Book of abstracts - day 3

Valentina Cotesta, Università di Palermo; CILS Università per Stranieri di Siena.

Mediation and plurilingual repertoires in LESLLA contexts: a preliminary study on inclusive language assessment in Italy

Across Europe, increasing language requirements for access to residence rights and citizenship have reinforced the use of high-stakes standardized language tests, often applied uniformly to highly heterogeneous migrant populations. Within this context, adult migrants with low or no literacy levels, commonly referred to as LESLLA learners, remain largely invisible, despite facing the dual challenge of acquiring a new language while simultaneously developing basic literacy skills. These learners are further characterized by limited written metalinguistic awareness and low test wiseness, which intensify their vulnerability within monolingual assessment regimes. This study addresses the ethical implications of current assessment practices for LESLLA learners in Italy, where an A2 level of Italian is required for long-term residence permits. Building on critical perspectives in language assessment and recent developments in literacy education, the paper argues for the need to reconceptualize assessment constructs in more inclusive and learner-centred ways. In particular, it explores the concept of mediation as a resource for connecting emerging literacy practices with learners' plurilingual repertoires. Drawing on the action-oriented approach and the CEFR Companion Volume, mediation is conceptualized as a dynamic, situated process of meaning co-construction through which social actors facilitate communication and address linguistic, cognitive, and cultural asymmetries. Integrating mediation into assessment practices allows for an additive perspective on learners' plurilingual repertoires, understood as expressions of agency rather than deficits. In contexts of superdiversity, such an approach has the potential to support social participation, inclusion and cohesion.The study adopts a mixed-methods research design conducted in third-sector organizations in Central and Southern Italy. Data collection includes non-participant classroom observations in literacy and pre-A1 courses, semi-structured interviews with teachers and course organizers, and biographical-narrative tools inspired by the CARAP framework. The analysis focuses on three interconnected dimensions: emerging mediation strategies in classroom interaction; teachers' perceptions of current language tests, literacy processes, and plurilingual practices; learners' use of their plurilingual repertoires in everyday communicative contexts, to identify their communicative needs from a mediation-oriented perspective. Preliminary findings highlight both the misalignment between standardized test demands and LESLLA learners' profiles, particularly with respect to implicit literacy and test awareness requirements, and the potential of mediation-oriented practices. By documenting situated forms of mediation and plurilingual resource mobilization, the study contributes initial empirical insights for the development of a mediation-integrated assessment construct. Such a construct aims to better reflect LESLLA learners' communicative competences and to contribute to more equitable assessment practices.

References

Clelia Farina, Università di Palermo.

From the other side of the desk: experiencing a repertoire-based approach to additional language literacy

The aim of this workshop is to foreground the limitations of monolingual approaches to additional language (AL) literacy and to demonstrate, by engaging participants in a simulated literacy lesson, how plurilingual teaching practices can more effectively support literacy development. Over the past two decades, research across SLA, applied linguistics, and sociolinguistics has consistently shown that multilingualism and translanguaging practices constitute central resources in language acquisition and education (García & Kano, 2014; Piccardo, 2024; D'Agostino & Mocciaro, 2021; D'Agostino, Mocciaro & Ba, 2026). Yet, despite various applied projects (Piccardo et al., 2022; Ortega, forthcoming, among others), the theoretical and empirical advances in this line of work have, to date, had only limited uptake in AL literacy provision for young and young adult multilingual LESLLA learners. Even when lingua francas such as English or French mediate classroom interaction, programme design—and the instructional materials through which it is enacted—often remains organised around the target language. For newly arrived migrant learners, this creates a dual demand: they must develop basic decoding and encoding skills in an AL to which they have had little or no prior exposure, while simultaneously building the vocabulary needed to participate in and make sense of literacy activities designed for/in that same language. Within this framework, literacy and language development therefore unfold in tandem, placing substantial cognitive loads on learners. To address these theoretical and pedagogical challenges, the workshop draws on an action-research project conducted in Palermo, which led to the development of a repertoire-based plurilingual literacy model for Italian AL. The model builds on phonemes shared across learners' repertoires to establish phoneme–grapheme correspondences in the Roman alphabet, strengthening early alphabetic skills and phonological awareness (cf. D'Agostino et al., 2025). Participants will take part in a simulated initial literacy lesson in Arabic. Arabic is selected because it is unfamiliar to most participants, providing an experiential point of entry into the challenges faced by newly arrived learners who must engage with a new language and, for many (i. e. Bangla and Arabic speakers), an unfamiliar script. The workshop is organised in two phases. In Phase 1, participants are introduced to a small set of syllables through a strictly monolingual approach designed to foreground the disorientation and cognitive strain often reported in the initial stages of LESLLA literacy learning, (cf. Farina, forthcoming). In Phase 2, the same content is revisited through a plurilingual approach that mobilises participants' dominant and shared languages (e.g., Romance and Germanic languages, and if relevant Slavic languages). By moving participants to the other side of the desk, the workshop seeks to create a shared experiential basis for collective reflection on how teaching choices constrain or enable what learners do and, more broadly, their participation at the outset of LA literacy—and how a repertoire-based approach can shift both the learning process and its early outcomes.

References

Franziska Förster, Friedrich-Schiller-Universität Jena.

Real-life digital literacy practices of Arabic-speaking adult migrants in Germany

In second language education for adult migrants, there is a growing emphasis on integrating digital competencies alongside the teaching of reading and writing skills across different domains (Minuz et al., 2022). However, despite the evident interrelation between digitality and everyday literacy practices, literacy courses often struggle to connect classroom instruction with real-life demands. Consequently, digital literacy instruction remains underdeveloped. This gap reflects the still limited research on literacy practices beyond the classroom, particularly regarding digital practices among LESLLA learners (Guichon, 2024). Digital literacy practices can be conceptualized in two complementary ways. First, the involve the use of digital tools—such as smartphones or translation apps—when engaging with texts in everyday contexts. Second, they refer to interactions with digital text environments, where digitality shapes the structure, organization, and surface features of texts. While previous studies have explored migrants' digital competences and smartphone use (Guichon, 2024; Eilola & Lilja, 2021), less is known about the specific characteristics of digital texts in comparison to paper-based texts and the demands they place on LESLLA learners in real-life situations. This presentation adopts an emic perspective to examine how learners navigate digital practices and texts in everyday life, namely searching employment via online job advertisements, registering for a COVID-19 test, and participating in chat-based communication. The study draws on qualitative-reconstructive data from a recent PhD project on everyday literacy practices of Arabic-speaking adult migrants learning German as a second language. The 12 participants have at least six years of schooling and benefit from moderate to good first-language literacy skills. Nevertheless, their experiences with (digital) literacy practices highlight the considerable challenges of everyday literacy demands—challenges that are likely even greater for LESLLA learners. Methodologically, the study is based on ethnographic, multilingual interviews grounded in a view of literacy as a social and cultural practice. Participants recounted personal literacy events and shared relevant literacy artifacts. The reconstructive and detailed analysis draws on literacy practices research (Purcell-Gates et al., 2011), text linguistics, and multimodality (e.g., Brinker et al., 2024; Beißwenger, 2020). The findings underscore the need to treat digitality as a core analytical category in literacy and text-linguistic research, as it fundamentally shapes textual structures and modes of presentation. Engaging with digital texts requires not only operational digital skills but also experiential "written genre knowledge" (Perry, 2009). Furthermore, everyday literacy practices rarely consist of isolated reader–text interactions; rather, they are embedded in complex communicative situations involving face-to-face interaction, multilingual negotiation of meaning, and the simultaneous use of digital devices. The presentation concludes by situating its findings within broader theoretical debates. It argues for a practice-oriented understanding of literacy that fully acknowledges the complexity of digitally mediated interactions in second language and literacy acquisition contexts.

References

Live Grinden, Western Norway University of Applied Sciences.

The distribution of burdens within Norwegian integration policy: an exploration of how LESLLA learners are impacted by language requirements for Norwegian permanent residence and citizenship

A LESLLA learner is an adult migrant who goes through the complex process of learning to become print literate, or consolidating her print literacy, in a language which is not her first language. This presentation uses a critical applied linguistics lens to examine learners' reflections on their preparation for and performance within a testing regime in relation to current language policy. Semi-structured interviews were conducted with 18 LESLLA learners. A Capability Approach (Sen, 1992, 2000) expanded with a relational and multidimensional aspect of disability (Terzi, 2005) including Buchanan et al.'s (2000) elements for disability analysis, served as a conceptual lens to analyse the data. Using reflexive thematic analysis (Braun & Clarke, 2022) the discussion suggests how participants, in various ways, assume and embody societal and political expectations embedded in the relevant policy framework.

Ann-Kristin Helland Gujord, Universitetet i Bergen; Linda Evenstad Emilsen, Høgskolen i Østfold

Early grammatical development in a non-academic sample

L2 research is strongly limited in what multilingual contexts and learning environments it has explored, raising critical questions about the generalizability of the research (e.g., Andringa & Godfroid, 2020; Plonsky, 2023, Bylund et al., 2023), its relevance for teaching (e.g., Bigelow and Tarone, 2004) and to what extent it contributes to social justice (e.g., Ortega, 2019). The present study investigates the grammatical development of an hitherto understudied group of adult learners; refugees with little or no previous schooling (n=40) with various linguistic backgrounds. Oral data is excerpted from conversations collected at two data points (DPs) in their first year of instruction in Norwegian (weeks 8 and 40). Analyses of the early L2 language production of this learner population have proven challenging as their language differs from learner language previously described in studies of L2 learners with academic backgrounds. This is due to, for instance, high frequency of uninterpretable speech, undercommunicated/elliptic (grammatical) information, and most importantly: limited amount of data due to early stages of language proficiency. To handle these challenges, we adopt an approach centered on the use of nouns. This decision is based on an in-depth case study of the language production of two of the participants (Emilsen & Gujord, forthcoming), demonstrating that elements identifiable as nouns occur more frequently than other categories. The aim is not to describe noun-phrase development as such, but to characterise the surrounding environment – e.g. the emerging syntax – in the segments containing at least one noun, and to further evaluate whether these segments contain indications of syntactic development. We suggest that this approach captures a sufficiently large and comparable share of the learners' production to support a principled description of their Norwegian during the first year of acquisition and to identify potential traces of emerging syntax. Overall, the quantitative analyses point to clear expansion in production from DP1 to DP2, coupled with a shift away from isolated noun realisations towards longer, structurally richer (though still often nonclausal) strings. At the same time, the results underscore that development is uneven: variability between learners is substantially greater at DP2, suggesting diverging trajectories rather than uniform growth across the dataset. The study provides a much-needed empirical account of how morphological and syntactic knowledge emerges in learners with a non-academic background. In addition, the study contributed in methodological developments for grammatical analyses in non-academic samples of early language. This knowledge will expand the theoretical understanding of the emergence of grammar and the factors that create variation among learners.

References

Markéta Hajská, Pavel Kubanik, CUNI-Univerzita Karlova, Praha.

Romani: one language, two stories

Romani has several million speakers worldwide, with hundreds of thousands in the Czech Republic and Slovakia. Across Europe, the Romani people constitute a highly heterogeneous community, and the same is true of the individual varieties of Romani. From the outside, the Romani people and the Romani language are viewed as a single entity, and policy (including language policy) typically treats them as such. The Czech Republic and Slovakia are the countries where we find two varieties of Romani with vastly different levels of linguistic vitality; two linguistic communities to which the same policy has been and continues to be applied, yet with entirely different outcomes. Using this example, our lecture will provide a basic overview of the situation of Romani in the Czech Republic and Slovakia, while also serving as a foundation for broader reflection on the factors influencing ethnolinguistic vitality and also situation of these varieties (without firmly established standardization) within the educational systems of the mentioned countries.

Jeanne Kurvers, Universiteit van Tilburg; Roeland van Hout, Centre for Language Studies, Radboud Universiteit Nijmegen.

Bridging the gap between everyday reasoning and logical deduction in emergent readers: updating Vygotsky

Does being literate make us smarter? Plato once doubted this. In Phaedrus he quotes Socrates who argues that the written word appears to speak with wisdom but remains silent when questioned for clarification. While written language can only repeat itself, teachers in an oral culture can explain, clarify, paraphrase, and provide examples. Over time, the evaluation of the value of writing changed radically. Literacy became a global technology, central to education and a defining feature of modern societies. The Russian scientist Vygotsky argued that writing not only transforms our knowledge, but also how our mind works. His colleague Luria tested the idea that cultural tool such as written language foster the development of 'higher psychological processes'. He found that literacy supports more abstract and logical reasoning, especially in tasks like syllogisms. A syllogism is a structured form of reasoning, for example: "All humans are mortal. Socrates is a human. Therefore, Socrates is mortal." Decades of research across cultures and time periods show a consistent pattern: adults without literacy skills approach syllogisms differently from those who can read and write. Literate adults tend to follow the deductive rules of logical argumentation. Non-literate adults often rely instead on personal experience and everyday knowledge, answering in ways that make sense within their lived context. This contrast has been documented repeatedly (Luria, 1976; Scribner & Cole, 1981; Oesterdiekhoff, 2022; Ong, 1982; Olson, 1994; Kurvers & Van Hout, 2026). In this presentation we examine more closely how adults without literacy skills approach syllogisms. While we acknowledge the strong influence of literacy on reasoning, we question Vygotsky's assumption that departures from formal, text‑based logic indicate lower cognitive functioning. We propose two perspectives that have been largely overlooked in earlier research. The first draws on Wittgenstein's later philosophy, which holds that the meaning of words and sentences is rooted in their use within everyday activities. This aligns with the strategies we observe among non‑literate adults: they treat a syllogism as a meaningful claim about the world rather than as an abstract, self‑contained puzzle. Their answers reflect practical sense‑making rather than formal deduction. The second perspective comes from Bayesian reasoning. In Bayesian terms, conclusions depend not on the truth or falsity of premises but on probabilities—on what is likely given one's experience. This framework captures the intuitive, context‑sensitive reasoning that non‑literate adults typically employ far better than strict deductive logic does. We will use data from several studies on reasoning of adult non-literates, including our own. By integrating Wittgenstein and Bayesian reasoning, we aim to develop a more nuanced account of how non‑literate adults think. Their responses do not reveal weaker cognitive skills but alternative forms of rationality that are well adapted to real‑life situations. This broader view helps clarify the relationship between literacy, logic, and meaning, and highlights the difference between formal reasoning and everyday language use and practices.

Marta Maffia, Università di Napoli L'Orientale; Raymond Siebetcheu, Noemi Lari, Università per Stranieri di Siena; Anna De Meo, Università di Napoli L'Orientale.

Embodied plurilingualism in adult literacy learners: insights from language portraits across Italy

Adult literacy education in migration contexts is now faced with increasingly diverse scenarios, in which the intrinsic value of plurilingualism, understood as a dynamic and integrated repertoire of multiple linguistic codes (Grosjean, 2010), has yet to be fully recognized. Especially in LESLLA settings, adult learners with limited or interrupted schooling are often mistakenly viewed through a deficit-oriented lens that implicitly associates social marginalization and low literacy with limited linguistic resources. However, according to the framework of democratic language education (De Mauro, 2018) and the principles of plurilingual and intercultural education (Beacco et al., 2016), recognizing and valuing learners' entire linguistic competences and their complex communicative practices supports social integration and facilitates a non-traumatic reconfiguration of individual linguistic repertoires after migration. Based on an autobiographic approach (Cognigni, 2007; Cavagnoli, 2014; Busch, 2012), this contribution aims to observe how plurilingualism is perceived and represented by adult literacy learners of Italian. The data comes from two ongoing qualitative surveys involving more than 200 adult migrant students residing in Italy, one conducted in Campania (Naples) and the other in various provinces of Tuscany (Florence, Pisa, Pontedera, and Siena). Drawing on an already tested methodological approach (Carbonara, 2023; Maffia & Noschese, 2024), the research combines several data collection tools: a) a short socio-biographical questionnaire; b) a language portrait task; c) an oral presentation of the portrait, recorded and subsequently transcribed using CLIPS norms (Savy, 2006). Involved learners are highly heterogeneous in terms of national backgrounds (45 nationalities including Ukraine, Sri Lanka, Bangladesh, Pakistan, Mali, Gambia, Guinea, and Senegal) and reported mother tongues, as well as in terms of biographical variables, educational background in their country of origin, communication needs, and motivations for language learning. At the time of data collection, all participants were attending Italian literacy and L2 courses offered by volunteer associations, reception centres for asylum seekers, or CPIA (Provincial Centres for Adult Education). From the overall sample, this paper examines the subgroup of literacy learners, consisting of about 70 subjects. The qualitative analysis of the graphic strategies used by each learner in their portrait and the information gathered through the questionnaires and oral presentations made it possible to observe how literacy learners position languages in relation to memory, emotions, and identity, and to reveal perceptual and affective patterns linked to repertoires and also to literacy/language learning itself. Despite limited formal education and precarious living conditions linked to recent arrival or refugee status, participants demonstrate considerable linguistic capital. On average, literacy learners report five languages in their repertoires, including heritage languages, official national languages, lingua francas, regional dialect/s and languages acquired in transit countries. Through the metaphor of the body, learners were also able to report how they flexibly employ different (oral) resources in various communicative contexts and to express the complexity and difficulties of their literacy development in a second language and as adults. The specific characteristics of the portraits of literacy learners within the surveys conducted in the two Italian regions will be discussed in detail, along with some critical issues relating to the use of this task with this peculiar target group.

References

Eva Malessa, Turun Yliopisto, Finland; Live Grinder, Høgskolen på Vestlandet, Norway; Skye Playsted, Jemima Rillera Kempster, The University of Queensland.

Co-creating LESLLA-specific research ethics guidelines with and for LESLLA stakeholders

In recent years, ethical considerations relating to LESLLA learner populations have increasingly been addressed and problematized, see e.g., Gujord et al. (2023); Michaud et al. (2022); Norlund Shaswar et al. (2024), as evidenced by the call for LESLLA focused research methodologies in the recently developed LESLLA Manifesto. In this workshop we first introduce the audience to such LESLLA-specific research ethics discussions, including informed consent and voluntary participation, e.g., Bigelow & Pettitt (2016); Malessa et al. (2026); Shepperd & Dalderop (2026). Compared to Western-based researchers, LESLLA learners most often have distinctly different backgrounds in multiple ways, including their limited experience with formal schooling, their cultural historical contexts, their linguistic repertoires, and, for many, their experiences of displacement. Responding meaningfully to these dimensions requires approaches that are sensitive to asymmetries in power, knowledge, and lived experience. We then present case studies and research practices with LESLLA stakeholders regarding ethical issues from various multilingual research contexts, e.g., Australia, Finland and Norway, as a starting point for discussions and workshop activities that might draw on participants' experiences with their previous research projects. We hope to provide a space for questions/provocations/puzzling over existing ethics-related issues and practices. With this workshop, we endeavour to open up conversations that could contribute to a LESLLA community-led creation of practical research ethics guidelines for current and future LESLLA stakeholders that ideally could support particularly early career researchers and consequently also encourage more experimental LESLLA studies. The aim of this workshop is thus to initiate the co-creation of LESLLA-specific research ethics guidelines, similarly to guidelines that are LESLLA-relevant, but not LESLLA-specific see e.g., BERA (2018), Harley & Wazefadost (2023) and Krause et al. (2024), based on LESLLA-specific case studies and practices and find partners to follow-up this workshop's output for continuing discussions and dissemination.

References

Fernanda Minuz, Independent Researcher, Bologna; Alessandro Borri, CPIA Montagna, Castel di Casio (Bologna), Italy.

Unaccompanied minors in adult education: listening to their voices

Unaccompanied minor refugees in Italy represent a constantly growing group in a particularly vulnerable situation (ML, 2025a). Under current legal provisions, after the age of 15, they are usually enrolled in the first cycle of CPIAs (State Adult Education centres) (aiming for the middle school diploma), often in heterogeneous classes not specifically targeted to them. Some of them attend vocational training courses alongside this. The still-inadequate national and international research on the educational environments in their host countries highlights the specific difficulties they face and the ways these contribute to the school dropout rate. Among the obstacles are poor knowledge of the school language, especially formal/academic registers, and the imposition of a uniform educational path independent of previous education and aspirations (Baruzzi et al., 2020; Borri et al., 2025; Cerrocchi e Porcaro, 2023; D'Agostino, 2022; DeCapua e Marshall, 2022; Heinemann e Vogt, 2021; Santagati e Barzaghi, 2021). This paper reports the results of a participatory action research (Chevalier J. M., Buckles D. J., 2019) involving 12 teachers and 23 unaccompanied minors from two CPIA classes (in Parma and Castel di Casio) differing in terms of size and teaching models (face-to-face only, blended learning). The objective of the action research is to give voice to learners themselves, focusing in particular on the difficulties they perceive in learning Italian and curriculum subjects, their representation of the learning environment, and their expectations. Teachers are involved in two roles: as privileged informants and as educators. Discussions of the research questions, learners' testimonies, and teaching practices aim to underpin improvements to educational contexts.

The research questions:

1. What difficulties and strengths do learners perceive in studying Italian curriculum subjects?
2. How do they evaluate the educational path they are following in terms of present effectiveness, future usefulness, and relevance for the life each envisages?
3. How familiar are they with the educational, institutional, and relational environments and teaching style?
4. To what extent does the school context facilitate or hinder learning?

The methodology of the ongoing research is qualitative. The research phases were the following:

• 2 focus groups (at the beginning and end of the action) with teachers of Italian, mathematics, science and English, plus tutors;
• 4 cycles of classroom observation in different subject lessons (note-taking and audio recording).
• 2 cycles of semi-structured individual interviews with Arabic-speaking minors (assisted by the CPIA's linguistic-cultural mediator).
• Data analysis and feedback to teachers.

The research was conducted in accordance with the code of ethics for research with minors, and constant contact was maintained with the tutors at the host reception centres.

References

Roeland van Hout, Centre for Language Studies, Radboud Universiteit Nijmegen.

Low-educated adult second language learners and their low sensitivity to L2 inflectional morphology: effects of sound discrimination, learner and target word characteristics

Mocciaro and Young-Scholten (2026) conclude that low-literate, low-educated adults may need up to eight times longer than literate adults to acquire a new language. Among available learner samples, inconsistent or absent use of inflectional suffixes has been frequently observed. Such patterns raise the question of whether low-educated learners' inflectional systems differ qualitatively from those of highly educated learners. Competing explanations include the Basic Language Variety, the Missing Surface Inflection Hypothesis, and the Prosodic Transfer Hypothesis. Research on low-educated adult L2 learners deserves renewed attention. Most of these learners today are migrants, making the acquisition of morphological inflection particularly relevant again: what do they learn, and how? To address this, we re-analyzed an older dataset from Oldenkamp (2013), which included an impressive large sample of 130 low-educated adult learners of Dutch: 44 native speakers of Turkish, 44 of Moroccan Arabic, and 42 of Mandarin Chinese. Participants were evenly distributed across three proficiency levels (A1, A2, B1). Learners completed two tasks: a morpheme discrimination test and a sound discrimination test, focusing on the Dutch suffix sounds /ə/ (schwa) and /t/. We applied new statistical techniques—d' combined with generalized linear mixed models (GLMM)—and introduced a novel perspective by using sound discrimination performance as a predictor of morpheme discrimination. The morphological test was a picture-selection task assessing sensitivity to nominal and verbal plural inflection. We examined effects of learner characteristics (L1 background, L2 level) and word characteristics (frequency, root-final consonant), analyzing nominal and verbal inflection separately. Overall, learners' performance was strikingly low. L1 influenced verbal inflection only, while L2 level affected nominal inflection only. Word characteristics were less important than learner characteristics: word frequency had no effect, and root-final consonant influenced nominal plural sensitivity only. These findings suggest a minor L1 effect but a strong negative impact of low education. Participant performance on the sound discrimination task was better, though still imperfect. Using sound discrimination as a predictor of morphological performance yielded a significant but modest effect, indicating that perceiving sounds (input) does not guarantee their use in detecting meaning distinctions (intake). We will discuss how these results relate to current theories of morphological acquisition in low-educated adult L2 learners. Our findings underscore the profound challenges faced by low-educated adult learners in acquiring morphological inflection, even when phonological perception is relatively intact. This disconnect between input and intake highlights the need for targeted pedagogical approaches that go beyond phonetic training and address deeper cognitive and educational constraints. Understanding these mechanisms is crucial for developing effective language programs that promote linguistic equity for migrant populations.

References

Marieke Vanbuel, Universiteit Gent.

How do LESLLA learners process multimodal texts?

Texts for beginning second language (L2) learners are often multimodal, i.e., they combine text and images (Montero Perez, 2020). While images have been found to support reading comprehension and incidental vocabulary acquisition in L1 and L2 contexts (Guo et al., 2020), little is known about how multimodal texts are processed by LESLLA learners. For emergent readers, multimodal texts may be particularly beneficial because they facilitate the decoding process. At the same time, however, multimodality may induce a split-attention effect, diverting focus from the text and hindering comprehension and learning (Mayer & Moreno, 1998). This effect may be more pronounced under complex task conditions (e.g., reading to locate multiple specific pieces of information instead of a single piece) (Sweller, 1988). However, complexity may also encourage careful reading, potentially enhancing comprehension and vocabulary acquisition (Jung & Lee, 2023; Jung & Révész, 2018). Since research on LESLLA reading processes in multimodal conditions is lacking, no findings on the presence or impact of these effects are available. In this talk, I present findings from an eye tracking study that examines how LESLLA learners (N = 27) process multimodal texts under simple and complex task conditions, focusing on their attention to images versus text. Texts also included 8 novel target words. Comprehension and incidental vocabulary learning (form recognition, form recall and meaning recognition) were assessed. Eye movements were recorded with an Eyelink Portable Duo (SR Research). Results show no sign of a split attention effect: learners predominantly focused on text, regardless of condition, in line with findings for other L2 learners (Pellicer-Sánchez, 2022). However, reading patterns varied. While some participants integrated information from text, images and target words, others attended to images only before or after reading, or predominantly focused on the novel words. Learners who managed to integrate the information from all the different sources at the appropriate moments achieved higher comprehension scores and greater gains in meaning recognition than those who primarily focused on the text and only briefly attended the images. These findings underscore the importance of teaching LESLLA learners to apply reading strategies to integrate information from different input sources in order to effectively navigate multimodal texts, and to design tasks that require and facilitate such integration.