Au milieu des bureaux empilés (In Between Desks): A Case-study about open Discursive Spaces in Schools and their Potential to Foster Agency and Gestures of Care

This paper analyzes the socially engaged art project Au milieu des bureux empilés and investigates if it can be possible to create open discursive spaces within schools for students to collectively develop conversations about the education system and their experiences of schooling. The article outlines one possible framework to do so and focuses on investigating the repercussions of these open discursive spaces in schools were students shared instances of lack 190 Research in Arts and Education | 4 / 2021 of care. It explores how the developed conversations might have the potential to foster a certain sense of agency and of community within participating students.

cursive spaces according to how they felt regarding their environment. After outlining these responses, this article will focus on investigating the repercussions of these spaces in schools were students shared instances of lack of care. It will explore how these spaces might have the potential to foster a certain sense of agency and of community within participating students.

Literature on socially engaged art as theoretical framework
The project's framework-in reference to the structure of the project rather than its frame of reference-has been elaborated by trial and error since 2016. It is discussed and adapted collaboratively with teachers and students I meet in every school that take part in the project.
As the project goes on, I try to acknowledge as best I can every instance where the framework might have failed by generating circumstances that were preventing meaningful conversations to arise. To elaborate this framework and analyze instances where it might have failed, I referred to the literature on Socially Engaged Art (SEA). SEA is an art movement that regroups artists who work to bring art back in living contexts outside of usual art institutions (museums, galleries, studios, etc). The literature referred to in this article will concentrate on analysis of projects that try "to enact change on a larger scale by fostering creativity through dialogue" (Jordan, 2013, p. 153).
These projects aim for participant to experience their "relational self" and experiment with perceiving themselves as active members of a community, in opposition with the "individualist self" which have been asserted by Modernist Aesthetics during the 20th century. As a pioneer in this field, Suzi Gablik situates her "Connective Aesthetics" as a form of resistance to the Modernist "objectifying consciousness of the scientific world" (Gablik, 1995, p. 79) that led to the removal of art and the artist from living contexts, teaching artists to experience the self as something separate and autonomous from others and its environment. She argues that Modernism asserted an individualist worldview through "Its nonrelational, noninteractive, nonparticipatory orientation [that] did not easily accommodate the more feminine values of care and compassion, of seeing and responding to need" (p. 80). In this sense, SEA theory resonates with critiques of the main education system elaborated by Freire, Dewey, and Noddings. Indeed, regular mandatory schooling-referred to by Freire as banking education (1996)-also actively supports this paradigm of "objectifying consciousness" which divides, distances, and creates the same kind of "nonrelational" environment inhibiting gestures of care between students (Beckett, 2018;Stone, 2018).
To reverse this Modernist process of division, SEA artists act as "context providers" (Kester, 2004, p. 1) by creating environments and frameworks that can constitute open discursive spaces and welcome transformative exchanges. Nato Thompson (2015), while elaborating on "sites of transversal becoming" (p. 129), describes how these discursive spaces can lead to the arising of a sense of agency that can be transformative: Social environments where power is in flux, and not a given condition, can make room for open-ended possibilities of self-production. To put it plainly, when it is unclear who holds the power in a room, the room becomes a space where anything is up for grabs. When one feels free and empowered to act, one is able to change. (Thompson, 2015, pp. 136-137) These discursive spaces that set power in a "flux" by allowing the exchange to fluctuate, evolve, and fail could be described as spaces for "becoming pedagogy", as developed by Shin and Bae (2019), who-like Thompson-refer to Guattari's (1987) notion of "becoming" to elaborate ways to reverse the alienating processes of division generated by the influence of modernism. Gablik (1995), Lacy (2010), Kester (2011), Thompson (2015), and Shin and Bae (2019), agree that we collectively need to re-learn how to experience the self as relational-connected and integrated with its environment and community-through verbal ex-changes. Therefore, group conversations are major spaces of experimentation of new ways of relating with each other that are compassionate, caring, and politically revolutionary. These authors emphasize how these power relationships need to be constantly re-negotiated between the interlocutors themselves for the conversations to acquire its generative potential.
In this sense, SEA artists working collaboratively need to resign some of their power over the discursive spaces they initiated for the exchange they welcome to prompt engagement. Pablo Helguera (2011) and Suzanne Lacy (1995) both assert this relationship between power over the project and engagement in the project. Rather than advocating for a complete mitigating of power imbalances, which they recognize as unattainable, both suggest that the level of power offered to participant over the project must be proportional to the level of participation that is expected from them. Helguera frames this relationship within the context of discursive spaces by stating that the power given to participants should take the form of "freedom given to the group to shape the exchange" and that this "openness of the format and content of the project must be directly proportional to the level of genuine interests that the artist shows toward the experiences of community and his or her desire to learn from these experiences" (Helguera, 2011, p.48). Gablik (1995), Kester (2004), and Lacy (2010) points in the same direction by advocating for artists to adopt approaches based on empathic listening rather than facilitating.
However, in addition to the impact of the artist's attitude and sincere curiosity, the literature recognizes the necessity for a discursive space-for it to be transformative-to manage existing power dynamics related to disparities in access to knowledge and discourse production. These reflections on cultural capital and its effects on power dynamics within group conversations are highly relevant to the relationship between me, as an adult artist, and the students. Suzi Gablik asserts how these disparities can be partly alleviated through a transformation of the notion of knowledge. She advocates for new understanding of knowledge as a concept that includes and recognizes the value of personal experiences, referring to the legacy of feminists movements (Gablik, 1995, pp. 81-82). Building on this notion while also referring to and problematizing the notion of "public sphere" developed by Jürgen Habermas, Grant H. Kester (2004) asserts the development of a "local consensual knowledge" (p. 112) as a key aspect of his "Dialogical Aesthetics" (p. 82): A dialogical aesthetic, for its part, does not claim to provide, or require, this kind of universal or objective foundation. Rather, it is based on the generation of a local consensual knowledge that is only provisionally binding and that is grounded instead at the level of collective interaction. (Kester, 2004, p. 112) According to Kester, this focus on accepting-within a discursive space -a knowledge that would be "only provisionally binding" (p. 112) rather than universal and everlasting allows for a successful rewriting of what is recognized as valid forms of knowledge. According to Kester, the group needs to take the time to understand the "speaker's history"-meaning the past experiences and the social context which have shaped the speaker's subjectivity-in order to successfully avoid the trap of holding speakers accountable to "some ideal or generalized standard" (Kester, 2004, p. 113) which would perpetuate forms of exclusion. Therefore, these transformative discursive spaces described by artists and critics of the SEA movement rely on their capacity to offer to the group the freedom to shape the structure as well as the content of the exchange, and to foster a more inclusive rewriting of what constitutes valid forms of knowledge to be shared within the group.
Project's framework I approach high schools and secondary schools by offering to set up an ephemeral tearoom where students can have conversations about schooling and education. As previously stated, in every school welcoming the project, the framework I suggest is discussed and adapted collaboratively with the teachers and students who take part in the project. Therefore, there are slight variations in the way the project happened in various school. However, the main aspects of the framework were common to all schools: the space was transformed by introducing new furniture, participation was voluntary, and the conversations were inspired by audio excerpts of other students' exchanges.

The space was transformed by introducing new furniture
Before each session, the students and me transform the space by pushing all desks and chairs to the side, and by installing the carpets and chairs I brought in the middle. In a regular classroom, furniture and the spatial organization of the space prescribe specific movements and assert where students have to direct their attention: towards the front, rather than towards each other. Therefore, I wanted to invite students in a space that would enable new possibilities in terms of relationships. Hence, I built foldable Russian birch plywood chairs that would be installed in a circle, directing students' attention towards each other. The use of these chairs was often combined with the use of traditional chairs or other adaptive furniture to accommodate all students, their needs, and their level of comfort. Tea is served during each session.

Participation was voluntary
When I suggest the project to a school, I mention that the initiative is an art project and I send pictures of my chairs. Once I establish a relationship with one or many teachers, I ask for a few minutes to present the project to the students during class: I introduce myself and the project and invite students to come and converse about the education system once a week for the next 4 to 8 weeks, stating that the activity is not integrated in any curriculum. I give them the time and place of the first session and invite them to show up if they wish to take part in the conversations, stating that there is no obligation to attend all sessions. I have also done this by email in some instances.
On the first session, I answer students' questions about the project and I re-establish the fact that I am not a school's employee and that the school did not initiate the project but rather accepted to participate in it. This helps establishing the fact that the course of the activity is not prescribed by the school like regular teaching is. I reinstate that they are not evaluated, that their participation is not monitored, and that they have no obligation to show up every week. I also ask them to not mention any name, if they happen to share experiences involving other people or other schools.

The conversations were inspired by audio excerpts of other students' exchanges
At the beginning of each session, rather than bringing a question or suggestion of subject myself, I give student's access to the collection of excerpts the project have generated in other schools to allow conversations to be inspired by other student's ideas and questions, rather than adults' ideas or questions. The students choose what excerpt they wish to listen to according to the subject they feel like exploring that day. These audio excerpts of conversation serve as catalysts as the foreign words from other students become a starting point for the students' exchange. Week after week, students build, together, an archive of interlacing experiences and ideas 1 .
Recording exchanges between students who are teenagers is a gesture with many ethical implications and managing these ethical implications is an important part of the framework that I work to constantly improve. On our first session, we discuss the possibility of recording our exchange. I explain that I offer the possibility to record because it can help create bridges between different schools. However, I insist of the fact that recording is not mandatory. The conversation is recorded only if all students agree for it to be. So far, 53 out of 56 conversations were recorded.
During this discussion, I give to all students a form for their parent to sign, and we read it together. This form asks parent's permission to use the recordings and explains in detail the project and the ways in which the recordings will be used and shared in the short-term (played in other schools, on the project's website, and in exhibitions) and how they might be used in the long term (research articles and publications). This form ensures that the students and parents have the right to pull their or their children's recordings from the project at any time in the future. They are given my personal contact information as well as resources to follow the project's evolution.
Finally, we discuss and decide how the selection of experts will be conducted. In all instances so far, students wished for me to select the excerpts, as they were not interested in doing it themselves. To still give them power over their representation in the project, I give them back the excerpts I selected for them to listen to and tell me if they wish for some to be taken out or modified. Even if they agreed for the conversations to be recorded, all students can choose to not be included in the excerpts. In this case, I do not select any excerpts where they are heard.
Students can also change their mind since I ask for every group to reinstate their consent to be recorded before every session.

Schools visited and their context
I spend 4 to 10 weeks in each school, meeting with the same group of 3 to 10 students weekly or bi-weekly, aiming for each group to meet at least 4 times. To this date, noting that the project is still ongoing, 83 students from 11 different schools took part in the project from various regions of Canada (Montreal QC, Carleton-sur-Mer QC, and Lévis QC) and of Switzerland (Ebnat-Kappel SG, Will SG, and Wattwil SG). Of these 11 schools, 2 were privately funded, 2 were in rural areas, 3 were in underserved areas, and 1 used alternative pedagogy. In total, 56 conversations took place of an average of 1.5 hours each. Therefore, the project led to approximately 84 h of exchange. Of these 56 conversations, 22 welcomed some students with diagnosed disabilities and special needs. Of the 83 participating students, 5 had repeated at least one grade. Various responses to framework I identified three possible responses to the framework elaborated previously: students do not feel the need to invest the space and stop coming after a couple of sessions, the students do not feel the need to formulate critiques of their school of the education system but still invest the space by broadening the subjects explored, and finally, the students invest the space to share personal experiences of instances of lack of care and formulate critiques and solutions.
Students do not feel the long-term need to invest the space I observed the first response on one occasion in the school I visited that was using alternative pedagogical approach (school 1 of table 01). The students came for the two first sessions to talk for this is that they already had access to open discursive spaces elsewhere in the school as the teaching was participative and based on self-led projects and direct democracy and so, their interest in the project was instead based on the possibility to advocate publicly for alternative pedagogy via the recordings. Once they considered that it was done, they did not feel the need to invest the space of the project anymore.
Students do not formulate critiques but still invest the space by broadening the subjects explored I observed the second response in three schools I visited (schools 3, 6, and 10 of table 01).
While discussing their relationship with schooling and the education system in the first and second session, participating students agreed that they felt comfortable and free within their Figure 2. Plywood chairs and carpets folded on a cart and ready to be transported to the next school. Credits: Anouk Verviers school. However, they kept coming and investing the space but broadened the subjects explored.
They discussed labour practices, economic inequalities, immigration, challenges of queerness in rural area, social medias effects, pressure to be unique, etc. Both of these responses derive from the capacity from these 4 schools to make students feel heard and taken into account in the institutions' decisional processes.
Students invest the space to share accounts of instances of lack of care and formulate critiques I witnessed the last response in the 7 remaining schools (schools 2, 4, 5, 7, 8, 9, and 11 of table 01) were students invested the space of the project to exchange about difficult experiences and formulate critiques of their school and the education system. I observed how these groups would start by sharing personal experiences and would broaden their analysis to speculate on other perspectives. Then, they would collectively try to elaborate solutions.
While an analysis of the level of harm conducted by these schools would be relevant and important, this project was not intended in this sense and the participating students were not gathered as a representative sample from the school's student population. Moreover, studies assessing harm in schools and in the education system use a method based on interviews (Lewis and Pearce, 2020;Dadvand and Cuervo, 2020;Archer and Yamashita, 2003), which greatly differs from the recordings collected during this project. Accordingly, I will instead focus my analysis on exploring how this possible openness of the discursive space could have allowed students to respond to imbalances in the conversation with some gestures of care and how the sessions could have been empowering for some of these students.
Students shared instances of lack of care. In every one of these schools, at least one student shared their experience of instances where the school as an institution failed to care for their mental or physical wellbeing, or that of a peer, to some degree. I chose to name and categorize the experiences shared by students in this section to exemplify how some subjects that were taboo could be discussed. Students from these schools notably shared experiences of bullying by other students, bullying by teachers or school employees, mental illness induced by school-related stress, homophobia, anxiety related to grades and their effect on career choices, lack of motivation, lack of time for developing deep friendships, negative repercussion on their self-esteem and their perception of their own abilities and potential, teacher and professional refusing to follow recommendations associated with their diagnosis, and migraines induced by noisy and tense classrooms environment.
Students broadened their analysis. All these groups of students spontaneously, from one week to another, started to extrapolate from the sum of their experiences to consider the broader context of the education system. They were accounting from various perspectives: successful students, struggling students, students with personal challenges, teachers, school employees, policymakers, etc. The students notably discussed: the validity and accuracy of evaluation, the impact and necessity of the grading process; the prevalent valorization of science and mathematics over humanities and arts, the dichotomy between objective or subjective modes of knowledge production; the dominant definition of intelligence excluding emotional intelligence; the lack of professional independence faced by teachers; the role played by schools in the reproduction of social, racial, and financial inequalities; the school's bias towards teaching and rewarding conformity; the impact of the schooling process on student's self-esteem and self-worth; etc.
Students responded to imbalances in the conversation's structure by experimenting with gestures of care. The project's framework preventing me from facilitating exchanges and preventing a conversation regulating structure to be imposed meant that conversations between students could go "wrong" and end up generating very imbalanced structures of exchange.
Each student could potentially monopolize speaking time, retreat from the exchange by refusing to participate, interrupt other students while they were speaking, etc. In fact, many of the first conversations happening between the groups would include one or many of these patterns, which I would intentionally restrain myself from compensating. As a result, students would often develop their own response to try and mitigate the imbalance. The following qualitative accounts exemplify some instances where that happened: * On our first meeting, this one student monopolized speaking time. Some students tried to speak with more conviction or to keep talking while he was trying to interrupt, but these attempts were unsuccessful. The following session, one of the students took full responsibility to show the over-sharing student that he was heard.
He sat next to him, would make eye contact, and would signify him that he was listening with some non-verbal gestures. The student was still speaking more than others, but he was slowing down, getting less anxious, and starting to listen. By the fifth meeting, the speaking time was almost equally shared. The group seemed to have gained his trust and had received and validated the difficult experiences he had shared.
* This group's first conversation was very balanced in terms of speaking time, but one of the students stood out to the group because she was both eloquent and emotive. In the second session, one other student started to systematically confront her.
He would play devil's advocate on almost all her interventions. The tension grew and their debates were monopolizing the conversation. On the third session, other students took to defending the first eloquent student, which seemed to provoke the confronting student. Finally, on the fourth session, one other student made a joke and enunciated the tension that he perceived. Other students laughed and signified that they had noticed too. They were not laughing at the confronting student, but with him. Yet, it still made him realise how his attitude toward her was overbearing to the rest of the group and, by the sixth meeting, he had stopped systematically confronting her.
* She did not speak for the first three weeks that the group met even though she was there every week and was clearly listening the whole time. The group seemed to notice that she did not feel the need to share anything but that she was still involved.
The other students remained unbothered and did not point out at her silence. They seemed to respect her listener's posture. On the fourth week, she shared a difficult experience. It was evident that she trusted the group and the students listened to her with great care. The following two weeks, she was speaking more, while remaining in a listener-based posture.
* These accounts exemplify how on many occasions students seemed to become aware of the developing imbalances and patterns harming the generative potential of the collective conversation and seemed to experiment with means to mitigate them. This analysis of these accounts is also correlated by the feedback received by students at the end of the project. Even though I did not constrain students to give feedback on the sessions and the conversations, I received feedback in an informal way. In 7 of the 11 schools (schools 2, 4, 5, 6, 8, 9, 10 in table 01), students formulated explicitly that they felt empowered by the experience. They identified the freedom given to them to explore the subjects they wanted and to develop their own structure of exchange as determining factors.
Therefore, in the upcoming sections, I will analyze what aspects of the project's framework described above might have made it successful in these instances to allow for certain gestures of care to arise and feel empowering.
Aiming to redistribute power and responsibility to foster the openness of a discursive space Trying to mitigate the potentially restrictive effect of the school's requirements The recordings show that many students participating in the conversations seemed to feel protected enough to share difficult experiences and formulate some critiques of the school or the education system, as mentioned previously. I would suggest that this was potentially supported by the fact that, by taking part in the project and the conversations, the students were conscious that they were stepping in a new type of art space that they never had experienced before, however this risk was tempered because they were still within an environment and a space that was theirs. This dislocation potentially made it easier to create a space that students were Figure 3. Students having a conversation about the education system while sharing some green tea. Credits: Anouk Verviers more likely to perceive as 'else' from the school-even though it existed spatially within the institution. The school was not initiator of the project but was welcoming the project, which that had been elaborated as a SEA project outside of the school's interests and requirements. There was no monitoring of the conversations' content or the students' attendance by the school as no teachers were mandated to be present during the exchange and as I did not share recordings with anyone other than the students. All recordings that were released publicly on the website or in exhibitions were made anonymous as they were never associated with any name or location.
The most obvious reason for sticking with these conditions was to protect students from being identified by the schools, the teachers, or the employees. However, these conditions were also necessary to establish the space as separate from the school apparatus, therefore unlocking new possibilities of actions and attitudes for students: acknowledging instances of lack of care, questioning the education system, formulating critiques, etc.
This distinction between the project and the usual use of the classroom was also asserted through the process of pushing the desks to the side to sit in a circle. This transformation of the space was collaborative as students would help me push the desks and chairs to the side, spread the carpets, assemble the plywood chairs, and prepare the tea. This process gave them the permission to break the usual order of their classroom by noisily relegating the desks and chairs to the back of the room. Through this simple gesture, students were taking responsibility in the creation of a new discursive space that would open new possibilities of interaction.
Experiencing, even briefly, this feeling of agency over the configuration and the use of the space contributed, moments to later, to ignite a sense of creative agency once the students sat to exchange.
However, the shielding of conversations from the school's monitoring and the transformation of the space were not in themselves sufficient for the discursive space to be experienced as open by students and to develop some sense of agency over the project. Knowing that there is no way to reach a completely balanced power distribution between me, an adult artist, and the students, I believe that I alleviated this imbalance at least partly by fostering the development of a "local consensual knowledge" (Kester, 2004) and by giving students the "freedom to shape the exchange" (Helguera, 2011) in accordance with my analysis of the literature on SEA projects outlined above.

Developing local consensual knowledge to foster a sense of agency
Fostering the "generation of local consensual knowledge" described by Kester (2004) as a way to provisionally develop an inclusive concept of knowledge that would include personal experiences within groups of students implied that I first needed to restrain myself from bringing any external sources of hegemonic knowledge. These sources would be perceived by students as representatives of an omnipotent adult-built discourse necessarily overshadowing their own.
I would also need to restrain myself from suggesting subjects or asking questions because I was not a member of the community, and my choice of relevant subjects could only be based on a superficial understanding of their context. This choice derived from trial and error. In the first schools I visited, I would bring suggestions of subjects on pieces of paper or video excerpts of tv reporting or of interviews with thinkers. I witnessed how this practice would lead to students feeling either uninspired, intimidated to react or contradict the sources, or constrained to formulate the right answer.
Instead, I started bringing a bank of audio excerpts of conversations from other schools sorted by subjects and concerns, which would allow each group to choose excerpts referring to what they felt like exploring. This use of excerpts would give students the power to determine their subject all the while being presented with new perspectives. Many of these excerpts would include mention of personal experiences, which would legitimize these types of interventions, allowing students to dive into their respective "speaker's history" (Kester, 2004). These pieces of exchange had a slightly relaxed and unrestrained quality to them because they were pulled from real conversations and were not formulated to be recorded. This relaxed quality would be perceptible in the form: students would interrupt each other, speak on top of each other, laugh, etc; as well as in the content: students would share memories, ask questions, contradict each other, etc. In this sense, these recordings that the framework would present as valuable contributed to deconstruct the "ideal and generalized standard", as mentioned by Kester, that students are used to encounter and perceive in the sources and pieces of information presented to them as valuable in general (Kester, 2004).
This contributed to reframe what would constitute a valuable contribution to the exchange-and extensively a valuable contribution to knowledge production-in a similar way to the one described by Gablik (1995) and Kester (2004). This would in turn have the potential to give students a sense of agency: they did not need to "learn" from an external source beforehand in order to be able to formulate valuable interventions, they were already in a position to contribute.
Keeping the conversation's structure open to allow for gestures of care to arise This project's aim was to give students the space and time to potentially initiate the acknowledgement of their "relational identity" (Gablik, 1995) -to start perceiving themselves as active members of a community of students that can practice care, empathic listening, and validating of personal experiences. In this sense, I would suggest that one of the most important levers in this process was the fact that I refused to suggest or impose to students any form of conversational organization and refused to act as a facilitator.
As explained previously, the literature reviewed emphasized how a conversational space can lead interlocutors to start experiencing their self as relational in the condition that power relationships in this space are constantly re-negociated between the interlocutors themselves (Gablik, 1995;Lacy, 2010;Kester, 2011;Tompson, 2015;Shin & Bae, 2019). In this sense, suggesting a form of previously elaborated conversational organizing would disrupt this relentless movement of modes of expression, enunciation, and reception; thus, disempowering students. They would be constraint to surrender their agency towards the imposed mode of conversational organization. This process would mostly prevent gestures of care from prevailing as caring for the participating students would no longer be a responsibility shared by the group, but rather a responsibility handed over to the imposed structure.
In this sense, on our first meeting, I would take the time to express how, as an artist, I viewed exchange as a creative medium; emphasizing how the conversation in itself could be more important than the conclusion reached, if any. This was the only guidance I would provide students, in addition to refusing to facilitate exchanges and trying to adopt a position based on empathic listening, as described in the literature. Student would, for example, expect me to allocate speaking time, ask me to approve or disapprove of something said, talk to me directly instead of addressing the group, etc. Each time, I would enunciate how I did not wish to act as a facilitator but as a listener because I wanted to learn from them. Obviously, I was not impeccable, and I would identify moments where I would deviate from this ground rule while listening back to recordings. Regardless, on many instances, my efforts seemed to be sufficient to create a discursive space opened enough for self-governance to arise-12 groups out of 14 ended up leading their own exchanges. They were in control of the subjects explored and were collectively elaborating the structure of the exchange, in accordance with the approach advocated for by Helguera (2011) and Kester (2011).
I would suggest that this was made possible by the fact that participation was not mandatory nor monitored. Indeed, engaged students that would wish for the activity to go on would become aware that imbalances could prevent other students to willingly show up the following week.
Lasting imbalances in the conversation's structure could thus put the viability of the project at risk, but I was still refusing to take charge of compensating them as an adult. Consequently, the responsibility to make sure that all interlocutors were comfortable enough to come back the following week was spread amongst the students and myself, rather than being solely detained by a single person in authority. I witnessed on many occasions how engaged students would react to imbalances. They seem to show up the following week with the intention of having a slightly "better" conversation, having figured out, in the meantime, some ways through which they could potentially prevent these imbalances from resurfacing again. Knowing that these instances did not take place systematically in every rendition of the project in all 11 schools visited, I still think that these gestures of care reveal a certain potential that the proposed framework might have in terms of fostering a sense of agency that can prove to be transformative and generative on certain occasions.

Conclusion
The analysis of the framework of the project Au milieu des bureaux empilés (In Between Desks) through the lens of the literature on SEA made it possible to consider how to create-within high schools-discursive spaces that are open enough for students to self-govern their exchanges by controlling the subjects discussed as well as elaborating the structure of their conversation. This could be rendered possible by creating a space mostly protected from the school's monitoring of the content shared, by relying on the development of a local consensual knowledge valuing personal experiences, and by keeping the conversation's structure open for the students to shape it themselves.
To aim at generating such spaces of exchange, the paper elaborated one possible framework which relied on transforming the space welcoming the conversations with new furniture, maintaining a voluntary participation, and allowing conversations to be inspired by audio excerpts of other students who are also teenagers. Although imperfect, this investigated framework seemed to be successful enough to lead students to feel empowered, as this feedback was spontaneously formulated by students in 7 of the 11 schools visited.
The main aim of this paper is to hopefully open the conversation about how the gesture of opening discursive spaces where subjects and structures are not determined by adults may have the potential to lead some students to experiment with the responsibility to both fill this discursive space and take care of the people within it-and how this can constitute a positive experience.