Between Participation and Disturbance: Multilingual Repertoires of Students in Special Schools for SEBD

Abstract

Multilingual students’ repertoires are often framed either as resources for participation or as possible sources of difficulty, yet their situated use in special education remains underexamined. This study investigated how students in German special schools for social, emotional and behavioral development (SEBD) mobilized migration-related multilingual repertoires to address practical interactional problems and how situational conditions shaped the functions and consequences of these practices. A multi-sited ethnography was conducted in six SEBD special schools in the Rhine-Ruhr metropolitan region (North Rhine-Westphalia) between April 2024 and March 2025. Data comprised participant observation, written and talking fieldnotes, and informal ethnographic conversations. They were analyzed through a constant-comparative approach informed by reflexive grounded theory. Six recurring practice types were grouped into four result domains: Participation and Meaning-Making, Relational Positioning and Communicative Accessibility, Affective Mobilization and Counter-Positioning, and Navigation of Institutional Language Orders. Repertoire use had no fixed inclusive, disruptive, or resistant function. Its meaning emerged within interaction and changed with participation structures, public visibility, adult comprehension, and peer responses. In SEBD settings, migration-related linguistic marking and behavioral interpretation could reinforce one another. Multilingual practices should therefore be analyzed sequentially and situationally rather than treated as direct indicators of competence or disturbance.

1. Introduction

Multilingualism has become reality in contemporary schooling; however, students’ linguistic resources do not enter educational situations on equal terms. While some languages are formally established as subjects of instruction, others remain largely confined to familial contexts, peer interactions, or informal educational environments—particularly migration-related family languages. In this article, migration-related multilingualism refers to linguistic repertoires associated with individual or family histories of migration and with languages that occupy different positions within the school context (Binanzer & Jessen, 2020; Geist, 2021). The term does not imply a fixed correspondence between migration history, language proficiency, language use, or individual identity. Rather, it directs attention to how linguistic resources linked to migration-related biographies become available, marked, or restricted in educational interaction.

Whether students can use their resources for educational participation depends on the situations and relationships in which those resources become available, as well as on their institutional position (Gogolin, 2008; Huxel, 2020; Sturm, 2013). Research on multilingual repertoires and translanguaging has challenged conventional conceptions of multilingual speakers as operating through separate and self-contained language systems. From this perspective, speakers draw on integrated repertoires of linguistic, embodied, material, and semiotic resources to construct meaning and engage in social interactions (García et al., 2021; Li, 2018; Otheguy et al., 2015). Research has linked such repertoire-based practices to meaning-making, participation, metalinguistic reflection, multilingual identity, and experiences of belonging (Aleksić & García, 2026; Mendoza et al., 2024). This work has redirected attention from presumed deficiencies in the dominant language of schooling toward what multilingual students accomplish with the resources available to them. A repertoire-oriented perspective does not, however, imply that multilingual practices are inherently inclusive or empowering. Their educational significance depends on the institutional and relational context in which they occur (Jaspers, 2018). Students may not experience pedagogically initiated multilingualism as affirming, while access to linguistic resources continues to be structured by language hierarchies (Dirim & Mecheril, 2018), social inequalities (Piller, 2016) and school-specific expectations (Gogolin, 2008; Huxel, 2018). Formal and implicit language policies therefore need to be examined together with everyday school practices and the power relations through which they are enacted (Knappik et al., 2026; Thoma, 2023).

These institutional dynamics are particularly important in special education. Research has established associations between language skills and social, emotional and behavioral development (SEBD) (Chow & Hollo, 2022; Chow & Wehby, 2018; Hollo et al., 2014; Schramm et al., 2025). In this study, SEBD is treated as the institutional and pedagogical context of the observed interactions rather than as an explanatory attribute of individual students. Its analytical relevance lies in a setting in which behavior, affect regulation, participation, and responses to perceived disruption are subject to heightened educational attention. This makes it particularly important to examine how multilingual practices may become connected to behavioral interpretations without assuming that such practices are caused by students’ identified special educational needs. It also points to communication demands that may remain overlooked when students’ actions are interpreted as defiance, withdrawal or dysregulation. Yet much special education research approaches language through individual proficiency, delay, or impairment. Such approaches provide only limited insight into how multilingual students use their repertoires in naturally occurring peer and classroom interactions or how these practices become connected to interpretations of behavior (Niehaus, 2025a). Critical and decolonial perspectives on SEBD further question interpretations that detach distress or resistant behavior from the conditions under which they emerge. Leitner (2025) shows how uncertainty, exclusion, and experiences of racism may disappear from view when young people’s responses are interpreted primarily through individual dysfunction. Dis/ability Critical Race Studies similarly examines how racialization and dis/ability are co-constructed through educational knowledge and institutional practice (Annamma et al., 2013). These perspectives draw attention to the possibility that migration-related linguistic marking and behavioral interpretation may become intertwined and connected to processes of racialization (Akbaba & Buchner, 2019; Amirpur & Doğmuș, 2022). This does not presume that every conflict involving multilingual speech is racist or disabling—but it does require examining the conditions under which particular linguistic practices become noticeable and how they are interpreted. Conversely, research on multilingual education has rarely focused specifically on students with formally identified SEBD. Recent research has identified a marked underrepresentation of multilingual students in the evidence base on social and emotional learning interventions and has called for greater attention to the relationship between linguistic minoritization and social and emotional education (Lucas et al., 2026; Niehaus, 2026). More broadly, research at the intersection of multilingual education and special educational needs remains limited, while linguistic, emotional, psychosocial, and institutional factors interact in the identification and support of multilingual students (Roux Sparreskog, 2025). Existing studies have primarily examined diagnostic categorization, professional perceptions, language-support models, and intervention outcomes (Haas & Amirpur, 2025; Leone-Pizzighella, 2022; Noriega Etxaide, 2025; Plöger & Barakos, 2021). Students’ own situated practices and perspectives in SEBD settings remain less visible. This research gap limits understanding of multilingualism in SEBD settings, where linguistic practices unfold within affectively charged classroom demands and institutional processes of behavioral interpretation and regulation. A shift within a student’s repertoire may help to formulate a contribution, establish proximity to peers, protect a private conversation, express frustration, or maintain a position during conflict. Classifying multilingual practices in advance as supportive, disruptive, or resistant risks overlooking both the practical interactional problems students are addressing and the sequence through which the practice acquires meaning. The present study therefore examines students’ migration-related multilingual practices as situated social action and addresses the following research questions (RQs):

• How do students in SEBD special schools mobilize migration-related multilingual repertoires to address practical interactional problems?
• How do situational conditions shape the functions and consequences of these practices?

The analysis reconstructs six recurring practice types across social situations and organizes them into four result domains: (A) Participation and Meaning-Making, (B) Relational Positioning and Communicative Accessibility, (C) Affective Mobilization and Counter-Positioning, and (D) Navigation of Institutional Language Orders. It examines how students use their repertoires and how school interactions shape whether these practices become available as participation, remain limited to peer-oriented communication or are interpreted as disturbance. In doing so, the study contributes a student-centered perspective on migration-related multilingualism, special education, and social and emotional participation.

2. Research Design and Methods

2.1. Research Design and Setting

Building on the theoretical framework of Knappik et al. (2026) on linguistic ethnographies in education, the present study followed a multi-sited ethnographic design focused on naturally occurring school situations (Hammersley & Atkinson, 2019). It formed part of a larger research project on migration-related multilingualism in special education (Niehaus, 2025b). The study involved six German special schools serving students with formally identified special educational needs in SEBD (the German support category Emotionale und Soziale Entwicklung). These institutions are referred to as SEBD special schools. Some students had additional formally identified support needs in learning and/or language development (the German support categories Lernen and Sprache). The SEBD classification defined the institutional setting of the study; it was not used as an individual-level explanatory variable in the analysis. Fieldwork took place in the Rhine-Ruhr metropolitan region (North Rhine-Westphalia) between April 2024 and March 2025. For this article, the analysis focused on practices initiated or co-constructed by students. Teachers and other professionals were included insofar as their actions, expectations and linguistic resources shaped the situations under examination. The social situation–rather than the individual student–constituted the primary unit of observation and analysis.

The sites differed in student age, composition, school organization, pedagogical structure, and the linguistic repertoires represented in the observed groups. To protect participants and institutions, the schools are referred to as Schools 1–6 (

Table 1. Overview of Research Sites, Observation Periods, Observed Groups, and Grade Levels Represented.

Site	Observation Period	Observed Groups	Main Grades Represented
School 1	April–May 2024	Primary and lower-secondary classes	Grades 3–8
School 2	June 2024	Lower-secondary classes	Grades 5–8
School 3	August–September 2024	Lower- and upper-secondary classes; intensive support group (RTI Tier 3)	Grades 1–10
School 4	October–November 2024	Primary and lower-secondary classes	Grades 3–8
School 5	December 2024	Primary and lower-secondary classes	Grades 1–8
School 6	March 2025	Primary classes	Grade 1

). Across the six sites, fieldwork was organized into 21 observation weeks covering regular lessons, open starts, breaks, meals, project work, social learning, sports, excursions, and transitions between activities. The resulting corpus comprised approximately 650 h of observation documented in fieldnotes. This article draws on the same broader ethnographic corpus as a related manuscript currently under review. Whereas that study reconstructs how professionals respond when migration-related multilingualism becomes relevant to classroom participation and order, the present analysis focuses on students’ practices and the practical interactional problems they address. It applies a separate coding framework and develops an independent, student-centered typology. This overlap is disclosed here for transparency; the analyses and substantive claims developed in the present article do not depend on that manuscript.

2.2. Data Collection

Data were generated through participant observation and documented in contemporaneous written and talking fieldnotes (Cloos, 2010). Talking fieldnotes consisted of audio-recorded research commentaries produced shortly after observation periods. These recordings were subsequently transcribed and incorporated into the fieldnote corpus. The material documented interactional sequences, spatial arrangements, bodily orientation, prosody, objects and materials, participants’ responses, and, where possible, short fragments of speech. The observations were complemented by informal ethnographic conversations with students or staff that arose during fieldwork. Information on students’ linguistic repertoires was reconstructed from observations, conversations with participants, and contextual school information. These sources were not treated as comprehensive assessments of individual language proficiency. Accordingly, descriptions such as “Arabic-speaking,” “English-speaking,” or “Italian-speaking” refer to linguistic resources that students demonstrably mobilized or that participants identified during fieldwork. They do not describe complete or stable competence profiles. Classroom interactions were not audio- or video-recorded. The excerpts presented in the findings are therefore renderings from the fieldnotes rather than verbatim interactional transcripts. Quotation marks indicate speech fragments documented contemporaneously in written fieldnotes or captured in the immediately subsequent talking fieldnotes; they should not be interpreted as transcription-level reproductions. Speech that could only be approximately reconstructed is presented in reported or paraphrased form. Excerpts not originally in English were translated for publication. Original linguistic forms were retained where the language choice or wording is analytically relevant. The language of a speech fragment was identified only when participants confirmed it or when it could be established with sufficient contextual certainty.

2.3. Analytic Strategy

The collected data were analyzed using a reflexive, constant-comparative ethnographic approach informed by reflexive grounded theory methodology (Breuer et al., 2019; Charmaz, 2014). The analysis proceeded iteratively rather than as a linear sequence of data collection, coding, and categorization. It involved repeated movement between discrete situations, within-school patterns, cross-school contrasts, analytic memos, and the evolving conceptual framework. In the first phase, the material from each school was examined to identify instances in which students mobilized, negotiated, offered, restricted, or visibly refrained from using multilingual resources. Non-use was included when a linguistic resource had been explicitly invoked, offered, requested, or otherwise made relevant within the interactional sequence. Relevant situations included explicit language switching, translation, language comparison, peer mediation, and the use of digital translation. Embodied and material practices were also included when they formed part of students’ responses to a linguistically structured practical interactional problem. In the second phase, the material was openly coded within each school. Coding focused on practical interactional problems addressed in the sequence:

• What were students trying to accomplish?
• Which linguistic or other semiotic resources were mobilized?
• How did other participants respond?
• What changed over the course of the interaction?

Initial coding included peer-supported formulation, multilingual meaning-making, affective switching, selective privacy, relational alignment, linguistic boundary-making, sanction commentary, and the monitoring of adult language expectations. In the third phase, related situations were grouped into school-specific practice bundles. This within-site reconstruction preserved local interactional orders and reduced the risk of detaching individual speech fragments or actions from their sequential and institutional contexts. Comparison considered contrasts between formal and informal situations, public and less publicly accessible interactions, different age groups, and varying degrees of affective activation or institutional pressure. In the fourth phase, the school-specific reconstructions were compared across all six sites. Codes and provisional categories were merged, differentiated, reformulated, or discarded through constant comparison. Negative and contrasting cases were used to test whether an emerging type referred to a recurring interactional mechanism rather than a pattern specific to one language, student, or school.

The analysis yielded six recurring practice types, organized into four result domains (

Table 2. Result Domains, Practice Types, Primary Practical Interactional Problems, and Site Distribution.

Result Domain	Practice Type	Primary Practical Interactional Problems	Sites Identified
(A) Participation and Meaning-Making	Repertoire-Mediated Participation	How can I understand, express myself, or participate in the situation?	Schools 1, 3, 4, 5, 6
(B) Relational Positioning and Communicative Accessibility	Relational Positioning	How can I establish proximity, belonging, or mutual acknowledgment?	Schools 1, 2, 3, 4, 6
	Managing Communicative Accessibility	Who should be able to understand and participate in the interaction?	Schools 1, 2, 3, 4, 6
(C) Affective Mobilization and Counter-Positioning	Affective Repertoire Mobilization	How can I express, share, or process affective activation?	Schools 1–6
	Linguistic Counter-Positioning	How can I maintain a position under conflict or sanctioning pressure?	Schools 1–6
(D) Institutional Language Orders	Navigating Institutional Language Orders	How do I respond to school language expectations and the unequal institutional status of linguistic resources?	Schools 1–6

Note: Site distribution indicates the schools in which each practice type was identified during the cross-case analysis. It documents the empirical basis of the typology but is not a frequency measure and should not be interpreted as indicating prevalence within or across schools. Observation periods and opportunities for observing particular practices differed between sites.

). The domains structure the findings, but do not constitute additional empirical types. Shared repertoires, public accessibility, affective activation, relational safety, institutional pressure, and task or material structure were reconstructed as cross-cutting situational conditions. Situations selected for detailed presentations were chosen for their analytical density and their capacity to demonstrate either a recurrent mechanism or an analytically revealing contrast. The selection was not intended to represent every school equally. Where a situation involved several practices, it was assigned according to the dominant practical interactional problem within the sequence; secondary functions were retained in the interpretation. The author carried out all coding and primary interpretation. Analytical reflexivity was supported through continuous memo writing, repeated comparison with the full corpus, and the inclusion of negative and ambiguous cases. In the analysis and presentation of findings, observed action, participant explanations, and theoretical interpretation were treated as distinct analytical layers. Throughout the approximately two-year period of data generation and analysis, selected fieldnote passages, provisional codes, and emerging interpretations were discussed in research workshops and peer data sessions at roughly two-month intervals. Participation varied across sessions and typically involved three to five colleagues from educational research, foreign-language education, research on multilingualism, critical inequality research, and sociology. These discussions accompanied the analysis across different stages rather than serving as a separate validation at its conclusion. Alternative readings, unresolved questions, and suggested follow-ups were documented in analytic memos and subsequently taken back into the researcher’s comparison with the wider corpus. The purpose of these discussions was not to establish inter-coder agreement or consensus. They served to expose the developing analysis to alternative interpretations and possible blind spots. In particular, the repeated discussion of contrasting readings contributed to sharpening the analytical focus from multilingualism as such toward the practical interactional problems students addressed through their repertoires.

2.4. Linguistic Access and Researcher Reflexivity

The researcher was not proficient in all the languages used in the observed school settings. The analysis therefore distinguished between directly accessible semantic content and interpretations derived from interactional sequences. The function of speech whose semantic content was inaccessible to the researcher was not inferred from prosody or volume alone. Interpretations were informed by preceding actions, bodily orientation, subsequent responses, participant translations, later conversations, and contrasts with comparable situations. Where several interpretations remained plausible, ambiguity was retained rather than retrospectively resolved. The researcher’s presence and linguistic positioning also formed part of the situations under study. In some cases, students translated for the researcher, explained the meaning of expressions, tested the researcher’s linguistic knowledge, or described school language expectations. These exchanges were treated as situated accounts produced within the relationship between researcher and participants, not as transparent access to pre-existing meanings. They could also shape which linguistic resources were displayed, translated, or made explicitly relevant during fieldwork.

At the beginning of each field stay, the researcher was introduced to the students as a university researcher and guest who would observe how teaching and everyday school life unfolded across different schools. The specific analytical focus on migration-related multilingualism was not foregrounded; the research interest was described more broadly as observing how lessons and everyday school routines unfolded in different settings. During classroom observation, the researcher usually remained in the rear part of the room and sought to maintain an unobtrusive position. The researcher did not assume instructional, disciplinary, or formal support responsibilities. When students initiated contact or asked questions directly, however, these interactions were not artificially avoided, and the researcher responded in a situationally appropriate manner. Such contact became part of everyday field relations, particularly during the longer observation periods. Familiarity with both students and professionals generally increased over time, while the researcher maintained a professionally bounded observer role. These field relations and researcher-participant encounters were treated as part of the ethnographic situation and considered in the reflexive interpretation of the material.

2.5. Ethical Considerations

The study received institutional approval from the Institute for Special Needs Education at the University of Duisburg-Essen. Access to the schools and school-based fieldwork was authorized by the responsible school supervisory authority and arranged in consultation with the leadership teams of the participating schools. The approved research design covered ethnographic observation at classroom and group level without separate individual recruitment. No student records were accessed, and the study did not collect diagnostic, anamnestic, or systematically compiled individual case data. Families were informed through the participating schools. An opt-out procedure was available. If parents or legal guardians informed the respective class teacher that they did not want their child to be included in the research, the student would have been excluded from observation and fieldnote documentation. No such request was received at any of the six schools. Students were informed at the beginning of the fieldwork about the researcher’s role and that everyday classroom interaction would be observed. Student dissent was treated as relevant throughout the fieldwork rather than only at the point of initial access. Verbal or behavioral indications that a student did not wish to be included would result in that student being excluded from observation and documentation. The extended periods spent with the same groups allowed the researcher to remain attentive to individual responses over time.

The fieldwork consisted of non-interventional ethnographic observation of everyday teaching and school life. Classroom interactions were observed while responsible educational professionals remained present. In affectively charged situations, the researcher sought to avoid increasing students’ exposure or interactional pressure. This included reducing visual attention, turning away where appropriate, and maintaining a physically unobtrusive position in the classroom rather than moving closer to the interaction. The researcher did not independently intervene in situations that remained under the direct supervision of responsible professionals. This approach was applied, for example, in the emotionally charged sequences reported in Section 3.2 and Section 3.4. The research did not introduce diagnostic tasks, standardized tests, formal interviews, or audio- or video-recording of classroom interactions. Informal ethnographic conversations with students and staff arose during fieldwork, as described in Section 2.2, and did not follow a structured interview format. The study did not systematically follow or assess individual students or staff members and did not construct individual research profiles. All schools, staff members, and students referred to in the fieldnotes and publications were pseudonymized. Potentially identifying details were omitted or altered where necessary to reduce the risk of deductive identification. The risk of deductive identification was considered not only in relation to external readers but also within the participating school communities. Particular care was taken where combinations of age group, linguistic repertoire, migration-related biographical information, and specific classroom events could make an episode recognizable to insiders. Exact dates and contextual or biographical details were therefore generalized, omitted, or slightly altered where they were not analytically necessary. Linguistic information was retained only where it was required for the analysis. These procedures reduce but cannot entirely eliminate the possibility that participants familiar with a particular event may recognize it.

3. Results

3.1. Overview of the Cross-Case Practice Typology

The cross-case analysis yielded six recurring practice types, organized into four result domains (Table 2). Similar repertoires and shared languages could serve different social functions across situations. A shift to another language might enable academic participation, establish proximity, limit access to an interaction, express affective activation, or support a counter-position under correction or sanctioning pressure. The typology was organized according to the recurrent practical interactional problems students addressed through their available repertoires. Here, a practice type refers to a recurring cross-case pattern in how students addressed a practical interactional problem, whereas situated action refers to the concrete enactment of such a practice within a particular interactional sequence. The concept does not imply conscious or strategically planned language use; many practices unfolded routinely in response to immediate interactional demands. The four domains structure the presentation, while the six practice types constitute the central empirical typology. The practice types describe recurring forms of situated action, not stable characteristics of individual students. Because a single sequence could involve several practices, its allocation followed the dominant practical interactional problem identified in the sequence. Secondary functions remained part of the interpretation (see Section 2.3). Table 2 also indicates the schools in which each practice type was identified. These site distributions document the cross-case basis of the typology rather than the frequency or prevalence of particular practices.

3.2. Participation and Meaning-Making

Students drew on their linguistic and other communicative resources when German alone did not provide sufficient access to a task or allow them to formulate an intended position. These practices included translation, lexical assistance, language comparison, gesture, modeling, objects, and digital tools. Participation was often achieved collaboratively through resources distributed across students and the situation, rather than through individual proficiency alone.

A detailed example occurred in School 1 during a social-learning activity. A school social education professional introduced a cooperative game in which wooden figures had to be placed on a movable balancing board without causing it to tip over. As the class jointly established the rules, one Arabic-, English-, and German-speaking student wanted to propose an additional time limit for the game. He paused, appeared frustrated, and turned to a peer: “How do I say it? M., please help me express myself.” The peer listened as the student explained the idea through English and Arabic and then reformulated the proposed rule into German for the class. The student ratified the paraphrase by nodding. Neither the class teacher nor the social education professional interrupted the sequence, although both appeared to follow the English explanation. The other students treated the switching between English, Arabic and German as an unremarkable part of the shared activity (Fieldnote rendering, School 1, lower-secondary class II, April 2024). The peer’s contribution extended beyond lexical assistance. The initial request to “help me express myself” indicated a gap between the student’s communicative intention and the German resources available to him at that moment. He selected a peer who could understand the code-mixing of Arabic, English, and German explanations and anticipate the linguistic requirements of the public group interaction. Although the peer supplied the German wording, authorship remained with the first student as he introduced the idea, selected the mediator, and ratified the formulation. The temporary language shift also moved the formulation work away from the public search for German wording at a moment when the fieldnote documented increasing frustration. Peer mediation kept him involved and enabled the proposal to enter the group discussion.

The episode was not exceptional within this class. During observations in this class, students regularly switched between German, English, and Arabic in informal exchanges and task-related cooperation. The two Arabic-, English-, and German-speaking students repeatedly used their repertoire, which was sometimes taken up by teachers. These practices were neither restricted nor treated as unusual (Fieldnote rendering, School 1, lower-secondary class II, April 2024). Language shifts generally proceeded without interruption and did not visibly attract special attention from students or teachers. Shared linguistic resources mattered, but so did an interactional climate in which their use did not itself attract attention or correction.

A contrasting constellation was observed in School 3 with a recently arrived student who used Ukrainian. No reliably shared linguistic repertoire between the student and those around her was apparent in the observed situations. Teachers frequently shifted from German to English, but neither language appeared sufficiently accessible to her. When the student was asked to place her communication folder in the designated location, she responded only after the teacher held up the folder and modeled the action. Before sports class, a question about sports clothing became understandable only after the teacher pointed to her own clothing and sports bag and added gestures and facial expressions (Fieldnote rendering, School 3, lower-secondary group, September 2024). Peers also attempted to create situational access. Before a mathematics lesson, several students responded to the teacher’s German instruction by saying “Matematika” to the student. One peer turned around, held up his mathematics exercise book, offered a worksheet, and attempted to formulate the offer in English. The student then took out a small mathematics book from her bag (Fieldnote rendering, School 3, lower-secondary class, September 2024). The interaction combined a cross-linguistically similar word with the visible object and the peers’ preparation for the lesson. The students did not translate the full instruction. Instead, they made the activity understandable through the coordination of speech, objects, and bodily orientation. The practical interactional problem differed from that in School 1. Here, participation began with basic orientation: Which subject was about to start? What material was needed? What action was expected? The limits of such peer-mediated bridging became apparent in a later sports class sequence. The student did not have appropriate sports shoes and was instructed to bring an English workbook to the sports hall instead. While the teacher was outside the classroom, a peer used his tablet to translate several sentences into Ukrainian and played the audio output to her: “Hello [name], how are you? What are your hobbies? If you want to take part in sports, you need sports shoes, like all of us.” According to the fieldnote, the student appeared to freeze when she heard the Ukrainian output. She turned away, burst into tears, and hid her face in her arms. When the teacher returned, the peer explained that he had used the translation application on his tablet to make the situation understandable to her (Fieldnote rendering, School 3, lower-secondary class, September 2024). The peer’s action was oriented toward linguistic access. He responded to the limited accessibility of German and English by introducing Ukrainian, which had otherwise remained largely absent from school interaction. The outcome, however, was ambiguous. The student’s reaction could have been related to the content of the message, the unexpected experience of hearing Ukrainian, the realization that she could not participate in sports, or meanings unavailable to the researcher. The episode nevertheless illustrates that introducing a presumed stronger language does not necessarily guarantee understanding or emotional relief. The peer initiative also pointed to a gap in the available institutional support. Students suggested or independently used translation technologies, while formal interaction continued to rely on German or English. Such mediation could bridge a particular moment, but it did not provide linguistic access across the school day.

Across the six sites, Repertoire-Mediated Participation appeared in more routinized forms. In School 4, a student reformulated an explanation of swimming movements in German and Arabic while demonstrating the movements to a peer in sports class. The teacher continued her German explanation without incorporating the bilingual mediation (Fieldnote rendering, School 4, primary class, October 2024). In School 5, students used another shared language to help a peer complete an account of St. Nicholas gifts and then supplied the missing information in German for the class teacher (Fieldnote rendering, School 5, primary class I, December 2024). In School 6, two students used English vocabulary while solving a German picture-word-matching task together (Fieldnote rendering, School 6, primary class, March 2025). Across these situations, students responded to local gaps in understanding or expression by combining available resources. At times, this produced a contribution for the wider group; elsewhere, it supported orientation, clarified a subject-related term, or enabled a peer to continue working. Much of this mediation remained spontaneous and only partly visible to staff.

Repertoire-Mediated Participation was repeatedly initiated and organized by students themselves. It ranged from collaboratively formulating a public contribution to making a task or situation accessible. No particular language or tool guaranteed participation; what mattered was how shared repertoires and peer support were combined within the sequence.

3.3. Relational Positioning and Communicative Accessibility

Students used their repertoires not only to access tasks or formulate contributions, but also to shape relationships and regulate access to interaction. This result domain includes two connected practice types. Relational positioning refers to practices through which students establish proximity, belonging, acknowledgment, or claims about who was entitled to use particular linguistic resources. Managing communicative accessibility captures how they opened interactions to others, limited access, or marked communication as private. The two practices overlapped when a shared language strengthened affiliation within a group while making participation more difficult for someone else.

This tension became particularly evident during the open start of a first-grade class in School 6. Students were free to select play activities and form small groups. Three students who shared English as a family language sat together in the play area and coordinated their activity through code-switching between English and German. When a fourth student joined them, the three initially shifted toward German. As the play continued, however, they repeatedly returned to English. The fourth student eventually interrupted them: “Don’t always speak English. I don’t understand. Speak German with me” (Fieldnote rendering, School 6, primary class, March 2025). The three students made renewed efforts to use German, but English continued to structure much of their coordination. During the approximately five-minute sequence, the fourth student gradually withdrew and eventually left the group (Fieldnote rendering, School 6, primary class, March 2025). The sequence does not support a straightforward interpretation of deliberate exclusion. The initial switch to German suggests that the three students responded to the newcomer’s need for access. Their return to English maintained a familiar and apparently more fluent mode of coordinating play and marked the proximity of the established group. For the fourth student, however, the effect was exclusionary. Although physically present, the student could not reliably follow the interaction through which the activity was organized. The request for German did not invoke a general monolingual norm. It concerned access to an activity in which the student was already involved. Fluent communication within the established group thus came into tension with another participant’s need to understand communication relevant to the shared activity. The group briefly accommodated this claim, but the shift to German was not sustained.

A second episode from the same class shows that limited communicative accessibility did not always call for wider disclosure. During breakfast, students coordinated activities and informal conversations by moving between languages their addressees could understand. A teaching assistant sat with them and showed interest without interrupting these shifts. Students frequently translated for her or supported one another when a German explanation became difficult. In one situation, however, a student declined to disclose the content of an exchange: “I don’t want to say it in German. It is private and only for J. I don’t want everyone to know.” (Fieldnote rendering, School 6, primary class, March 2025). The student appeared hesitant at first while maintaining the boundary of the exchange without directly rejecting the adult. The teaching assistant accepted the refusal and explained to another child that students did not have to share everything with everyone. The child replied: “But I can’t do that secretly because everyone understands German and I don’t speak other languages.” The teaching assistant responded that this did not justify preventing others from speaking an additional language. The other students listened without challenging her response (Fieldnote rendering, School 6, primary class, March 2025). Unlike the play sequence, this exchange did not concern a jointly organized activity. The student marked the communication as addressee-specific and private. A shared repertoire created a bounded interpersonal space within the classroom. The teaching assistant did not treat the refusal to translate as gossiping, insulting, or rule-breaking; she separated language choice from the right to keep some communication private. The German-speaking student’s objection nevertheless exposed an asymmetry. Students with an additional shared language could create communicative privacy more easily than someone whose available language was understood by almost everyone present. The sequence affirmed a right to privacy while also showing that students had unequal opportunities to create communication that others could not immediately access. Shared accessibility was necessary for the collective play activity, but not for every interpersonal exchange taking place in the same room.

Older students in School 3 articulated these distinctions more explicitly. In an intensive-support group, students described Arabic and other shared linguistic resources as ordinary parts of friendship and informal school life. While returning from a break, one student explained that “Kurdish-speaking friends spoke Kurdish, Turkish-speaking friends spoke Turkish” (Fieldnote rendering, School 3, upper-secondary class, August 2024), and that such shifts usually occurred without conscious planning. If someone wanted to know what had been said, they could simply ask: “You can always ask if you want to know. They will tell you if they won’t, but only if it’s very private. I don’t think anyone feels excluded. I and the others are happy to translate” (Fieldnote rendering, School 3, upper-secondary class, August 2024). The student distinguished this from brief exchanges in which another language could be useful precisely because not everyone nearby should immediately understand. While returning to the classroom, he spontaneously exchanged several Arabic words with a peer standing nearby. He then turned to the researcher, laughed, and commented that this was exactly what he meant. The spontaneous shift illustrated the routine character of such repertoire use in this interaction (Fieldnote rendering, School 3, upper-secondary class, August 2024). His account presented accessibility as relationally negotiated rather than absolute. Translation was offered where participation or reassurance required it, while some personal exchanges could remain bounded. The younger students negotiated this boundary through requests, temporary accommodation, or withdrawal; the older students formulated it as an explicit distinction between collective communication and interpersonal privacy. The same student also evaluated the conditions under which particular linguistic resources or expressions could be used appropriately. He described German-speaking peers’ use of isolated Arabic or Turkish expressions to present themselves as “cool” as “cringe” (Fieldnote rendering, School 3, upper-secondary class, August 2024). By contrast, he responded positively when teachers attempted to use Arabic, provided that their interest appeared sincere and they accepted correction. Appropriateness was assessed not through proficiency alone, but through intention, relationship, and respect for the language’s biographical significance. His evaluation differentiated routine peer use and respectful participation from the use of isolated expressions detached from the experiences of those for whom the language held personal meaning (Fieldnote rendering, School 3, upper-secondary class, August 2024).

Comparable practices occurred during breaks, in changing rooms during sports class, during partner work, and in other less publicly accessible settings. Students adapted their language choices to addressees and friendship relations, often returning to German when new participants joined or a contribution became relevant to a wider group (Fieldnote rendering, School 2, lower-secondary class, June 2024). Under closer adult monitoring, some older students preemptively stated that they were not insulting or discussing others (Fieldnote rendering, School 4, upper-secondary class, November 2024), indicating that limited adult access could attract suspicion. Conversely, a shared family language did not always determine affiliation (Fieldnote rendering, School 4, upper-secondary class, November 2024). Students sometimes rejected adults’ assumptions that linguistic similarity should determine working groups (Fieldnote rendering, School 1, lower-secondary class, April 2024), while peers without fully shared languages established proximity through youth-cultural expressions and codes circulating across repertoires (Fieldnote rendering, School 3, lower-secondary class, September 2024).

Relational positioning and communicative accessibility were connected but distinct. Shared repertoires could support fluent coordination, acknowledgment, and privacy. However, they could also leave others unable to follow a collectively relevant interaction. Across these situations, communicative inclusion did not depend on permanent transparency. Rather, access became relevant where shared understanding was necessary for participation, whereas interpersonal communication could remain bounded in other situations.

3.4. Affective Mobilization and Counter-Positioning

Multilingual practices also became visible under conditions of affective activation. Students mobilized different parts of their repertoires when expressing or responding to anger, frustration, sadness, joy, and relief. Two related practices emerged. Affective repertoire mobilization refers to the use of linguistic resources to express, share, intensify, or process affect. Linguistic counter-positioning describes how students maintained a position under public correction, peer confrontation, or institutional sanctioning. The practices could overlap within the same sequence, but they addressed different practical interactional problems.

Their overlap was particularly evident in School 1 in sequences involving a newly arrived student who frequently used Italian. During a plenary phase, the teacher instructed him to stop tapping on his desk. When he did not immediately comply, several peers joined the correction and addressed him in an increasingly confrontational manner. The student responded rapidly in Italian, directing his comments toward the peers. The semantic content was not directly accessible to the researcher, but a second teacher who understood some Italian and the subsequent peer responses treated the comments as part of the dispute. Students mimicked his intonation and one student shouted: “Stop using that fu**ing language” (Fieldnote rendering, School 1, lower-secondary class I, April 2024). The teachers did not address the imitation or the devaluation of Italian and mutual provocation continued. The conflict had shifted as the desk-tapping was no longer the object of conflict; Italian itself had become open to ridicule and public correction. The multilingual exchange and the peers’ responses contributed to the escalation and became part of how the student’s position within the peer group was challenged (Fieldnote rendering, School 1, lower-secondary class I, April 2024). Another sequence in math class showed that this pattern could not be explained through general inability or refusal to use German. During a whole-class mental-arithmetic activity, the student initially commented in Italian and received little attention from peers. After a teacher ensured his inclusion, he solved a multiplication task and gave the result in German. He then formulated a more difficult task in German for another student. While the classmate was calculating, the student spoke rapidly in Italian. The teacher asked him to be patient because the others were not as fast, thereby treating his speech as commentary on the calculation. When he later called out correct answers in German, another (multilingual) peer insulted him and demanded: “I don’t speak Española, you idi*t. Speak German like everyone else!” (Fieldnote rendering, School 1, lower-secondary class I, April 2024). A teacher moved closer to the Italian-speaking student and remained beside him. She did not correct the peer’s designation of Italian as Spanish, although the student was known in the class to frequently use Italian. During the later individual work phase, after public attention had diminished and a teacher sat next to him, he engaged more readily in the task (Fieldnote rendering, School 1, lower-secondary class I, April 2024).

The contrast within the lesson is analytically relevant. The student had demonstrated that he could use German for academic participation. Across these observed sequences, Italian became particularly visible when the student was publicly corrected, ignored, or confronted in front of the group. The shift allowed him to continue responding while reducing others’ immediate access to what he was saying. In this sequence, affective expression and counter-positioning were closely connected.

This interpretation does not frame the use of Italian as inherently resistant or empowering. The exchanges included mutual insults and contributed to escalation. At the same time, reducing the language shift to non-compliance or insufficient German would downplay its place within the unfolding conflict. A later classroom discussion during computer science class showed how the function of Italian changed with participation structure. After insults had been exchanged in several languages, the subject teacher led a class-wide conflict resolution. She asked whether a proposed German translation was accurate and invited the student to offer a conciliatory expression in Italian. He followed the discussion, confirmed translations by nodding and apologized in Italian. At the teacher’s request, he demonstrated the proper pronunciation and use of the expression. The fieldnote described him as calmer and physically more relaxed after the exchange (Fieldnote rendering, School 1, lower-secondary class I, April 2024). Italian was now incorporated into the repair rather than treated as an obstacle to it. The student did not have to abandon the repertoire before participating. His linguistic knowledge became relevant to resolving the conflict.

A contrasting form of affective repertoire mobilization was observed in School 6. During the multi-day absence of the regular class teacher, the first-grade group appeared increasingly unsettled. The substitute teacher addressed one student directly: “I know you miss Ms. X. We all do. It makes us sad, and that’s why we’re in a worse mood.” The student began to cry and found it difficult to calm down. A peer who shared Farsi approached him and began speaking in Farsi. A third student translated the exchange for the teacher without being asked (Fieldnote rendering, School 6, primary class, March 2025). Here, Farsi supported contact rather than opposition. One peer addressed the crying classmate through a shared language, while another kept the exchange accessible to the teacher. The students organized both affective support and mediation themselves. The primary practical interactional problem was not how to maintain a position against correction, but how to respond to emotional overwhelm without excluding the adult from the situation. The return of the class teacher the following morning generated immediate expressions of relief. Students sought physical proximity, hugged her, or reached for her hands. Their greetings drew on different languages and registers: “Massallah, Ms. X is back. I kiss your eyes, Ms. X, because you are finally back.” An English-speaking student exclaimed: “Oh my God, look who’s back. Ms. X is back. I am shocked. Finally, she’s back.” Other children shifted into Farsi. The fieldnote documented a visible reduction in bodily tension among several students. Several students who had appeared particularly unsettled during the teacher’s absence appeared more composed and cooperative during the open start (Fieldnote rendering, School 6, primary class, March 2025). The multilingual greetings accompanied and communicated relief across the group. They marked the teacher’s return as the restoration of a familiar relationship.

The contrast between the two schools clarifies the distinction between the practice types. In School 6, shared languages supported reassurance and collective relief. In School 1, Italian became a recurrent resource under public correction and peer devaluation. The difference lay in the interactional problem addressed (see Table 2): responding to affective distress in one case and maintaining agency under pressure in the other. Older students displayed more explicit forms of counter-positioning. Students in School 4 explained that they sometimes used insults in languages teachers were unlikely to understand, reducing the likelihood of immediate sanctioning. If questioned, they would downplay what had been said or challenge an inaccurate translation (Fieldnote rendering, School 4, upper-secondary class, November 2024). According to the students’ own accounts, limited adult access could be used strategically to delay or complicate accountability. A related mechanism appeared in School 5 under a strongly sanction-oriented teaching arrangement. Students who had been removed from the class circle and seated separately maintained contact across the room through shared additional languages. They commented on the trainee teacher and maintained contact around the behavior that the separation had been intended to interrupt. During more open phases of the same class, similar resources supported cooperation, play and academic assistance (Fieldnote rendering, School 5, primary class I, December 2024). Counter-positioning arose here from the combination of separation and monitoring as well as peer alignment. The repertoire itself did not determine the function.

Across the material, students used multilingual repertoires to express affect, seek support, and maintain a position under pressure. A language shift could calm an interaction or intensify a dispute. It could also make a contribution less accessible to adults. Its function became apparent only within the sequence, through what preceded it, whom it addressed, and how peers and professionals responded.

3.5. Institutional Language Orders

Across sites, students repeatedly adjusted their language choices to locally enacted expectations. Explicit rules were only one source of orientation. Students also attended to which languages adults understood, how they reacted to linguistically inaccessible exchanges, and when communication became relevant to the wider group. They used German to widen access, shifted other languages into less publicly accessible interactions, or challenged expectations they regarded as inconsistent. Students in the upper grades of School 4 articulated these processes during an informal conversation while walking to the schoolyard during the lunch break. They stated that languages other than German were not formally prohibited by a written school rule. Nevertheless, teachers repeatedly said: “German is spoken here” (Hier wird Deutsch gesprochen) (Fieldnote, School 4, upper-secondary class, November 2024). The students linked this expectation to previous conflicts. Insults in languages not understood by teachers had been difficult to attribute and sanction, while adults sometimes assumed that students were gossiping or insulting others when they could not access the content. They therefore understood the expectation to use German as serving both shared accessibility and institutional control. Their criticism focused on the different treatment of English and migration-related family languages. English was used by teachers and taught as a regular school subject, whereas Kurdish, Arabic, and other languages were more readily treated as problematic when adults could not understand them: “Some teachers distinguish between languages—others don’t. For instance, one teacher speaks Turkish and even uses it with students, yet forbids other languages. That’s really strange. Why do people do that?! Turkish isn’t German, after all—nor is it English. But it’s still more accepted than Kurdish or Arabic.” (Fieldnote, School 4, upper-secondary class, November 2024). One student explained that students sometimes turned the logic back on teachers: “When they speak English, we say: ‘German is spoken here.’ English is another language too.” (Fieldnote, School 4, upper-secondary class, November 2024). The students did not dispute the curricular status of English. They pointed instead to an inconsistency. English remained acceptable in public interaction, while other languages attracted suspicion when they moved beyond private exchanges. Students also distinguished between individual adults. Students described some social education professionals who used additional languages or shared youth-cultural expressions as more accessible and relationally close. Conversely, students criticized multilingual teachers who used their own family language with selected students while restricting others’ repertoire use. Institutional language orders were therefore experienced less as a uniform policy than as expectations enacted differently by individual professionals.

A contrasting orientation was found in School 2. One student believed that speaking another language at school was prohibited, although he could not identify a corresponding rule or teacher instruction. This caused little conflict for him because he was proficient in German and had no school peers with whom he routinely spoke Russian (Fieldnote rendering, School 2, lower-secondary class, June 2024). Other students described the linguistic division of school life more explicitly: “We use languages with each other whenever we want. But we do the learning in German. Otherwise, Mr. Y wouldn’t understand us” (Fieldnote rendering, School 2, lower-secondary class, June 2024). The statement was met with laughter from other students and presented the arrangement as familiar and largely uncontroversial. German was used for plenary interaction and communication with the teacher, while Arabic, Russian, Romani, and other resources remained available in peer exchanges and open activities (Fieldnote rendering, School 2, lower-secondary class, June 2024).

Both schools therefore operated with German as the expected public language of teaching. The difference lay in how students interpreted this arrangement. In School 2, they explained it mainly through communicative reach; in School 4, they connected it to control, unequal language status, and contradictory adult practices.

Students in School 5 similarly described German as the language shared by the largest number of students and professionals. Middle-grade students stated that they felt free to use other languages at school and had not been instructed to use German exclusively. During lessons, however, they “naturally” used German because it was the most widely comprehensible option (Fieldnote rendering, School 5, lower-secondary class, December 2024). A conflict at the same school showed that the consequences of language use also depended on whether adults could understand it. A student switched to English to criticize what she considered unfair treatment. Because the teacher understood English, the comment became immediately accessible and subject to sanction. The student argued that the same comment would have gone unnoticed in Romanian or Serbian. She thereby distinguished the curricular status of a language from the likelihood that its content would become accountable to adults (Fieldnote rendering, School 5, lower-secondary class, December 2024). Even first-grade students in School 6 oriented themselves toward person-specific expectations. In the presence of the regular class teacher, students moved between German, English, Farsi, and Arabic with little visible restriction. During breakfast time with a substitute teacher, an admonishing look prompted the English-speaking students to shift into German without an explicit rule being stated (Fieldnote rendering, School 6, primary class, March 2025). In a later emotionally charged interaction around bullying, the same teacher did not intervene in relation to multilingual comments. Her expectations were inferred from the immediate situation and problem being addressed rather than from a stable prohibition (Fieldnote rendering, School 6, primary class, March 2025).

Across the schools, no explicit German-only policy was required for an institutional language order to take effect. It was reproduced through routines, the curricular position of languages, adults’ repertoires, previous conflicts and responses such as interruptions or looks (see Gogolin, 2008; Wiese et al., 2020). Students learned when a language shift was likely to pass without comment, require explanation, or attract sanctions. Their practices went beyond compliance. Students used German when wider access mattered, retained other languages for peer communication, oriented to what adults could immediately understand, and criticized perceived inconsistencies. German could function as a shared resource, an implicit requirement or part of a hierarchy that assigned unequal status to linguistic resources. Which function prevailed depended on the school, the professional, and the situation.

Across the four result domains (Section 3.2, Section 3.3, Section 3.4 and Section 3.5), shared resources and material support were associated with wider opportunities for participation. Public and institutional pressure appeared in sequences involving alignment with German, withdrawal, or counter-positioning. Affective activation made repertoire shifts particularly visible but did not determine whether an interaction stabilized or escalated.

4. Discussion

The study examined how students in special schools for SEBD mobilized migration-related multilingual repertoires to address practical interactional problems (RQ1) and how situational conditions shaped the functions and consequences of these practices (RQ2; see Section 1). The six practice types capture recurring forms of situated action rather than stable characteristics of multilingual students. Taken together, they show that repertoire use acquired its function within interaction.

4.1. Multilingual Practices as Responses to Practical Interactional Problems

The findings are consistent with repertoire-based approaches that conceptualize multilingual meaning-making as flexible and multimodal rather than confined to separate named languages (García et al., 2021). Students combined their linguistic repertoire with gesture, bodily modeling, objects, and digital tools to gain access to tasks, formulate contributions, and support peers. Yet the presence of several languages did not in itself produce participation or empowerment (Jaspers, 2018). The findings are consistent with research showing that multilingual repertoires can support participation, meaning-making, and belonging (Aleksić & García, 2026; Mendoza et al., 2024). They add, however, that these functions could change across situations within the same school context as participation structures and responses from others changed. The central analytical question was therefore not which language students used, but what they were trying to accomplish in the situation. Peer mediation enabled public participation in School 1, multimodal support provided basic orientation in School 3, and shared English supported fluent peer coordination while limiting another student’s access in School 6. The function of repertoire use emerged from the interactional problem being addressed, the organization of participation, and the responses of others. The typology extends existing repertoire approaches by foregrounding the situational emergence of these functions. Multilingual practices did not carry a stable social meaning independently of the interaction in which they occurred. Their effects changed with the interactional arrangement. The same resources could support participation in one sequence and become involved in exclusion or conflict in another. This variation was not socially neutral. Repertoires entered school interactions under unequal conditions of institutional recognition and intelligibility. A practice that passed without comment in peer interaction could be interpreted as behaviorally problematic in a publicly evaluated situation. The typology therefore captures both what students accomplished through their repertoires and how school interactions shaped the meanings attributed to those accomplishments. These two dimensions become clearer when repertoire use is examined in relation to institutional frontstage and peer-oriented backstage interactions.

4.2. Frontstage and Backstage: Unequal Linguistic Legitimacy

Across the cases, multilingual practices were unevenly distributed between institutional frontstage and peer-oriented backstage situations (Goffman, 1959). Frontstage refers here to interaction that was publicly visible or oriented toward teachers. Backstage situations offered more scope for peer communication. These were not fixed spatial zones, nor did they simply reflect students’ language preferences. They were produced through the organization of participation and through expectations concerning who should understand what was being said. German was usually expected when communication concerned the wider group or required professional oversight. English could also enter public interaction relatively easily because it was taught as a school subject and used by some professionals. Migration-related family languages depended more heavily on shared peer repertoires. They attracted attention when adults could not follow an exchange or when limited access became associated with previous conflicts. Their institutional position was therefore shaped by more than communicative reach. Curricular status and social meanings attached to particular languages also mattered. This finding extends previous work on language hierarchies and school-specific expectations (Binanzer & Jessen, 2020; Thoma, 2023) by showing how students themselves anticipated and navigated these distinctions in everyday interaction.

These distinctions were not produced by students alone. They emerged within everyday interactions in which professional responses helped determine which repertoires could enter institutional frontstage communication without explanation or scrutiny. These boundaries were not neutral. From migration-pedagogical perspectives and critiques of racism and linguicism, these unequal conditions cannot be reduced to proficiency or communicative reach (Dirim & Mecheril, 2018; Piller, 2016). Linguistic resources were assigned different institutional positions and spaces (Buchner & Akbaba, 2023). Some could enter public interaction without explanation while others became conspicuous when professionals did not understand them. Contributions could therefore be assessed not only by what they accomplished but also by the repertoire through which they were expressed. The distinction between participation and problematic behavior could thereby become linguistically structured. Backstage communication should not, however, be treated as inherently resistant or romanticized as emancipatory. It allowed students to maintain privacy, support peers, and communicate with greater fluency. Limited accessibility could also support exclusion, insults, or the avoidance of immediate sanctions. Conversely, frontstage interaction was equally ambivalent. It became restrictive when public participation required students to suppress available resources, but it could also widen participation when teachers incorporated those resources into the shared interaction. The relevant inequality therefore did not consist in a simple opposition between German and other languages. It concerned the terms on which certain repertoires could become publicly usable without attracting suspicion, correction, or demands for justification.

4.3. Affect, Agency, and Doing Dis/Ability in SEBD Settings

The distinction between Affective Repertoire Mobilization and Linguistic Counter-Positioning (Section 3.4) is particularly relevant to SEBD settings. Previous research has identified associations between language and communication difficulties and social, emotional, or behavioral difficulties (Chow & Hollo, 2022; Chow & Wehby, 2018; Hollo et al., 2014). Such associations, however, do not explain how behavior acquires meaning within an interaction. The present findings direct attention to the linguistic and relational circumstances under which students’ responses become visible and may be interpreted as behavioral disturbance. In School 1, Italian allowed the student to maintain a position under public correction and peer confrontation. Its interactional consequences changed when the participation structure changed. During the later conflict resolution, Italian was incorporated into the exchange and supported the student’s contribution to repair. In School 6, Farsi enabled peer reassurance while spontaneous translation kept the interaction accessible to the teacher. These cases show that repertoire mobilization cannot be assigned a behavioral function independently of the sequence in which it occurred.

This situational perspective resonates with Leitner’s (2025) account of how young people’s distress and resistance may be pathologized or psychiatrized when the conditions under which they emerge recede from view. Drawing on research concerning young refugees and racialized students, Leitner shows how uncertainty, exclusion and racism may become obscured when responses to them are interpreted primarily as individual dysfunction (also see Kollender, 2020; Teigler & Flory, 2019; Tißberger, 2022; Velho, 2015). Institutional conditions may contribute to escalation, while the resulting response is attributed to the individual as evidence of disorder (Willmann, 2012).

The present study does not establish that the observed practices resulted from prior experiences of racism. Nor does it reconstruct the formal processes through which students’ SEBD classifications were produced. It does, however, reveal an interactional process in which linguistic devaluation and behavioral interpretation became connected. In School 1, a correction of desk-tapping developed into the public ridicule of Italian. The student’s subsequent counter-positioning could then appear primarily as further non-compliance. Within the sequence, attention shifted from the devaluation of the student’s multilingual repertoire and his public exposure toward the behavior that followed. The sequence also contained the potential for public shaming. Research in SEBD and trauma-sensitive education has emphasized the significance of humiliation, exposure and peer audience in pedagogical interactions (e.g., Hafeneger, 2013; Herz & Heuer, 2014; Neckel, 1991; Zimmermann, 2012). The material does not include the student’s own account and therefore does not establish that he experienced shame. The following interpretation therefore concerns the interactional sequence rather than the student’s subjective emotional experience. The interaction can nevertheless be examined as a possible process of second-order shaming. The term is used here as a case-sensitive analytical heuristic rather than as a generalized theoretical construct. It draws on existing work on shame, humiliation, and public exposure in educational interactions (Hafeneger, 2013; Herz & Heuer, 2014; Neckel, 1991; Zimmermann, 2012) and describes a specific sequential pattern in which a response to public devaluation itself becomes the object of renewed correction and exposure. First, the student’s desk-tapping and subsequent use of Italian were publicly challenged. His affective and counter-positional response could then become available for interpretation as another potential sign of disturbance. A response that had developed within the interaction could be read as an individual behavioral problem. This matters in SEBD settings, where defensive, aggressive or oppositional responses may attract immediate institutional attention while their relational origins remain less apparent. Because this pattern is reconstructed from one analytically dense sequence, its relevance beyond the present case requires examination in further research.

Dis/ability Critical Race Studies (DisCrit) provides a framework for interpreting this process. DisCrit conceptualizes racialization and dis/ability as historically and institutionally interdependent rather than as separate categories that occasionally intersect (Annamma et al., 2013). DisCrit emerged from the specific historical and institutional context of racialization and disability classification in the United States and is not transferred unchanged to the German context here. Rather, it is used as a sensitizing framework for examining how migration-related linguistic marking may become connected to dis/ability-related interpretations of behavior. In the present analysis, racialization refers to processes through which language use becomes linked to migration-related attributions, assumptions about belonging, and unequal social positioning within schooling (Akbaba & Buchner, 2019; Dirim & Mecheril, 2018). This does not equate German migration-related categories with the racial formations from which DisCrit originally emerged. From this perspective, migration-related linguistic marking and dis/ability-related interpretations of academic performance or behavior can be examined as potentially interconnected processes. The use of Italian was not marked because of any inherent property of the language. It became marked through its relation to the expected language of public classroom interaction, adults’ uneven linguistic access, and the meanings attached to migration-related multilingualism. Some peers who themselves used multilingual repertoires nevertheless framed the use of Italian as irritating and incompatible with accepted classroom participation. Once the repertoire had become conspicuous in this way, the student’s continued use of it could become readable as further disruption, dysregulation or communicative refusal. Linguistic marking and behavioral interpretation thereby reinforced one another. At the level of theoretical interpretation, this convergence can be understood as a situated doing of dis/ability. Disturbance was not produced by the student’s actions alone. It emerged through behavioral expectations, peer responses, professional interventions, and the unequal institutional valuation of linguistic resources. Leitner’s (2025) notion of being “verrückt gemacht” (driven crazy) captures this process: an interaction may contribute to escalation while its outcome is simultaneously individualized as evidence of disorder. This interpretation does not classify every expectation to use German or every behavioral intervention as racist or disabling. Harmful behavior also requires a pedagogical response. The relevant distinction lies between addressing that behavior and treating an already devalued repertoire as additional evidence of disturbance. The later conflict-resolution sequence demonstrates the difference. Once Italian was included as a resource of repair, the student could address the conflict without first abandoning the repertoire that had previously been constructed as part of the problem.

4.4. Implications for Educational Practice

The findings suggest that educational responses to multilingual practices should begin with the interactional situation rather than with the language choice itself. Where communication is relevant to a shared task or collective decision, professionals should attend to whether the students have sufficient access to what is being negotiated. This does not mean that every peer exchange must become transparent. Personal communication and linguistic affiliation may remain bounded without constituting exclusion or misconduct. Particular care is required when multilingual speech occurs during conflict or affective activation (see Section 4.3). A shift within a student’s repertoire should not automatically be interpreted as an additional behavioral symptom. Professionals can instead examine what preceded the shift, whether the repertoire itself had become an object of correction or ridicule, and whether the student still had an interactionally accepted way to participate. This does not diminish the need to respond to harmful behavior. This distinction separates the response to harmful behavior from the evaluation of the linguistic resources through which the student responds. At school level, implicit expectations concerning German and other languages should be made explicit and become objects of professional reflection (Skintey, 2023). Teams may need to distinguish situations in which a shared language is necessary for participation, safety, or accountability from situations in which broader repertoire use can remain available. Peer mediation and digital translation can provide situational support, but the findings also show their limits. They should complement rather than replace reliable arrangements for linguistic access.

4.5. Limitations

Several limitations should be considered when interpreting the findings. First, all observations, coding, and primary interpretation were conducted by one researcher. This enabled continuity between fieldwork and analysis but also means that the typology reflects a single researcher’s analytical perspective. Reflexive memoing, repeated comparison with the full corpus, negative and ambiguous cases, and regular interdisciplinary peer data sessions throughout the analytic process were used to challenge developing interpretations. As described in Section 2.3, these sessions exposed provisional readings to alternative interpretations and possible blind spots, but they did not constitute independent coding or inter-rater validation. Second, classroom interactions were not audio- or video-recorded. The analysis therefore cannot make transcript-level claims about precise wording, timing, prosody, or the sequential organization of interaction. Fieldnote excerpts draw on contemporaneous written documentation and immediately subsequent talking fieldnotes rather than verbatim transcripts. Interpretations of tone, affect, and interactional function were consequently based on the wider observed sequence and participants’ responses rather than on fine-grained conversational analysis. Third, the researcher was not proficient in all languages used during fieldwork. Semantic interpretations were restricted accordingly, and ambiguity was retained where the content of an exchange could not be established with sufficient confidence. Participant translations and explanations provided important contextual information but were themselves treated as situated accounts rather than transparent access to meaning.

The six schools were selected for ethnographic comparison rather than statistical representation. The findings therefore do not indicate the prevalence of particular practices across SEBD special schools. Nor are the sites represented equally in the detailed vignettes. Schools 1 and 6 provide several analytically dense sequences in the present analysis, whereas other sites appear more often through shorter contrasting or supporting examples. The typology was developed through cross-site comparison, but the prominence of individual cases in the presentation should not be interpreted as evidence that the respective practices occurred more frequently at those schools.

Finally, the researcher’s presence may have affected participants’ language choices and interactions, particularly in small classrooms where an observer remained visible. Although the researcher generally occupied a peripheral position and did not assume instructional or disciplinary responsibilities, interaction with students could not and was not intended to be eliminated completely. Familiarity increased over the extended field stays, and students occasionally translated for the researcher, commented on language use, or responded directly to questions. Such observer effects cannot be separated entirely from the ethnographic material and were therefore treated reflexively as part of the conditions under which the ethnographic material was produced.

5. Conclusions

The study examined how students in special schools for SEBD mobilized migration-related multilingual repertoires to address practical interactional problems. Its central contribution is to show that the meaning of these practices emerged within interaction and through the responses of others; they therefore cannot be treated as direct indicators of participation or behavioral disturbance. This is especially important in SEBD settings. When a migration-related repertoire was publicly devalued or treated as incompatible with accepted classroom participation, the student’s response could become readable as further non-compliance or dysregulation. Linguistic marking and behavioral interpretation could thereby reinforce one another and contribute to a situated doing of dis/ability (Köbsell, 2016). The educational implication is not that all communication should become transparent or that peer-oriented linguistic spaces should disappear. Schools need to distinguish between communication that must be accessible for collective participation and exchanges that may remain private. They must also distinguish between addressing harmful behavior and devaluing the repertoire through which a student responds. Students’ repertoire use cannot be understood independently of how schools regulate participation and respond to migration-related multilingualism (Gogolin, 2008; Huxel, 2020; Wiese et al., 2020). Institutional practices help determine which contributions are accepted or require explanation and which become associated with disturbance. Future research should examine how multilingual practices and behavioral interpretations develop together in SEBD settings and how institutional conditions shape when linguistic difference becomes associated with disturbance. Misreading or devaluing students’ repertoires may therefore contribute to the exclusionary dynamics that schools seek to address.