ABSTRACT. In many countries, the number of children growing up with a heritage language is steadily increasing. In Germany, around 40% of children under ten have a migration background (Statistisches Bundesamt, 2020). Migration status often overlaps with ethnic background, low socioeconomic status, and linguistic background (e.g., Autor:innen Bildungsberichterstattung, 2024). Previous studies comparing English achievement typically show that children with Arabic or Turkish backgrounds perform lower than German-speaking peers (e.g., Hopp, 2019; Keßler & Paulick, 2010), with even larger gaps reported for reading in German (e.g., McElvany et al., 2023; Stanat et al., 2022). It is unclear whether these differences also appear in bilingual German–English primary programs across reading and listening comprehension, receptive grammar, and vocabulary. This study reports data from 150 fourth graders assessed on German reading (Ein Leseverständnistest für Erst- bis Sechstklässler; Lenhard & Schneider, 2006) and English reading and listening (Primary School Assessment Kit; Little et al., 2003), receptive vocabulary (British Picture Vocabulary Scale II; Dunn et al., 1997), and grammar (Test for Reception of Grammar; Bishop, 2003). Half attended an intensive bilingual program (50% English instruction), and half received two hours of regular English lessons. Each group included equal numbers of children with Arabic, Turkish, or German backgrounds (N = 50 per background). Socioeconomic status (parent questionnaire) and nonverbal cognitive ability (Standard Progressive Matrices; Raven, 1976) were controlled. Results showed few or no significant differences between children of different backgrounds in German or English tests. However, bilingual program participants outperformed regular program students on English measures, regardless of linguistic background. These findings indicate that children with Arabic or Turkish heritage benefit from primary school foreign language programs, including high-intensity bilingual instruction. Results are also discussed in the context of third language transfer models (e.g., Rothman et al., 2019).

ABSTRACT. As migration transforms educational landscapes, multilingualism and social diversity have become invaluable assets in primary education (Atmaçasoy, Akar & Gogolin, 2025; García & Wei, 2014). This study delves into the potential of harnessing these assets in bilingual schools, with a focus on a superdiverse Dutch-German elementary school in Germany. By embracing the linguistic and social diversity of students as a foundation for learning, bilingual programs have the potential to create inclusive environments that boost both academic and social outcomes for all students, regardless of their gender and ethnic background (Cummins, 2021; Crenshaw et al., 2013). Through a qualitative approach combining classroom observations and focus groups with teachers, parents, and pupils, we explored two key questions: How are multilingual strategies and social diversity utilized in (bilingual) language classes, and how do stakeholders perceive multilingualism and social diversity at home and school? Additionally, we investigated the current use of generative Artificial Intelligence (genAI) in language classes (Berson et al., 2025), its perceived benefits and risks in light of diversity, and the practical needs of educators. Our preliminary findings, grounded in existing research (Nguyen, 2021), reveal that while Dutch and German are systematically integrated, students' diverse home languages are often overlooked in rather andro- and ethnocentric teaching practices, educational media, and curricula (Creese & Blackledge, 2015; Duarte & Günther-van der Meij, 2018). However, pupils themselves frequently leverage their home language to support peers, demonstrating the potential for more inclusive learning environments. We argue that thoughtfully implemented (genAI-generated) content can enhance inclusivity, catering to students with diverse gender and ethnic identities and promoting equitable teaching practices in superdiverse elementary schools.

ABSTRACT. The growing participation of older adults in language learning highlights the need to better understand how they approach the learning process, particularly in multilingual contexts. While research on language learning strategies (LLS) has expanded considerably, studies focusing on older multilingual learners remain scarce (cf. Gabryś-Barker, 2017; Pawlak et al., 2017). This study seeks to address this gap by examining the strategy preferences of multilingual learners aged 50-65 in Spain. The main objective of the paper is to present the language learning strategies most frequently used by older multilingual learners. The participants were 68 multilingual adults aged 50-65, with knowledge of three to seven languages, attending language centres in northern Spain. Data were collected using the Multilingual Strategy Inventory for Language Learning ([M-SILL], Dmitrenko, 2016), a questionnaire specifically designed for multilingual learners and based on an extended version of Oxford’s (1990) taxonomy (Dmitrenko, 2019). Results revealed that older multilinguals reported high or moderate frequency of strategy use for nearly all items. The most frequently used strategies included: compensation strategies (inferring meaning using linguistic and contextual clues), cognitive strategies (translating into L1 or any other learner’s language, pattern recognition, deductive reasoning, establishing cross-linguistic relationships), metacognitive strategies (self-monitoring, control of negative transfer, learning to learn, goal setting), affective strategies (risk-taking and keeping calm), and social strategies (practising languages, strategic code-switching to any language L1 or Lx). Multilinguals used strategies significantly more frequently than bilinguals, supporting a “breakthrough” (Hufeisen, 2000) or “threshold” (Kemp, 2007) effect in learning a third language. A significant positive correlation emerged between strategy use frequency and multilingualism degree. While younger learners showed higher overall strategy use, older multilinguals' extensive experience appeared to mitigate age-related differences, demonstrating that accumulated linguistic knowledge could support continued language learning in later life.

ABSTRACT. The paper investigates how lower secondary school pupils conceptualise multilingualism in relation to learning, with a particular focus on material environments. Drawing on the concept of the material culture of multilingualism (Aronin, 2018), the study situates pupils’ perspectives within the sociolinguistic context of the Czech Republic. It addresses two objectives: to explore how pupils represent multilingual elements in their everyday surroundings, and to examine how language practices are negotiated across learning experiences within and beyond school. The study employs a creative, participatory methodology inspired by Gauntlett (2007). Pupils were invited to construct, using LEGO bricks, a learning space in which they feel comfortable and safe. These material artefacts provided access to pupils’ subjective and imaginative understandings of multilingualism, highlighting the interplay between material objects and lived experience. Following an uninterrupted building phase, semi-structured interviews were conducted to encourage reflection and verbalisation. The dataset comprises 46 eighth-grade pupils from four lower secondary schools in Brno, Czech Republic. LEGO models were photographed and analysed following Purkarthofer (2018), while interview data were examined using qualitative content analysis (Mayring, 2010). Findings indicate that pupils’ imagined learning spaces are shaped by biographical experiences, affective orientations, and individual preferences. A notable discrepancy emerged between language use in perceived safe spaces and in formal school settings, offering insights into learners’ negotiations of multilingualism and learning.

Aronin, L. (2018). Theoretical Underpinnings of the material culture of multilingualism. In L. Aronin, M. Hornsby, & G. Kiliańska-Przybyło (Eds.), The material culture of multilingualism (pp. 21–45). Springer International. Gauntlett, D. (2007). Creative explorations: New approaches to identities and audiences. Routledge. Mayring, P. (2010). Qualitative Inhaltsanalyse. Grundlagen und Techniken. Beltz. Purkarthofer, J. (2018). Children’s drawings as part of School Language Profiles: Heteroglossic realities in families and schools. Applied Linguistics Review, 9(2–3), 201–223.

ABSTRACT. This talk discusses multilingual identity development in adolescent students aged 15 to 19 in the Ladin context in South Tyrol, northern Italy. The talk focuses on Ladin students‘ idiosyncratic multilingual repertoires and their self-identification as multilinguals. A critical analysis of the monoglossic native-speaker model is undertaken, and a new dynamic and creative multilingual speaker model is proposed. Recent research investigating Ladin students‘ multilingual identities (n=145) and attitudes towards their multilingualism with Ladin is presented together with methods and procedures (an online questionnaire and interviews with students provided the data for the current study) as well as selected quantitative and qualitative results based on statistical and thematic analysis. The Ladin school system with its explicit focus on multilingual proficiency and competence building is presented as a model case for the successful implementation of multilingual educational targets and multilingual identity formation in complex minority settings like South Tyrol and beyond. Dynamic systems and complexity-informed perspectives (Herdina & Jessner, 2002; Hofer 2023) provide the theoretical underpinning for my discussion of multilingual identity. References Herdina, Philipp & Jessner, Ulrike (2002). A Dynamic Model of Multilingualism. Changing the Psycholinguistic Perspective. Clevedon: Multilingual Matters. Hofer, Barbara (2023). A Multilingual Development Framework for Young Learners. Early Multi-competence in South Tyrol. Mouton De Gruyter. Language Contact and Bilingualism. ISBN:9783111104652 Hofer, Barbara (in prep.). Multilingual identities. Life in three languages. A Portrait of Ladin youth in South Tyrol.

ABSTRACT. This session explores the intersection of meme culture, family multilingualism, and digital communication. Viewing multilingual parenting as a diverse set of practices through which two or more languages are introduced, maintained, and negotiated within the family (King & Lanza, 2017; Curdt-Christiansen & Palviainen, 2023), we discuss how such practices are framed in online environments. Specifically, we examine these representations in the Polish context by analyzing memes posted regularly on a social media account dedicated to raising multilingual children.

The study follows an eclectic approach (Palviainen & Kędra, 2020) and addresses two research questions: (1) What themes and purposes characterize memes shared on the social media channels of a Polish platform dedicated to multilingual upbringing? and (2) What insights can these memes provide into multilingual parenting policies, ideologies, and practices in Poland? Drawing on research on the digital dimension of FLP (Little, 2020; Palviainen & Räisä, 2023), we argue that memes function as “capsules of cognitive and affective expression” (Wagener, 2024, p. 101) that may both reflect and shape parental language beliefs and practices.

The discussion is grounded in content analysis of a sample of memes (n = 64), together with thematic analysis of a semi-structured interview with the account’s content creator. By focusing on recurring themes, multimodal features, and communicative strategies, we discuss how memes are used to represent and negotiate language ideologies and everyday challenges experienced by multilingual families in a context often perceived as predominantly monolingual, such as Poland.

Overall, we aim to show how humor, intertextuality, and multimodal design are used to raise awareness, debunk myths, cultivate positive attitudes, and foster community among parents. In so doing, we seek to contribute to ongoing discussions on the digital dimension of FLP by positioning memes as an underexplored tool for language policy negotiation, community-building, and grassroots advocacy.

ABSTRACT. Switzerland’s political landscape provides an interesting context for examining how multilingualism shapes institutional and political communication. With four national languages and significant dialectal diversity, language regimes differ widely across regions and institutions. In the first part of this talk, I outline patterns of language choice and regulation at the federal and cantonal levels, distinguishing between mono-, bi-, and multilingual regimes. Particular attention is given to German-speaking cantons, where the coexistence of Swiss German dialects and Standard German leads to contrasting practices: some parliaments use the local Alemannic dialect as the official norm, while others mandate the standard variety. These institutional differences reveal how linguistic norms and ideologies are embedded in political structures. The second part of the talk investigates how language choice affects the tone and emotionality of political discourse at the national level. Drawing on research on the foreign language effect (FLE)—the tendency for speakers to reason more analytically and less emotionally in their second or additional language (Lx)—I analyse a corpus of over 2,100 conversational turns from ten Swiss politicians who appear in televised debates in both French and German. For each politician, one of the two languages is their L1 and the other their Lx, enabling within-speaker comparisons. Each utterance was analysed for sentiment, subjectivity, and polarizing potential using Natural Language Processing tools, human raters, and a large language model (LLM). The results reveal a consistent pattern: when speaking in their Lx, politicians tend to use less negative, less subjective, and less polarizing language. These findings suggest that Switzerland’s multilingual language regimes not only structure who speaks which language but also influence how ideas are expressed. The institutional and cognitive perspectives discussed reveal that multilingualism shapes both the organization and the emotional texture of political communication.

ABSTRACT. Teachers’ beliefs about multilingualism are increasingly recognized as central to how linguistic diversity is addressed in education. Recent research in teacher cognition and language ideologies has shown that these beliefs shape not only classroom practices but also broader orientations toward inclusion and language policy (e.g., Fischer & Lahmann, 2020; Paquet & Woll, 2021; Villabona & Gartziarena, 2025). The province of Quebec presents a unique context for such exploration, as it has been historically marked by language tensions and strong socio-political efforts to preserve the French language. The present study is set in the Montreal area, one of the most linguistically diverse cities in Canada. Within this complex landscape, we aim to explore how teachers perceive multilingualism not only as a pedagogical issue, but also as a broader societal phenomenon. This study examines the beliefs of pre-service and in-service language teachers regarding multilingualism and explores whether these beliefs differ according to teachers’ dominant language, professional experience, and the language(s) they teach. Drawing on data from 57 participants, the study combines an online questionnaire with follow-up focus group discussions. Adopting a mixed-methods design, the analysis captures both broad belief patterns and the ways teachers make sense of multilingualism.

Fischer, N., & Lahmann, C. (2020). Pre-service teachers’ beliefs about multilingualism in school: an evaluation of a course concept for introducing linguistically responsive teaching. Language Awareness, 19(2), 114–133. https://doi.org/10.1080/09658416.2020.1737706 Paquet, P.L & Woll, N. (2021). Debunking student teachers’ beliefs regarding the target-language-only rule. Multilingualism and third language acquisition: Learning and teaching trends, 2, 95. Villabona, N. & Gartziarena, M. (2025). What are the languages in my school? Pre-service teachers’ decisions when designing ideal schools in the Basque context. Current Issues in Language Planning, 26(4), 509–524. https://doi.org/10.1080/14664208.2024.2391717

ABSTRACT. We report on a study of L1 Norwegian, L2 English, L3 French learners’ realization of negation in written production. Data come from a corpus of learners (age 16-18) in their fourth and fifth year of learning French in a formal, non-immersion context. A previous study (authors, in press) investigated verb placement in non-subject initial clauses and relative to adverb placement in the same learners/corpus and found some indications of non-target lack of verb movement in the latter sentence type, compatible with transfer from L2. In the present study, we investigate verb placement relative to negation, and, furthermore, analyze other error types associated with realization of negation in L3 French. For negation, we find no indication of non-target lack of verb movement, as finite verbs invariably appear before the negative element pas. However, we see a number of non-target realizations of negation in other respects: Learners sometimes produce non-finite verb forms in finite contexts, arguably indicating an issue with morphology (Prévost & White, 2000; Slabakova, 2019), as well as non-target contracted forms that may reflect problems associating phonology and orthography. Finally, we see non-target placement of clitic elements relative to verbs and to the negative elements ne and pas. In line with previous research, our findings indicate that verb movement is less problematic with negation than with sentence-medial adverbs (Hawkins et al., 1993; White, 2024). Errors do not seem to reflect transfer from L1 or L2; rather, they are linked to the idiosyncratic realization of negation in French. We hypothesize that it is exactly the complexities of French negation that contribute to its saliency, which may in turn facilitate target-like verb movement in sentences with negation before other sentence types.

ABSTRACT. Although cross-linguistic influence is considered multidirectional, limited research has examined how acquiring a third language impacts previously acquired languages. Proposed modulators of regressive CLI (rCLI) include age of acquisition, L1–L3 similarity, typological proximity, and structural optionality (e.g., Puig-Mayenco, 2023, Olszewska & Długosz, 2025).

Participants were speakers of L1 Polish, L2 English, and L3 Scandinavian (Swedish, Danish, Norwegian). Two groups matched for L2 English were compared: L3 learners (n=163) and L2 learners without Scandinavian (n=71). They completed an untimed acceptability judgment task (AJT), a self-paced reading task (SPRT), and an oral production task (OPT). Proficiency was assessed with Dialang, and all tasks were completed in both languages.

We investigated three V2-related phenomena that differ in frequency and input ambiguity: subject-verb inversion in main clauses with fronted adverbials or objects, and clause-medial adverb placement. V2 is obligatory in Scandinavian, absent in English, and marked in Polish; adverbials are routinely fronted, while objects rarely appear clause-initially. Clause-medial adverbs are post-verbal in Scandinavian, pre-verbal in English, and variable in Polish.

Linear mixed-effects models reveal strongest rCLI in L2 English for V2 with adverbial fronting, weaker effects for adverb placement, and no effects for V2 with object fronting, mirroring L3 input frequency. In AJT, non-target L2 English items receive higher ratings only at intermediate L3 levels, returning to baseline at advanced levels. In SPRT, both groups show sensitivity to violations, but L3 speakers recover more quickly. Group differences in OPT are present.

We interpret these findings within the Linguistic Proximity Model (Westergaard et al., 2017), arguing that V2 in the L3 grammar is co-activated during L2 processing. However, because L2 grammars are more stable, co-activation does not lead to L2 grammar restructuring. We propose that lasting rCLI requires favourable conditions that involve optionality in the target language, where the L3 reinforces an existing option.

ABSTRACT. This study examines crosslinguistic influence (CLI) in L3-English ambiguous scope interpretation by L1-Kazakh L2-Russian and L1-Russian L2-Kazakh speakers as well as simultaneous Kazakh-Russian bilinguals who grew up speaking both languages. Research on scope ambiguities reported that there are crosslinguistic differences in interpretation preferences. For example, the sentence The boy did not get all the balloons is ambiguous between a partitioned set (i.e. some of the balloons’) and a full set interpretations, (i.e., none of the balloons’). Languages in which the negator c-commands the quantified phrase (QP) such as English and Russian prefer the partitioned set interpretation (Lidz & Mousolino, 2002), while languages like Korean and Kazakh in which the QP c-commands the negator prefer the full set interpretation (Han et al., 2007). This study looks at how crosslinguistic influence manifests in L3 acquisition of scope interpretation by sequential vs. balanced simultaneous bilinguals and what motivates transfer selectivity between two available linguistic systems (proficiency, language dominance and use)? This study employs a forced-choice picture selection (FCPS) task in which participants are asked to choose a picture that matches the meaning of a sentence. The FCPS task presents a sentence with two story-strips illustrating an event, as in Figure 1 (12 targets and 24 fillers). Participants’ proficiency levels in English, Kazakh, and Russian are measured using modified versions of CEFR and a modified/translated version of the Bilingual Language Profile questionnaire (Birdsong et al. 2012) is used to measure Kazakh-Russian simultaneous bilinguals’ language dominance and use. Data collection is ongoing and will be completed by the conference date.

ABSTRACT. While grammatical processing has been extensively examined across languages, relatively few studies have linked it to psychological variables such as personality. We propose that one relevant trait, Tolerance of Ambiguity (TA), reflecting comfort with uncertainty and multiple interpretations (Herman et al., 2010), may predict how structural ambiguity is resolved during grammatical processing.

We investigate the interpretation of ambiguous relative clauses in L2 English (e.g. I saw the servant[NP1] of the actress[NP2] who lives in the neighbourhood) among 126 young adult native speakers of Polish (L1) learning L2 English only (bilinguals; n = 63) or L2 English and L3 German (trilinguals; n = 63). The two groups were matched for L2 English proficiency (Dialang). Prior work shows that Polish and German generally favour high attachment (NP1 modification), whereas English more often favours low attachment (NP2 modification) in ambiguous NP1-of-NP2-RC constructions (Hemforth et al., 2015).

Both groups completed an English sentence-interpretation task comprising globally ambiguous and unambiguous RC sentences (e.g., They called the waiter of the princess who crowned herself), responding to a subject wh-question by selecting NP1, NP2, or ‘not sure’. TA was measured using the 12-item short form of the Tolerance of Ambiguity Scale (Herman et al. 2010).

We modelled the proportion of high-attachment responses in L2 English using generalized mixed-effects regression. The results show that TA interacts with L3 German knowledge: its effect is significant only in trilinguals. At higher TA levels, trilinguals display greater variability than bilinguals, fluctuating more often between NP1 and NP2 interpretations. Moreover, trilinguals select NP1, the interpretation preferred in German, significantly more frequently. We propose two possible explanations: (1) a direct influence of L3 German, modulated by TA, and (2) an indirect effect of multilingualism. Finally, increasing L2 proficiency leads to convergence on target-like RC attachment preferences, contrary to Clahsen et al. (2003).

ABSTRACT. Previous studies examining the role of case marking and word order in cross-linguistic structural priming obtained inconsistent results mostly in bilingual context and between languages with typologically close similarity. The present study extended the research into a trilingual context and between languages that are typologically different by examining cross-linguistic priming of ditransitive structures among Tibetan-Chinese-English trilinguals whose three languages are different in case marking (Tibetan is a case marking language while the other two are not) and word order (Tibetan is SOV but the other two are SVO). The results showed that the priming effect occurred across the three languages, indicating case marking and word order did not hind cross-linguistic priming among the trilinguals. Besides, language proficiency was found to have no moderating effect on the priming process. The results also suggest shared representation of ditransitive structure among Tibetan-Chinese-English trilinguals.

ABSTRACT. In this presentation we apply Dominant Language Constellations (see e.g. Aronin & Vetter, 2021) to study students in higher education (universities and universities of applied sciences) in Finland. At the outset of the study, the participating students were identified as Finnish-Swedish bilinguals, either grown up with the two languages at home, or introduced to the Swedish language by education in two languages in Swedish immersion (see e.g. Björklund, 2019). Regardless of the type of bilingualism (simultaneous or successive bilingualism), the collected data clearly shows that the students are users of multiple languages going beyond only Finnish and Swedish, and that bilingualism tends to have an increasing effect on the acquisition of additional languages (see e.g., Otwinowska et al., 2023).

The study contributes to previous studies on DLCs in education at an institutional level in Finland (Björklund, 2026; Björklund et al., 2020; Björklund & Björklund 2021, 2024) with a new methodological approach that combines individual and institutional DLC charts during on-line co-constructions with individual respondents. Data comprises 19 DLC charts combining self-reported languages practices of simultaneous and successive bilingual higher education in and outside their on-going higher education studies. A systematic inductive analysis of content (Barkhuizen, 2019) was conducted to study the data.

First results show that the number of languages in institutional DLC charts is smaller (on average 1,7 languages per respondent) compared to the number of languages displaying the DLCs of respondents’ leisure time (on average 2,3 languages per respondent). The institutional DLC charts consist mainly of two languages representing the national languages (Finnish and Swedish), one of them being the main language of the respondents’ on-going studies. English is reported in DLCs at an institutional level by 6 of 19 respondents, belonging more often to their linguistic repertoire together with other additional languages such as Norwegian and German.

ABSTRACT. Cognates, words sharing form and meaning across languages, are processed faster and more accurately than non-cognate words leading to cognate facilitation effect. While facilitation is robust for identical cognates (e.g. BANK-BANK), findings for non-identical cognates characterized by partial form overlap (e.g. TUNEL-TUNNEL) remain inconsistent (Arana et al., 2022; Comesaña et al., 2015; Tiffin-Richards, 2024; Vanlangendonck et al., 2020).

This presentation reports results from two studies with Polish-German-English trilinguals that jointly address the sources of variability in non-identical cognate processing. In both studies, participants translated single words from L3 into L1 and from L3 into L2.

The first study extends earlier work by AUTHORS (2021). In that study, 44 trilinguals translated Polish-English and German-English non-identical cognates (e.g., AKTOR-SCHAUSPIELER-ACTOR; PIWO-BIER-BEER) and English non-cognate controls (e.g., ROWER-FAHRRAD-BIKE). Cognate facilitation was observed in accuracy but not in translation latencies, a pattern attributed to relatively low L2/L3 proficiency. To directly test this explanation, new data were collected from an additional group of 20 higher-proficiency trilinguals who completed the same tasks. To increase statistical power, the new data were combined with data from the most accurate participants from the original study. Analyses of the combined dataset showed that increased proficiency did not yield facilitation effects for non-identical cognates in translation latencies, nor did it strengthen accuracy-based cognate effects.

Motivated by these findings, the second study examined whether stronger cross-language form overlap leads to more robust facilitation. To that end, 52 Polish-German-English trilinguals were asked to translate Polish-German-English cognates (e.g., BALON-BALLON-BALLOON), and matched English non-cognate controls. Results revealed significant cognate facilitation in translation latencies irrespective of response language, while cognate effects in accuracy were modulated by translation direction.

Together, the two studies indicate that the degree of cross-language form similarity, rather than language proficiency alone, is a key determinant of cognate facilitation in translation production in trilinguals.

ABSTRACT. Plurilingual approaches in education are widely regarded as key to the development of plurilingual competence and as facilitators of L3 acquisition. Nevertheless, from a research perspective, the field of plurilingual assessment remains considerably less developed than research on plurilingual pedagogies or plurilingual learning strategies (see Vogt & Antia, 2024 and Melo-Pfeifer & Ollivier, 2024, for exceptions). To further complicate this state of affairs, a recent large-scale European quantitative study found that, even when language teachers are willing to implement plurilingual teaching practices in order to value their students’ linguistic repertoires, they are far less inclined to adopt plurilingual assessment practices (Cortéz Velasquez et al., 2024). This gap between the adoption of plurilingual pedagogical practices and the persistence of a monolingual mindset in evaluation jeopardizes both the potential positive outcomes of plurilingual pedagogies and the sustainability of such practices, due to washback effects. This presentation addresses plurilingual assessment from the perspectives of language teachers and plurilingual students (Borg, 2017 on beliefs), drawing on two qualitative studies: one focusing on teachers’ beliefs and the other on students’ beliefs on plurilingual assessment. The corpus of teacher interviews was collected in 2017 and comprises nine participants, while the student data were collected in 2024 and include seven students from secondary schools in Hamburg. Using qualitative content analysis (Mayring & Fenzl, 2019), we examine how teachers and students conceptualize plurilingual assessment, how they perceive its potential contribution to reducing institutional discrimination against plurilingual students who are still acquiring the language of schooling, and how they evaluate the advantages and challenges of its implementation. The findings suggest that both teachers and students, while critically reflecting on the monolingual and standardized assessment regimes in schools, experience difficulties in envisioning alternative assessment approaches and ultimately tend to accept the existing assessment habitus without fundamentally questioning its status quo.