The multiplicity of preservice music teachers’ positioning in a participatory action research project

This article reports the results of a participatory action research study into Norwegian generalist music teacher education, that intended to develop spaces for preservice music teachers to foster agency and prepare for future teaching. We aimed to challenge the discursive practice of generalist music teacher education through participatory action research conducted from January to April 2020 at two central teacher education institutions in Norway. In this article, we present extracts from transcribed video recordings of the completed participatory action research that identify preservice music teachers’ positioning in interactions as a response to the challenges posed by action research events. Through our analysis, which draws on positioning theory from discourse psychology, we identify three primary positions taken up by preservice music teachers: (a) novices, (b) not yet independent, and (c) resource persons. The study identifies a need to interrupt traditional music teaching as a discursive practice that maintains power relations that obstruct preservice music teachers’ agency in their education. We conclude that more systematic long-term work is needed to change both educator and student habits and mind-sets.


Introduction
A challenge for music teacher education is to integrate past, present, and future possibilities in an effort to refine relevant knowledge from the past and adjust content and practices to future needs. As Elliott and Silverman (2020) suggest, music teacher educators must "read the future" and consider what today's preservice music teachers require for meeting the needs of tomorrow's pupils (p. 81). At the same time, scholars have addressed a certain resistance to change in music teacher education due to the dominance of strong conservatory traditions (Gaunt & Westerlund, 2013;Sandberg-Jurström et al., 2021;Väkevä et al., 2017). The same tendencies can be seen in Norwegian generalist music teacher education (Saetre, 2014). We therefore deem it necessary to "remix the classroom" (Allsup, 2016) and ask ourselves whether we are teaching traditions or students.
The main objective of the study discussed in this article 1 was to develop spaces for fostering preservice music teachers' agency to prepare them for future teaching. Other scholars have addressed the need for stronger teacher and student agency in music education (Allsup & Westerlund, 2012;Powell, 2019) to move beyond the circle of reproducing the same practices (Bowman, 2007). More recently, Conway et al. (2020) have encouraged music teacher educators to create atmospheres in which "future teachers are actively pushed to develop an agentic, forward-thinking teacher identity" (p. 914). Tucker (2020) responds to this encouragement by providing a framework for understanding and supporting the development of student agency using Emirbayer and Mische's chordal triad of three components: iterative (past), practicalevaluative (present), and projective (future) dimensions of human agency. Biesta and Tedder (2007) also draw on Emirbayer and Mische in their outline of an ecological understanding of agency, particularly when emphasizing the teacher's "critical judgement" as part of agentic learning and suggesting that teachers loosen themselves from past patterns of interaction and reframe their relationships within existing constraints (p. 139).
In earlier research, Powell and Parker (2017) have found that music student teacher identity and agency are inherently intertwined, which indicates that the phenomena may be best studied concurrently. Other studies on music student teacher identity have found that students tend to view themselves as either musicians or teachers (Ballantyne, 2006;Ballantyne & Zhukov, 2017;Bouij, 1998;Hargreaves et al., 2007;Pellegrino, 2009). There are also studies suggesting that students identify more as musicians and performers in the beginning of their education, but gradually develop a teacher identity during their education (Ballantyne et al., 2012;Georgii-Hemming & Westvall, 2010;Kenny, 2017;Kos, 2018). Although these studies do not address agency as part of student identity, Tucker and Powell (2021) have recently investigated how preservice music teacher values are inseparable from identity and agency. Their study presents values and beliefs students bring with them into their education and how they are influenced by their professors' values within instrumental music teacher education. Findings indicate a lack of consistent agentic identity development across participants, and the researchers suggest that preservice teachers' awareness of structures and agency may not be sufficient.
In addressing the challenge of developing agentic, forward-thinking music teachers capable of teaching in diverse contexts, we chose to conduct participatory action research with music student teachers at the bachelor's level, which is referred to as cycle one in the 5-year master's program that comprises the new generalist teacher education in Norway, modeled on Finnish teacher education (Skagen & Elstad, 2020) and introduced in 2017. We thereby contribute using a methodological approach that is seldom used in Nordic music education research. Internationally, action research in music education is used primarily to improve the practice of music teachers (Cain, 2012;Conway & Borst, 2001;Julia et al., 2020;Regelski, 1995), and not to promote preservice music teachers' agency in their education, as we have explored in this study. Therefore, we carried out two different series of exploratory events at two teacher education institutions from January to April 2020. In contrast to previous research into teacher and student agency, our analysis used positioning theory (Davies & Harré, 1990) and focussed on how students position themselves in the action research events. This is in line with Powell and Parker's (2017) claim that identity and agency are inherently intertwined. In this article, we explore how the positioning approach can give insight into the development of student agency. Our main research question is the following:

RQ1.
What can be learned about agency through preservice music teachers' positioning in two different cases of a participatory action research project?

Data generation through participatory action research
Two of the authors carried out participatory action research in generalist music teacher education at two central teacher education institutions in Norway, while the two others served as critical research partners to safeguard a certain distance in the critical reflection upon the progress of the cycles. Generalist teacher education in Norway is twofold: one program caters for those who want to become teachers for first to seventh grades, and one for fifth to tenth grades. In both institutions, the action research was conducted with one group from each program. All together the participants numbered 48 (23 males/25 females). We held six meetings with each group, approximately twice a month. The first three meetings took place on campus, while the subsequent meetings had to be held online due to the pandemic situation arising in March 2020. The meetings at one of the institutions took place as an extracurricular project where the students were asked to address future challenges for music in the elementary school and suggest ways to work with such challenges in their education (Case 1). In the other institution, the meetings were undertaken within the existing curriculum in music history and were subject to the corresponding conditions. The students were asked to select content and suggest ways to work with it (Case 2). Like Norwegian music student teachers in general (Nysaether et al., 2021), the groups were quite homogeneous, as almost all participants came from ethnic Norwegian middle-class families. The main difference between them was their prior musical knowledge.
The data were documented through video recordings and written logs from the students after each meeting. As researchers, we also wrote logs and held reflection and evaluation meetings between the meetings with the students. In this article, we mainly use extracts from video recordings. All data are anonymized, and students signed a form about the use of their data and their rights as participants in the research study. The safe and legal collection, storage, and sharing of data has been approved by the Norwegian Centre for Research Data. 2 We conducted participatory action research using a critical approach, which combines social analysis, self-reflection, and transformational action to challenge existing discourses and practices, as well as promote change in education emerging from the practitioners themselves (Kemmis et al., 2014). Critical participatory action research is intended to change social practices, including research practice itself, to make them more rational and reasonable, productive and sustainable, and fairer and more inclusive (Kemmis et al., 2014, pp. 2-3). It encourages change by breaking down power hierarchies inherent to the customary teacher-student or researcher-subject relationship. It also aims to empower participants through the process of constructing and applying their knowledge to act in and on their world. Freire's Pedagogy of the Oppressed has been an important source of inspiration for the development of this methodology (Orlowski, 2019).
In this study, we are not completely true to all the dimensions of participatory action research as described by, for instance, Pant (2018). Although our intention was to strengthen students' participation in their own education, we did not engage them in all aspects of the research process, from design to dissemination, as is the ideal action research approach (Pant, 2018, p. 4). For instance, the research questions and study aims were already determined when the researchers met the participants. Nonetheless, we did follow the action research spiral, as each meeting was treated as an action, followed by an immediate reflection session. This in turn served as a metaanalysis and evaluation of each action, supplemented by individual logs. However, the planning of new actions was not always done in collaboration. Some of the actions were planned by the researchers, mainly based on reflection sessions and logs, while others were planned solely by the students. In this article, our focus is directed toward how the action research we managed to produce, as well as other structures and frames, conditioned students' positioning. It is through interaction that positioning takes place and can be identified. We have therefore chosen extracts from the data that demonstrate interaction between the students and ourselves as researching teacher educators. The extracts serve to illustrate the positioning most prominent in the action research, and to prompt analysis and discussion of student agency.

Positioning in discursive practices
Teacher education can be regarded as a discursive practice. This concept was developed by Foucault (1981) and refers to the practices of specific knowledge (discourses) in which objects and subjects have historically been shaped and reshaped in interaction. A specific discursive practice, such as generalist music teacher education, is shaped by a complex net of both micro and macro discourses. The micro discourses, sometimes called interpretative repertoires (Potter, 1996), are constituted by students' and educators' interactions within institutions, and include both speech acts and attitudes. Macro discourses, which relate more to Foucault's concept of discourse, in this case denote broader educational, societal, and institutional ideas of music teacher education and also influence the discursive practice. Students and educators within a discursive practice position themselves within each of the available discourses, which means that their agency is discursively constructed. In this study, we sought to uncover and describe the positions that were taken up when an alternative discourse was introduced through participatory action research, which offered agentic positionings within the discursive practice of teacher education. Davies and Harré (1990) introduced positioning as an attempt to find a more dynamic concept than "role" to analyze and understand selfhood in social psychology. People's positioning occurs on concrete occasions of language in use. From this point of view, we can understand preservice music teachers' development of professional selves as a path of positionings within discursive practices aimed at making sense of the music teaching profession and themselves as future music teachers. Davies and Harré (1990) also refer to the fact that the production of the self takes place within multiple and contradictory discursive practices and is brought into being by speakers and hearers as they engage in conversations (p. 62). In our case, students are offered multiple and contradictory positions in their everyday teacher education experiences on campus through lectures, readings, discussions, and in school practice. Teacher education is part of a discursive practice in which certain positions are made available, which constitutes certain ways to think and act, which again enacts certain conditions for possible agency. Educational situations with contradictory requirements provide students with possibilities of choice: the possibility to act "agentically," as Davies and Harré put it (p. 59). In other words, students have agency within certain discursive frames.
When studying the transcriptions and logs from our completed action research through the lens of positioning theory, we drew on Davies and Harré's (1990) description of developing "the multiplicity of the self " (p. 47). The concept of "positioning" provided us with a tool for talking about the discursive production of a diversity of selves conjured up in the course of conversational interaction. Preservice music teachers are in the process of becoming professional selves as music teachers, and by identifying their positionings in the conversations constituted by participatory action research, we "freeze" certain fragments of an ongoing process. The positions we present in the following analysis must therefore be viewed as an artificial fixation of moments from particular situations that are part of something that is developing and moving, and should not be viewed as stigmatizing characteristics. By reading transcriptions of the data repeatedly and asking questions like, "how do the students talk about themselves and each other as becoming professional selves?," the three positions of novices, not yet independent, and resource persons gradually emerged.

Analysis of empirical examples
Through participatory action research, we aimed to challenge cultural stereotypes, such as teacher/student, to open a space for preservice teachers to act agentically and take up positions that are not usually expected of them in their education. In the following analysis, we highlight three positionings that we found to be most prominent in our research. These are the positions of novices, not yet independent, and resource persons. The three positions do not necessarily suggest a development or continuum by which students progress from one position to another. In other words, the novice position not only occurred during the first part of the participatory action research and the resource position during the last part. Instead, the three positions occurred repeatedly for students to take up and engage in.

Positioning as novices
In both cases of participatory action research, the researchers started the first meeting by claiming that the students were considered important contributors in developing and changing music teacher education for the future. We were clear about our agenda to invite them to share their own perspectives, opinions, and suggestions. The students responded to this initiative in various ways. Some found the invitation exciting, as one said, "No one has ever asked for our contribution before." This was particularly the sentiment among students in Case 1, in which participatory action research was carried out alongside regular teaching and the students were under no pressure in terms of assessment. We found the same tendency in Case 2 among students who were already well-skilled in music. Others were more skeptical, especially in Case 2 in which the action research was part of regular teaching with predetermined content and fewer opportunities for change and experimentation.
In Case 1, several students expressed uncertainty and discomfort when challenged to speak frankly about their experiences with the participatory action research, as shown in the following interaction during the reflection session after the third meeting.

Educator:
When ?" position the students as novices in the sense that they may not feel capable of contributing good suggestions for the development of their education. The action research in Case 1 challenged the students to think in a visionary manner, address their own obstacles to becoming future music teachers, and suggest ways to make their education more relevant and future oriented. Even though some of the students' statements could represent a nonagentic positioning, as when they claim not to have much to contribute, we observed that, to some degree, they took on the challenge and dared to take risks and dive into the deeper waters of their vocation. This shows that positioning as a novice does not necessarily limit students' agency, but rather enables them to face new challenges (such as this participatory action research project) according to their current capabilities. The contradiction in finding participatory action research exciting yet challenging shows a complexity and ambivalence in the novice position. In Case 2, the research was conducted during a course on the history of Western classical music with two very different groups, not only in terms of size (Group 1 was only a third the size of Group 2) but also regarding the level of activity and approach to the lessons. Here we use examples from students in Group 1, who showed resistance toward the project and expressed more pronounced positions than Group 2. The examples all originate from conversations about the method used for the lessons in which the students were encouraged to propose different approaches. They decided to work chronologically, starting with medieval and renaissance music, and wanted the researching teacher educator to find something they could sing from these time periods. The teacher selected the Hymn to St Magnus, and two of the students volunteered to rehearse the song in two voices with the rest of the group. As they worked on the piece, an interesting discussion took place that included, for instance, ways to rehearse songs with more than one voice. As the conversation developed, it also addressed other traits of medieval and renaissance music, but some students reflected that they struggled with participating in this conversation: When these two students positioned themselves as novices, they did so against an unspoken opposite. The experience of shortcoming expressed here was most likely due to the fact that there were other students in the group who could sight-read the music and had learned about music history previously, and therefore mastered the task of singing the piece with little difficulty. The silence from the other students in this exchange seemed to emphasize this difference, much like a silent agreement. These students' positioning as novices does not necessarily mean that they are musical novices, but rather that this genre was so alien to them that they could not see how they could make use of their reservoir of musical knowledge when working with the song. Even though they position themselves as novices, they still act agentically, suggesting ways to solve their lack of knowledge. Their positioning could be considered critical, as they question which aesthetic definitions should be relevant in the classroom setting. This is another example of the complexity of the position assumed by these students and of how agentic actions and speech can take different forms. At first glance, positioning as novices may seem like a nonagentic response to the insecurity and discomfort involved in participatory action research. But as illustrated through the extracts in this section, the position involves resistance, critique, risk-taking, and making suggestions.

Positioning as not yet independent
In Case 2, the positioning as not yet independent shares some obvious traits with the positioning as novices, but there are a few important differences. While the novice position results from a lack of knowledge and being new to the field of practice, the positioning as not yet independent revolves more around attitude and a need for reassurance and acknowledgment. When the researching teacher educator asked the students what they wanted to do next time, one of them answered, I prefer that you lecture [. . .] that you highlight some of the factors and then work with them: "What do you think about this? Here you have some specific examples of time periods." And so on. So that you end up with a bit more on the screen.
Many of the other students agreed, and at the end of the third meeting, the same student elaborated, It's clear that there's a lot of information available in the field since you can talk endlessly about the different composers and types of music, and it's exciting to listen to. But it would be very nice to concretise this using the screen-at least, this is my opinion and there may be others here who think the same. It's nice to have notes you can actually read. Because it's difficult to take good notes while engaging in discussion. If you only sit and take notes, you won't be able to participate in the discussion. These statements emphasize the notion of knowledge as something "out there" to be acquired. The idea of participation involves a constructive view of knowledge, which in turn requires a certain amount of independence on the part of the students. However, this student's desire for the teacher educator to point out the most important information shows that not all students consider themselves fit for this responsibility. From a broader perspective, sentiments like these might result from unclear communication from the researching teacher educator. At the very start of the course, the invitation for the students to participate in developing their own education was given with the intention to transcend the tendency to focus on factual knowledge and have the students construct more of the course content. Yet their participation was unquestionably inhibited by the immensity of the subject. A curriculum that consists of more than 1,000 years of music and an abundance of composers, genres, and subgenres is difficult to navigate as a beginner. Two of the students described the music history book as "impossible to understand," mostly because of an abundance of unfamiliar subject-specific theoretical concepts. Notwithstanding these difficulties, the excerpt above shows that instead of trying to explore the field themselves, the students wanted a guided tour.
These students' positioning as not yet independent corresponds to the banking model of education that Freire (2005) criticized for turning students into passive recipients of knowledge, into "containers" or "vessels" to be "filled" by the teacher. In banking education, the teacher teaches and the students are taught as if the teacher knows everything, whereas the students know nothing. Banking education attempts to control thinking and action, leading students to adapt to the world as it is and inhibiting their creative power using propaganda-like methods. While this model is traditionally viewed as imposed by the teacher, the aforementioned students in Case 2 positioned themselves as self-declared, voluntary, and empty vessels. Their lack of previous knowledge was, of course, very real and was unlikely to have been addressed properly during lesson planning. Yet, together with the desire to be lectured by their teacher educator, this example shows how the students reinforced the notion of a fixed curriculum-a canon that is "out there," that they need to know about-despite its key role in what this participatory action research project sought to change.
Toward the end of the course, the upcoming exam was given more attention during classroom discussions. In the first lesson, it was made clear that the students would take part in formulating the exam text and that it would be based on the content of the course. Yet the students seemed to worry about this task and asked to see previous exams. Together with the presumption that there is fixed knowledge to be acquired, the emphasis placed on the exam by the students can be considered an example of "learnification" (Biesta, 2015), in which students act as customers. Being brought up in a school system that values quantifiable output, preservice teachers demonstrate less resistance to practices such as testing and seem to consider them more important than previous generations of teachers (Mausethagen, 2013). This might explain the increased emphasis on the exam by these students. Although less pronounced, uneasiness about the exam was also visible in Group 2, where resistance to participatory action research was linked to the risk-taking experienced by the students, like customers expecting a dividend.

Positioning as resource persons
The third position we observed involved how students started to view themselves as resource persons in their education. This seemed to come as a direct response to the participatory action research and stands in contrast to the two other positions in which the students were more concerned about fulfilling a predefined role. During the last meeting of Case 1, the students participated in a group discussion on their various dream scenarios for music teacher education. The following extract is taken from the plenum conversation after the group discussions.

Student 1:
One thing we talked about was using students as resources. I think there is no better arena for doing so than at a university. After all, we are here to become teachers. The best thing would be to use those students who already have a good grasp of theory as resources to teach those who do not. I think students should really be used as resources more often.

Educator:
Yes, you bring a lot of knowledge with you to your studies. And you are training to become a teacher. One could assume it would be relevant to have a community of learning exchange. Student 3: As far as using students as resource personnel and assistant teachers in our studies is concerned, we could also organise study groups. We could arrange things so that the study groups could meet with the resource personnel present and, if necessary, with an experienced teacher monitoring and joining the various group discussions. At any rate, you could arrange to have groups using the resource persons available in the class.

Educator:
Absolutely. For example, one could use fourth-year students as resource persons for first-and second-year students, and fifth-year students for third-year students. Student 3: Yes, that sounds like an excellent idea. The fourth-year students would also get to practise teaching at different levels.

Student 5:
It would almost be teaching practice. By inviting the students to envision dream scenarios for future music teacher education, the researching teacher educator encouraged them to challenge existing positions and consider possibilities beyond the positions discursively constructed in music teacher education as they know it. In this extract, Student 1 starts by opening up the possibility for students to position themselves as resource persons in their own education. This position is strengthened by the researching teacher educator, who confirms Student 1's statement, and by other students, who follow up with statements that indicate the possibility of greater agency for students than what they have experienced in their education so far. Through interaction in the discussion, an opportunity is created for students to act more agentically and position themselves as resource persons for their education, for each other, and for the learning environment on campus. In doing so, the students enact a more active and autonomous version of agency by assuming greater responsibility in their education, as opposed to the not yet independent position observed in Case 2. This can also be described as a form of transformative agency-a concept devised by Haapasaari et al. (2014) and referred to by Lund and Vestøl (2020)-which stems from encounters with and an examination of disturbances, conflicts, and contradictions in collective activity. Our participatory action research was a disturbance in ordinary teaching and, in this last example, made the students envision new ways of thinking about themselves. This corresponds to the claim of Haapasaari et al. (2014) that transformative agency develops participants' joint activity by explicating and envisioning new possibilities.

Discussion
Several issues from our analysis provide further insight into what it takes to develop spaces for future teacher agency in generalist music teacher education. One issue is the relationship between agency and knowledge. When students position themselves as novices and not yet independent, a relevant question to ask is whether these positionings reflect the progress they have made in their knowledge development-as classroom practitioners in Case 1 and as music experts in Case 2. This is in line with earlier research on music student teachers identifying as either musicians or teachers (Ballantyne, 2006;Ballantyne & Zhukov, 2017;Bouij, 1998;Hargreaves et al., 2007;Pellegrino, 2009). In our cases, many of the students lacked identities either as musical experts/musicians or as classroom practitioners/teachers. This can probably be partly explained by the fact that these students were only in their second year of a 5-year education. Although previous research shows that teacher identity develops gradually during education (Ballantyne et al., 2012;Georgii-Hemming & Westvall, 2010;Kenny, 2017;Kos, 2018), our study suggests that students' lack of knowledge and teacher identity does not prevent them from acting agentically. If we follow Davies and Harré's (1990) description of agency as the possibility of choice within educational settings of contradictions (p. 59), some of the students in our study chose to cling to traditional and familiar approaches (e.g., lecturing), while others suggested more innovative ways to learn by relinquishing the conservatory tradition (Bowman, 2007;Gaunt & Westerlund, 2013;Sandberg-Jurström et al., 2021;Väkevä et al., 2017) as well as the traditional teacher-student dyad (Freire, 2005). These examples show that quite different forms of agency are possible in education. The question is whether these forms of agency are all fruitful and relevant for students' future music teaching in schools. Considering Barton and Riddle's (2021) argument that varied modes of communication in the classroom are important in making music education meaningful for diverse groups of learners, different forms of agency might be fruitful as well.
We have also seen that conducting participatory action research as part of the existing curriculum, as in Case 2, or outside of the ordinary music course, as in Case 1, makes a tremendous difference to its outcomes. We found that in Case 2, the space for agency was limited by the content of the course and by the fact that the course concluded with an exam. These limitations were largely self-imposed by the students, who seemed to assign them more importance than the educator did. The space for agency seemed to be more open in Case 1, where the possibilities for the students to perform the actions with their choice of content and activities were abundant. Nonetheless, not all students took advantage of the opportunities available to them to act agentically in the ways intended by the participatory action research.
This raises a third issue about developing spaces for agency: what kind of agency are we aiming for, and what kind of agency is it possible to develop spaces for? The participatory action research methodology and further scholarship (Biesta & Tedder, 2007;Tucker, 2020) offers optimistic views on participants as agentic individuals, while the theoretical lens we used in this article was more limited in identifying possibilities for agency, being dependent on resistance toward preexisting frames and structures constituted by hegemonic discourse (Davies & Harré, 1990). The development of participatory action research in educational settings has generally been influenced by Freire's (2005) suggestions for "problem-posing" or "liberating" education that resolves the student-teacher contradiction, by which "the teacher is no longer merely theone-who-teaches, but one who is himself taught in dialogue with the students, who in turn while being taught also teach" (p. 80). When using participatory action research, we therefore tend to think of agency as something liberating and empowering, first and foremost about change, while we found in our study that it can also be the students' choice to let things stay as they are. As Barnes (2000) stresses, the choice to stay inside the rules or, in our case, cling to traditional teaching must also be considered a form of agency.
It should also be acknowledged that preservice music teacher agency is not solely dependent on how we as researchers are able to develop certain spaces using participatory action research situations together. The students also bring with them certain conditions. For example, they may be formed by institutional traditions that, in recent decades, have been influenced by neoliberal thinking and values. As indicated by Mausethagen (2013), student teachers brought up in a neoliberal school system, like the students in our study, show less resistance toward practices such as testing and accountability than previous generations of teachers. The impact of neoliberal thinking is also addressed in several music education studies (Allsup, 2015;Tucker & Powell, 2021). Freire's banking concept of education can be seen as preparing "the student as customer," wherein knowledge is understood as a form of capital and students are portrayed as objects in which to invest. A view of education in terms of economic transactions is at the heart of the criticism of neoliberal policies (Apple, 2005(Apple, , 2018Gustavsson, 2012), but participatory action research represents resistance to this way of thinking and serves as a counterdiscourse to its ideological regime. With participatory action research, learners are empowered to transform the world by progressing from being silent to speaking out, from being spectators to becoming actors (Freire, 2005, p. 65). To promote a socially just and culturally diverse music education, this active competence must be seen as important. For instance, as Westerlund et al. (2021) point out, it is important for student teachers to have opportunities to engage in dialogue and learn to interact in uncertain situations to develop essential intercultural competence. Intercultural competence is just one of several competences for future music teaching that can be promoted through situations such as the participatory action research we conducted. As Tucker and Powell (2021) suggest, preservice music teachers lack awareness of the structural conditions for their education and how they can act agentically in their practice. From the students' positionings in this study, we learnt that developing spaces for agency through participatory action research and trying to make changes in generalist music teacher education is harder than expected. We conclude by highlighting three learning points.

Implications and concluding remarks
First, preservice music teachers' positionings and agency were more complex and multifaceted than we had expected. Although the students who participated in our study were homogeneous as a group, they showed varied ways of interacting based on background knowledge, values, and attitudes. The possibility of even greater complexity is something music teacher educators should consider for the future when more diverse groups of students are likely to apply for generalist music teacher education. We suggest challenging students to reflect on their own positioning as an important step in moving away from their previous pupil position toward a teacher position that is more multifaceted and future oriented. Agency advances in close relation to developing a professional identity, so taking small steps during the entire education program is needed. One example of such a step is to put agency on the agenda in all courses of the generalist music teacher education program.
Second, we experienced how difficult it can be to change discursive practices (Foucault, 1981). It takes much longer than the weeks we had available for our participatory action research to develop spaces for student agency and promote change. Music teacher educators should therefore consider, in line with the last suggestion, implementing the philosophy of participatory action research inside the curriculum during the whole generalist music teacher education, so that student agency can develop over time. We therefore suggest that the discourse of agency should explicitly be verbalized in curriculum texts, not only for music but also for all subjects in generalist teacher education and teaching practice. Because resistance and challenge promote agency, we suggest working systematically with explorative methods such as participatory action research, collaborative action research, and inquiry-based learning in teacher education. This demands solid preparation both from educators and students. In Foucault's final works (1984Foucault's final works ( /1990, resistance becomes active not against exercise of power, but in relation to oneself, through the ability to affect oneself. It opens the possibility for creating oneself in ways that differ from the present knowledge. By letting student teachers work with the question of how they could produce themselves otherwise could, therefore, be a way to promote agency.
Our final point concerns our experiences of teaching after finishing this study. We realize that our own teacher practice has changed after engaging with participatory action research. With new students, we continue to develop spaces for preservice music teacher agency and offer them alternative positions in the same way that we started with the participants in this study. This shows that this participatory action research study has, first and foremost, changed us.

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

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support for the research of this article from the Norwegian Research Council.