-is visible throughout the episodes she retells and her reflections on those events. However, rather than lean into this discomfort and the disequilibrium that it generates, both Silje and her students struggle against it and attempt to alleviate it by 'getting things back into order' and 'expecting [her] to take the lead'. This raises the question: How does one construct pedagogy that teaches towards ambiguity, openness and the emergent unknown? How does one pass those dispositions on to students who perceive us, their teachers, as guardians of the known and unambiguous?

Discussion
As Sperling and Appleman (2011) suggest, student voice work is not enacted in isolation. It unfolds within a context of social order and hierarchical relations. Both the students and the researchers in the two cases seemed to be strongly aware of the invisible institutional and hierarchical restrictions that governed their ways of being together. The students were invited to contribute to change their own education (Sperling and Appleman 2011, 74), but the structural policies that existed beyond the immediate classroom environment (Taylor and Robinson 2009), such as the looming exam at the end of the music history course, may have hindered them from freely imagining and sharing new ideas. As the students come from a somewhat performative school culture (Chua 2009) prior to teacher education, it is not surprising that they were concerned with the exam; it is what they have been taught. Similarly, some of the students may, out of convenience or comfort, have been tempted to simply 'reproduce the dominant voices' of the educational space (Pearce and Wood 2019). In Silje's case, they wanted to take (more) responsibility for the planning and content of the course only partly, and they wanted Silje to take the lead. In Bendik's case, the students in the second group just wanted to continue to do music history as usual and take a chronological approach using the traditional mix of PowerPoint presentations and music examples.
Although the circumstances of Silje's sessions are radically different from Bendik's, there are some striking similarities. In both cases, we see that the students want the teacher educator to take the lead and teach. This finding supports the notions of Harvey (2001), Shah and Nair (2006) and Seale (2010), who question the effect of student voice work on educational change in higher education due to strong structural institutional forces. Within the confines of regular teaching, this is not surprising, but these hierarchical relations seem to persist, even when the sessions are voluntary and extracurricular. It seems as if the presence of the teacher educator, even when she tries to be an equal participant, is enough to maintain the hierarchical order of higher education.
The renegotiation of the roles of both students and teachers with the aim of empowering students to engage in 'dialogue and action' (Pearce and Wood 2019, 117) seems to be an undertaking that requires time and investment. The hierarchical structures inherent in music teacher education cannot be simply ignored or swept aside. Recall, for example, how in Bendik's study, the students with previous knowledge of 'how music history is done' became the 'class voice'. Recall how Bendik, without even noticing, readily provided supplementary information after the student presentations, thereby acting as an expert and examiner. Such subtle micro-exchanges-and with Bendik being unconscious about their impact-serve to maintain the classical teacher-student dyad. It is no accident that this happened during the sessions in which the students taught music history traditionally, as this pedagogical situation in itself is hierarchical, with predefined roles and usually a disparity in subject knowledge.
These hierarchical structures, along with institutional requirements and structural policies, are important barriers that hinder student participation. However, the insecurity that both researchers showed while working in a participatory action framework may have offered ruptures of the ordinary (Fielding) and contributed enough to destabilising these same structures to afford some glimpses of 'pedagogical anomalies' (Ellsworth 2005, 5): glimpses into radically new ways of enacting teaching and learning. For example, the 'future challenges' generated in the student logs in Silje's case could be seen as 'generative themes' (Freire 1970) that genuinely came from the students themselves, although the idea of having students write logs came from Silje. Thus, this braveness, this 'openness to uncertainty' and the opening up of spaces for the students' voices by both educators are arguably the keys to promoting the students' participation. Further, the student logs may have functioned as 'tools for transgression' that afforded students a certain sense of ownership over their educational experiences (Pearce and Wood 2019, 21).
A central aspect of the renegotiation of roles that participatory action research demands is the teacher's and students' rethinking of the notion of participation. Després and Dubé (2020) outline a five-stage continuum of participation in educational settings. Merely keeping students informed of teacher decisions is at the lowest stage on the continuum. Next is consulting students on the efficacy of teacher decisions, followed by involving students in the decision-making process. Stage four of the continuum involves collaborating with students on 'all aspects of the decisionmaking process'. The final stage reflects empowering students to 'set agendas for change, selforganize and [take] responsibility over the management' of their learning (Després and Dubé 2020, 3). This fifth stage is most aligned with the goals of participatory action research.
All of these stages are evident in the cases of Bendik and Silje. Bendik informs the students of the aspects of the course that must remain intact (the content and the exam, for example). He offers the students options for how to proceed and consults with them regarding what to choose; he involves and collaborates with them in the decision-making process on how to organise the learning. Silje's case is more open, given that the meetings take place outside course constraints. She also involves the students in decision-making. She designed the core elements-the musical life-stories, tableaus, logs and dream scenarios-in advance with little collaboration from the students, but using these logs for the specific purpose of documenting the 'challenges they envision for future music teaching' paves the way for empowering the students to 'set agendas, self-organise and take responsibility' by generating themes that were explored in later meetings.
Both of the researchers struggle with recognising the type of participation that they are striving for, that they might offer their students and that the students are embracing or rejecting at a given moment. They are both attentive to expressions of informing, consulting, involving and to some degree collaborating, and they tend to assume these to be the participation that they are hoping for from the participants. (As Bendik writes, 'The participatory aspect appeared to be accounted for'). Students also gave 'positive feedback' on the type of participation that Silje offered them, leading her to initially conclude that her efforts 'were succeeding'. This raises the question of whether educators and students can even recognise ultimate expressions of participation in participatory action researchthat is, student empowermentif they have never experienced it before. It appears natural to fall back on understandings of participation that are familiar and may be lacking in normative teaching and learning and thus appear novel by comparison.
It is also interesting to note that in those moments when empowerment became viable, both Silje and Bendik found themselves uncomfortable (Burdick and Sandlin 2010). Silje reflected on her 'discomfort with the chaos' and her desire to 'get things back into order'. Bendik found himself feeling 'impatient' and responded by assigning tasks in an attempt to bring order to the students' own confusion over the new responsibility they were offered. This points to another demand of participatory action research: the need to lean into the uncertainty, ambiguity and disequilibrium of the moment and take the time to negotiate these openly with all who are present in the group. This involves a level of vulnerability and risk-taking on the part of both teachers and students. Silje and Bendik's accounts are filled with exactly such moments of risk-taking and being vulnerable with their students. The success of the research cases and the compelling findings generated are in great part due to the researchers' willingness to 'go there'. Unfortunately, the students did not always go with them.

Implications
This study started with a participatory action research framework, and although we failed to conduct the method comprehensively, the research points to several challenges to student participation. In Narrative 1, it is apparent that institutional constraints, such as an exam, can hinder students from thinking freely and envisioning new possibilities for their own education. Institutional constraints were also experienced by the teacher educators, as they felt responsible both for their students and for course requirements. This raises the question of whether it is even possible to conduct participatory action research in a setting limited by curricular goals and within a relatively short timeframe, as presented here. Both researchers bravely jumped into the tenuous landscape of participatory action research without any previous practical experience of this research approach, but with a strong desire to create a space for the students' voices. They felt the urge to move forward, which was visible in their impatience and discomfort with chaos. It is likely that sentiments such as these thrive in an institutional setting.
The sessions in this study sought to shift the power hierarchies inherent in the student-teacher dyad. Both Bendik and Silje found it difficult to find the right balance, as they occupied several roles, which have been labelled in the reflections following the narratives as 'the impatient manager', 'the conflicted gatekeeper', 'the balancing artist' and 'the reluctant host'. Each of these roles affected the preservice music teachers' participation in various ways. At times, the researchers themselves inadvertently claimed the teacher role through more or less subtle exchange with the students, and at other times, they were placed there by the students. This indicates that in participatory action research and in other proposals for change in teacher education, the role of the teacher educator needs careful consideration. What should be of interest for studies similar to this is that even in an extracurricular context, where the students have volunteered to take part in a project with the aim of enhancing student participation, we see that the teacher educator is expected to be the leader. The imbalance in the teacher-student dyad is further amplified by disparities in content knowledge. This is mainly a problem when participatory action research takes place in a setting with prescribed content knowledge, as in Narrative 1. Longstanding topics in music education, such as the history of Western classical music, also come with a pedagogical tradition, a sort of script for how the subject should be taught. This structural aspect further cements the teacher at the top of the hierarchy and needs to be accounted for in similar studies.
Nevertheless, in both cases, we see indications of new possibilities that emerge from the students' participation. A challenge for us as teacher educators is to recognise what the students bring to the table, even if it does not fit our expectations.

Notes
1. This study is part of the research project FUTURED (2019-2022), a collaboration between HVL and OsloMet that is funded by the Norwegian Research Council. 2. The term 'session' includes both the regular teaching lessons in the first institution, and the extracurricular meetings in the second institution. 3. We describe this concept more thoroughly in the following paragraphs when we connect our methodology to self-study in teaching and teacher education. 4. https://www.nsd.no/en, approval number 294194.

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

Funding
This work was supported by the Norwegian Research Council under grant number 288436.

Notes on contributors
Bendik Fredriksen is an associate professor of music at Oslo Metropolitan University, Department of Primary and Secondary Teacher Education, where he teaches music within generalist music teacher education. He holds a PhD in music education from the Norwegian Academy of Music and has an MA in musicology from the University of Oslo. He has worked as a teacher in Norwegian compulsory education for several years and has conducted children's and youth choirs. Dr Fredriksen is currently engaged in the FUTURED (Music Education for the Future) (2019-2022) research project. His research interests include generalist music teacher education, music teacher turnover and the philosophy of music.