Becoming a multilingual health professional in vocational education - two adult migrants’ translanguaging trajectories

The number of migrant students in vocational education in Finland is rising. Their educational backgrounds and language resources are diverse, and research is needed to gain better understanding on how their existing knowledge can be acknowledged, supported, and deployed when they become multilingual professionals in their own ﬁ elds. In this ethnographic study, we drew on the theoretical approaches of translanguaging and language architecture (García, O., and T.Kleyn. 2016. “ Translanguaging Theory in Education. ” In Translanguaging with Multilingual Students: Learning from Classroom Moments , edited by O. García, and T. Kleyn, 9 – 33. New York, London: Routledge; Li, Wei. 2018. “ Translanguaging as a Practical Theory of Language. ” Applied Linguistics 39 (1): 9 – 30; Flores, N. 2020. “ From Academic Language to Language Architecture: Challenging Raciolinguistic Ideologies in Research and Practice. ” Theory Into Practice 59 (1): 22 – 31). Both concepts promote the integration of students ’ multilingual resources in their content studies. We focused on two multilingual students studying to become practical nurses. We applied small stories analysis (Georgakopoulou, A. 2015. “ Small Stories Research: Methods - Analysis - Outreach. ” In The Handbook of Narrative Analysis , edited by A. De Fina, and A. Georgakopoulou, 255 – 271. Wiley Blackwell) to ascertain how translanguaging practices support them in developing their vocational competence and language architecture. The data comprised ethnographic observations, interviews, and audio-recorded interaction. Although the o ﬃ cial language of instruction is Finnish, students are encouraged to integrate their multilingual resources into their vocational development. The analysis illustrates how translanguaging not only deepens multilingual students ’ understanding of ﬁ eld-speci ﬁ c content but also enables them to strategically use and strengthen their multilingual resources.


Introduction
The proportion of migrant students in vocational education (upper secondary education) in Finland is increasing: currently, 23% of all vocational education students nationwide and 39% of those in the metropolitan area have a first language other language than Finnish or Swedish or do not have Finnish citizenship (StatFin 2022). Unlike in the other Nordic countries, students with migrant backgrounds in Finland are more often guided to vocational than higher education irrespective of their educational background (Dunlavy et al. 2020;Hoffman et al. 2020). Moreover, students from migrant backgrounds with higher education degrees often choose vocational paths in order to find employment. Vocational education enrolments also include a growing number of late arrivals, i.e. students who have only been in the Finnish educational system for a couple of years, and whose previous schooling paths may have been interrupted (Mustonen 2021). Thus, students in vocational education are highly diverse in their educational and biographical backgrounds, language and literacy skills, content knowledge, vocational skills, and wishes and goals for the future.
Finnish vocational education provides students with qualifications and capabilities for entry into various fields of working life. Before entering vocational education, students will have completed at least their basic education either in Finland or in their former home country. Instruction is offered mainly in Finnish (or Swedish in Swedish-speaking municipalities, due to the Finland's official bilingualism). Some institutions also offer a few programmes and qualifications in English. Preparatory education is available for those whose language and literacy skills need further development before they enter vocational programmes; this transition phase may also include gaining some acquaintance with the field of interest. However, as national language requirements for vocational education no longer exist (Ministry of Education and Culture 2019, 30), many students with migrant backgrounds find themselves developing their Finnish language skills in parallel with their field-specific content studies.
The current socio-constructivist and socio-cultural alignments and ideals of education call for a deeper understanding of the diverse knowledge and skills students already have. Accordingly, the value of multilingualism as a valuable learning resource has, with the exception of vocational education, for several years been acknowledged across curricula (EDUFI 2016) in the Finnish educational system. Thus, the experiences and existing skills of migrant adult learners may be overlooked (Pöyhönen, Suni, and Tarnanen 2019).
Many recent studies on multilingualism in education have applied a translanguaging approach (e.g. García and Kleyn 2016). In this article, we draw on the notions of translanguaging (Li 2018) and language architecture (Flores 2020), both of which emphasise recognising and encouraging the flexible and most optimal use of students' multilingual resources to support their equality and participation in education along with their learning processes. However, the need remains to investigate and develop ways of implementing these ideas in practice, both pedagogically and in relation to learners' own translanguaging practices. In this article, we use the concept 'translanguaging trajectory' to analyze how students strategically leverage their multilingual resources in building new knowledge and skills, and how their language architectures develop over time. The term 'trajectory' highlights the idea that while translanguaging practices are situational and dynamic, they nevertheless develop and exhibit coherence over time (see Wenger 1998, 154). While teachers' translanguaging trajectories (David et al. 2022) and multilingual workers' transnational trajectories (see Garrido and Sabaté-Dalmau 2020;Kraft 2020) have previously been studied, the perspectives of multilingual students have received less attention.
We focus on two multilingual students in a group of students studying for a vocational qualification in social and health care: one from Somalia who had completed some higher education studies before coming to Finland and one who came from Afghanistan to Finland as an unaccompanied minor with an interrupted schooling path. Both participants are developing their L2 skills (Finnish) alongside their vocational content studies. We also investigate their educational paths and experiences before they started vocational education, as these experiences are an integral part of their trajectories and should be recognised in education. Our research question is 'How do multilingual students' translanguaging practices support their developing vocational competence and language architecture?' Methodologically, our study is based on team ethnography (Blackledge and Creese 2010) and the small stories approach (Georgakopoulou 2015). We apply narrative positioning analysis (Bamberg and Georgakopoulou 2008) to ethnographic data, including interviews with the two students and ethnographic observations and audio-recordings gathered during their vocational education programme. Our findings suggest that translanguaging supports multilingual students in their objective of becoming multilingual health professionals and independent users of Finnish.

Theoretical background
The multilingual turn in language learning research (May 2015;Conteh and Meier 2014) calls for critical consideration of pedagogical practice. Language learning is perceived as a complex, multilingual, multimodal, situated social process (see, e.g. Meier 2017) in which the learner's agency, personal biography, and identity struggles, intertwined with wider socio-political contexts and power relations play a crucial role (Darvin and Norton 2015;Meier 2017). Thus, a student's whole linguistic repertoire should be considered as a resource for interaction and learning.
However, tensions between the monolingual language policies of education institutions and multilingual students' translanguaging practices have been widely noted (Li and Lin 2019, 209). This issue calls for more research in order to develop new kinds of critical and flexible multilingual pedagogical practices.
We contribute to this discussion by drawing on the theoretical approaches of translanguaging (García and Kleyn 2016;Li 2018) and language architecture (Flores 2020). Both concepts challenge the existing rigid boundaries and hierarchies between different languages (and varieties of languages) and normative views on the use of different linguistic resources in interaction and learning.
Translanguaging as an approach focuses not on languages per se, but on how multilingual individuals deploy their resources to gain their communicative goals in education and other contexts, and to invest for their future (García and Kleyn 2016, 12, 14;Canagarajah 2011;Meier 2017). Translanguaging can also be defined as a pedagogical practice that values and encourages learners to leverage all of their linguistic resources strategically and selectively (García and Kleyn 2016;Li 2018;Cenoz and Gorter 2020). The prefix trans refers to the transformative nature of language practices on both the individual and educational levels: the aim of translanguaging practices is to critically consider and dismantle old structures and hierarchies and promote the equal participation of all (García and Li 2014, 3;Li and Lin 2019, 210, 212).
In the classroom, creating a translanguaging space means that teachers adopt a stance that allows students to build on and develop their linguistic repertoires and design learner-centered, collaborative learning activities with multilingual resources that promote learners' creativity and criticality (García and Kleyn 2016, 20-24;García, Johnson, and Seltzer 2016;Li 2011). Translanguaging also promotes amplifying and deepening their understanding of the academic content (see also Walqui and van Lier 2010;Meyer et al. 2015;Cenoz and Gorter 2020). Thus, translanguaging is, in essence, an approach that empowers learners, i.e. positions them as agents of their own learning and identity development (see also Gynne 2019). One translanguaging practice is brokering, i.e. a peer student crossing the linguistic and cultural borders and thus also mediating the content to their peers (Orellana and García 2014).
A translanguaging space enables students to draw on their histories, experience, outlook and knowledge (Li 2011). Here, we use the concept translanguaging trajectory (see David et al. 2022) to highlight the changing and developing nature of multilingual students' translanguaging practices and language architectures over time and across spaces. This choice also allows us to take into consideration previous life experiences that affect these students' translanguaging practices.
The concept language architecture seeks to dismantle the traditional hierarchy in which learners' prior language practices are perceived as less abstract and complex than their academic (L2) skills (Flores 2020; see also Flores and Rosa 2015). According to Flores (2020), educators should recognise the complexity of students' language and literacy skills, and support them in using these resources in interpreting, analyzing, and producing texts. As multilingual students constantly encounter negotiations on choosing between language varieties for which they already possess the requisite strategies, the educator's task is to recognise these skills and promote their students' development as 'language architects' (Flores 2020). Thus, multilingual students' resources should be central in the learning process and be genuinely integrated into the task of building their academic competencies and identities (Flores 2020;Rosa and Flores 2017; see also van Lier 2007, 47). The concept language architecture is not only intertwined with but also specifies the concept of translanguaging, because translanguaging has, despite its good intentions, been criticised for being too polysemic, and also for being a dominating approach rather than a liberating force (Jaspers 2018). Language architecture focuses on genuinely valuing and integrating students' entire linguistic repertoire into their professional development while also giving them tools to resist raciolinguistic ideologies (Flores 2020).
Within a broad field of translanguaging research, Cenoz and Gorter (2020) acknowledge a need to develop practices that take the specific linguistic, social, and educational context into consideration. In the following section, we describe the contexts of this study.

Data and methods
This study forms part of an interdisciplinary project Building Blocks -Developing Second Language Resources for Working Life, funded by the Academy of Finland (2019-2023), targeted at exploring the language resources needed and developed by adult immigrants whose aim is to enter the Finnish labour market. To obtain rich data on these issues in vocational education, we applied team ethnography (Blackledge and Creese 2010) and undertook fieldwork in different vocational domains, such as health and social care, building maintenance technology, surface finishing, hairdressing, cleaning, and security. The previous studies of the project found that in many fields the multilingualism of the students was not valued or deployed (see Mustonen andPuranen 2021 andMustonen 2021). In this study, we wanted to make multilingualism visible and analyze how translanguaging practices support learning. Our observations indicated that the most varied multilingual practices were deployed by a group of adult migrants who were studying the vocational qualification in social and health care to become practical nurses. Therefore, we decided to focus on this group.
The study programme in which this group was enrolled had been tailored for adults with migrant background. It had begun with a three-month period of preparatory studies aimed at strengthening the basic skills needed to succeed in vocational training, including Finnish language classes. We conducted ethnographic fieldwork with this group during the academic year after their preparatory studies, starting in August 2020 and ending in June 2021. Observational data were gathered on the 27 days during the 10-month period when we attended their lessons, wrote fieldnotes, took photographs, conducted interviews, and held informal discussions with the participants. Thus, the data are longitudinal and also included a retrospective perspective, as in their interviews, the students reflected on their past life experiences and educational paths.
Each student (16) in the group signed an informed consent agreeing to be observed and audiorecorded, and seven students also expressed interest in taking part in the individual interviews. With them, we conducted 14 interviews in total. For this article, we focus on two key participants: Ali and Omid. Ali was interviewed twice. The questions were initially asked in Finnish and in the few instances when Ali replied in English, the interviewer also replied back in English. Omid was interviewed three times, twice in Finnish and once in Persian. We also interviewed their teachers. Unfortunately, the COVID-19 pandemic hindered our access to the students' practical training in day care and elderly care centres. However, we were allowed to attend and record four students' evaluation discussions, including Omid's, at the end of their practical training period.
Several researchers were involved in this team ethnographic study. Maiju Strömmer followed the key participants' classroom days and interviewed them with research assistant Rosariina Suhonen, whereas Sanna Mustonen was conducting ethnographic research in other vocational education fields but had not met the present participants. We believe that this benefited the analysis, as it enabled the combining of two different perspectives on the data: that of an ethnographer who has lived experiences with the group with that of a researcher who is familiar with the educational context, but whose observations on the data as an outsider strengthens the validity of the findings. Maiju, Rosariina and Sanna are white women who were born in Finland and whose first language is Finnish. This means that we have to carefully consider our position when analyzing the data. Another member of our research team, Amin Rasti Behbahani, who is originally from Iran and speaks Persian, was able to interview our Persian-speaking key participants in Persian, including Omid. The language, culture, and experiences of migration shared by the participant and Amin meant that the interview on the topic of a multilingual speaker's experiences were more profound and multifaceted. In this article, we present the interview extracts in the original Persian but confine our analysis to their English translations. In the English translations of Persian and Finnish extracts, the non-standard varieties are not visible, as the target of the analysis is not the linguistic forms, but the narratives told.
To analyze the multilingual students' translanguaging trajectories, we applied the small stories approach (Georgakopoulou 2015;Bamberg and Georgakopoulou 2008), which offers tools for analyzing non-canonical stories, such as brief and interrupted tellings, tellings of ongoing, future, or hypothetical events, or refusals to tell. Telling activities can take place through different channels in different modes and do not necessarily have a clear beginning, middle and end (Georgakopoulou 2015, 258). Here, we apply, in particular, the positioning analysis developed by Bamberg and Georgakopoulou (2008). By exploring how the participants perceive themselves in relation to others in their narratives, we gain understanding of how they see their (social) role and how they want to be understood (Bamberg and Georgakopoulou 2008, 380). It is also important to analyze how the multilingual participants are positioned by others (e.g. teachers and supervisors) in narratives (see Smith 2019) and in interaction situations (see Bamberg and Georgakopoulou 2008, 386-388), as the idea in pedagogical translanguaging is to provide everybody present with opportunities for participation and agency. The concept language architecture, in turn, renders visible the role of ethnic and other background factors in positioning language learners and valuing their skills.

The translanguaging trajectories of two multilingual adult students
Ali: deploying prior expertise in vocational education Ali is originally from Somalia but lived in other African and European countries before moving to Finland. He has strong prior medical and anatomical knowledge as he spent three years in medical school in an East African country. There, while the language of instruction was English, he also learned field-specific terminology in Latin. However, he could not afford to finish his studies. He moved to Finland in 2018 to live with his family: he had met a Finn in another European country. They had got married and had two children but were now divorced. In Finland, he entered integration training, which included studying Finnish for six months alongside practical training in elderly care for 2 months and in a day-care centre for 2 weeks. In the interview, he stated that he was strongly motivated to learn Finnish as he wanted to pass the language test for the practical nursing programme in vocational education. He passed the test and started his studies in March 2020. His plan is to continue his studies after vocational education, first in a university of applied sciences to become a registered nurse, and then in a medical school to become a medical doctor.
Ali's language repertoire includes Somali, English, Amharic, Orom, Finnish, Arabic and to some extent Latin, and he has also lived in several countries. However, as Li (2018) and Wu and Lin (2019) have argued, the process of translanguaging goes beyond named languages and is multimodal in its nature. Thus, different multi-sensory communicative resources are combined when making meaning. The extract below is a fieldnote describing Ali's self-portrait, which was a classroom activity designed by their teacher. In an art class, the students were asked to make self-portraits using the collage technique: Example 1: fieldnote, art class (originally in Finnish, English translation below) Ali näyttää omaa paperiaan ja kertoo, että se on puoliksi valkoihoinen ja puoliksi tummaihoinen ja välissä on arabiviikset ja egyptiläinen parta. Tosiaan: paperin yläreunaan on leikattu vaaleaihoisen ihmisen kuvasta yläosa, leukaosa on tummempi, ja välissä on tummat viikset. * * * Ali shows me his paper and tells me that it is half white and half black and with an Arabic mustache in the middle and an Egyptian beard. In fact, the upper part of the paper shows a cut-out of the upper part of a white person's face, the chin is darker, and between the two is a dark mustache.
This multimodal small story about a self-portrait is a mixture of different ethnic elements which the student also verbally points out. It illustrates that he positions himself both visually and verbally in relation to different cultural elements. He has a history of living, studying, and working in different cultures; his life trajectory has not been linear but has included many changes such as interrupted studies in medical school in Ethiopia (where Amharic and Orom are spoken), relocating to new environments in Europe, and a transnational marriage with a Finnish woman, and a divorce. Thus his self-portrait may be interpreted as an illustration of how these fragmented elements are embodied in how he positions himself. Ali's prior life experiences were acknowledged by his teachers. Based on our observations, he was positioned several times as knowledgeable on field-specific content, as evidenced in the next example drawn from an anatomy lesson. The main teaching method used by the anatomy teacher is collaborative classroom interaction. Such interaction can be viewed as an example of the co-constructed telling of an on-going event. As García and Li (2014, 3) have argued, the suffix ing in translanguaging draws attention to such ongoing, momentary activities that occur in classroom interaction. The teacher acknowledges the language skills and previous knowledge of this multilingual group, in this example Ali's knowledge of Latin terms in particular: Example 2: classroom interaction, anatomy class (originally in Finnish, English and Latin; the English translation is given below, with the original English words underlined) T = the teacher, S = a student, A = Ali 1 T: sitten otamme kolmannen elinjärjestelmän joka myöskin poistaa aineita ja se on meillä tämä iho 2 S: iho 3 T: ja nyt vaikka tällä minun kurssilla emme opiskele hirveesti latinaa paitsi mitä meillä sulla tulee hyvin niitä vierasperäisiä termejä niin voidaan niitäkin tässä kääntää niinku pancreas on latinasta suomeen käännettynä haima 4 A: haima joo 5 T: tai hepar on 6 A: liver 7 T: hepar on maksa 8 A: maksa joo 9 T: j-ja niin edelleen mutta nyt yks semmonen termi joka otetaan tässä nyt oikeestaan muutama termi jotka liittyy tähän ihoon 10 A: dermis ei dermis 11 T: jees mutta nyt ensin katotaan tämä iho tähän kokonaisuutena (.) ihon latinankielinen nimi on tämmönen ku cee uu tee ii äs 12 All: cutis 13 T: cutis C-kirjaimella kirjoitettuna (.) ja se rakentuu kahdesta kerroksesta joihin nämä kaksi ylintä viivaa tässä viittaavat (.) ja näillä ihon kahdella kerroksella on sitten taas 16 omat nimensä (.) otetaan tämä paksumpi alempi kerros tästä ensin tässä (.) paksummassa kerroksessa tällä alueella tällä vyöhykkeellä on erittäin tiheä verisuonitus ja siitä tämä alue on saanut suomenkielisen nimen verinahka mutta nyt emme kirjoita sitä suomenkielistä nimeä vaan otetaan tämä ammattitermi jonka mainitsitkin jo se on 14 A: dermis ja epidermis *** 1 T: then we take the third organ system that also secretes waste substances which is skin 2 S: skin 8 A: 'maksa' yes 9 T: a-and so on but now there's one term that we take or actually a few terms that are related to the skin 10 A: dermis no dermis 11 T: yes but first we'll take a look at the skin as a whole (.) the Latin word for skin is like this C U T I S 12 All: cutis 13 T: cutis written with C (.) and it is formed from two different layers indicated by these two lines at the top (.) and there two layers of skin again have their own names (.) let's take this thicker lower layer here first (.) in the thicker layer in this area in this zone there is a very fine net of blood vessels and that is why this has the Finnish name 'verinahka' [direct translation: blood skin] but now we won't write the Finnish name but we'll take this academic term that you already mentioned it is 14 A: dermis and epidermis In this excerpt, the teacher positions Ali as having expertise not only in anatomy but also in the international language of medicine. The teacher offers him space to use his multilingual resources and prior knowledge. As Flores (2020) emphasises, a multilingual student's existing resources should not be conceived as less abstract or 'academic' than the language of schooling; in fact, as this excerpt demonstrates, it may be quite the contrary. Here, in this kind of classroom interaction, Ali also builds and narrates his role and position as a (medical) professional. This is indicated by his use of Latin as the professional language of medicine, sometimes used in anatomy instead of Finnish (the Latin term 'dermis' is used even in the Finnish-medium instruction instead of the Finnish term 'verinahka'). Ali also translates the Latin term into English (line 6), which the teacher acknowledges by translating the term from English into Finnish (line 7). As Cenoz and Gorter (2020) have pointed out, this kind of translanguaging promotes both linguistic (multilingual) and content learning; translanguaging does not mean placing less emphasis on the target language (Finnish in this case), but rather, the languages reinforce each other. Makalela (2019, 246) has also shown how shared moments of translanguaging reveal and deepen understanding of the content. In this way, Ali's previously acquired linguistic resources are valued and integrated in building new knowledge.
It is noteworthy that the previous excerpt illustrates the usage of shared multilingual resources between the teacher and the student (western languages, English as a common lingua-franca and Latin as the language of medicine). However, the following examples show that Ali's entire linguistic repertoire, including Somali, which the teachers did not understand or speak, was valued during his studies. As Johnson (2019, 344) has argued, integrating students' own linguistic repertoires into classroom interaction is important because it also indicates to the students that these resources are valued in education and in society. This promotes identity development and agency. In the following excerpt, the student narrates the role of translanguaging in his learning process: In this small story, Ali describes the process of becoming an independent Finnish language user. He positions himself in the beginning of studies as dependent on his Somali-speaking peer ('I had an interpreter who helped me and told me'). Here, he foregrounds the role of translanguaging: he received brokering from his peer, which helped him to survive. His need for support decreased during the academic year we observed, and this change in self-positioning is also indicated in his small story: 'now I don't need any help', 'because I understand everything'. Ali's word choices are quite uncompromising: he positions himself as an independent Finnish language user. This small story also illustrates how teachers position multilingual students: they actively encourage them to translanguage, negotiate and collaborate in building their knowledge of Finnish and the course content. According to Cenoz and Gorter (2020), support of this kind is needed: even if the students spontaneously use their multilingual resources for learning, their whole potential may not be leveraged without active backing from their teachers.
However, although Ali became quite independent in Finnish during the first year of vocational studies, he sometimes struggled to fully understand field-specific content if he had to process it solely in Finnish. After a year of vocational studies, he started asking Maiju about the meanings of some of the Finnish terms he had studied for the exam. The words were related to asthma, which was one of the diseases they had studied for the exam. He said that he remembered these words and phrases by heart even though he didn't understand their meaning. The words he mentioned were 'suussa liukenevalla' (oral soluble), 'pahoinvointi' (nausea), 'sekavuus' (confusion), 'uloshengityksen huippuvirtaus' (peak expiratory flow). He did not, however, give them in their basic form but inflected, which indicates that they are being repeated directly from a continuous text and have not been fully processed by the student. Ali had passed the exam by memorising these medical terms in Finnish without fully understanding the phenomena they refer to. In narrating this memorising process and asking for help in understanding, he positioned himself not only as a learner of Finnish but also as a motivated student who wanted to deepen his surface-level knowledge of asthma. In this case, he might have benefited from using his multilingual resources to search for information about the disease and thereby gain a better idea of how to apply this knowledge in his field.

Omid: becoming a multilingual health professional by having a brokering role in different communities
Omid is originally from Afghanistan. He has an interrupted educational path: he went to school in Iran where he lived as a child but moved back to Afghanistan as a teenager and thereafter only managed to study in private schools. In 2014, aged 17 he came to Finland to live with his brother who had already moved to Finland earlier. Omid resumed his educational path by enrolling in 9 months of integration training for adult immigrants that included L2 Finnish courses and a 5-week work practice period in a youth club. In 2017, he started high school, but discontinued after 1,5 years partly due to inadequate Finnish language skills and lack of support for these, and partly due to culture shock that unsettled his identity and religious beliefs and diverted a lot of his energy and resources away from studying: 'on some occasions, I was confused between what's right and what's wrong. I couldn't decide on what side I have to stay on. I was not even sure that I am in the middle of right and wrong.' Despite his interrupted educational path and severe cultural shock, Omid's orientation to language learning was strong. In addition to Finnish, his language repertoire includes Dari (Persian), Pashto, Arabic and English. During the data gathering process, he made observations on his language architecture, such as his language proficiency levels, by comparing his Finnish and Arabic skills and the different modes of language use. He also reflected on the history and development of Persian, and the domains of language use, thereby indicating a strong metalinguistic awareness, which is evident for example in the following fieldnote extract: 'He says that his Finnish and Arabic skills are equally strong, but his oral skills in Finnish are weaker. He begins to tell me about Persian language: it is a widely spoken and old languageit had a writing system already 2000 years ago. A lot of loan words. But not as much vocabulary for technology, for instance, as in Finnish. ' Omid's orientation to language learning is illustrated in the following extract in which he tells Amin that he was almost excluded from a Finnish language course that was part of the integration training programme because of categorisation based on his educational and ethnic background. Visiting the social office, I was informed that a Finnish course was to begin soon but because the students were graduates from other European countries I might not be proficient enough to take the course. However, I insisted on participating in that course. The students were university graduates, and the teachers taught differently, focusing much more on grammar. If I personally want to learn a language, I focus more on its grammar at first. In the other Finnish courses, the main focus was on speaking whereas in this course the focus was on written grammar. It was a nice course.
In this small story, Omid describes how he was positioned as less capable than the 'European graduates'. The story illustrates not only how his prior knowledge was devalued but also the prevalence of raciolinguistic ideologies in the formal education system, manifested in this instance by the assumption that a student from outside Europe who has an interrupted educational path may not be able to orientate linguistically or make rapid progress in learning. While this might be a relevant issue, it also demonstrates that categorising learners without looking for any further evidence is problematic. Recent studies applying a raciolinguistic approach have shown how racialized student's linguistic competence and literacy skills are often overlooked by the 'white gaze', and how these discourses have been naturalised both at the individual and institutional level (e.g. Flores and Rosa 2015). Omid narrates how he actively resisted being positioned in this way by this stakeholder: he 'insisted on participation', as he had a strong motivation to learn Finnish and especially through its grammar. Thus, Omid was falsely positioned and had to re-position himself. Smith (2019) argues that this kind of struggling can affect individuals' identity and cause them to sacrifice part of their personhood, such as their ways of speaking and thinking.
Language and ways of thinking are also tightly intertwined in Omid's story, as illustrated in the following extract: This narrative may be interpreted as evidence of a monolingual ideology and the dominant role of Finnish language. Rather than enhancing multilingual awareness, multifaceted points of view, and translanguaging, the students were encouraged by the teacher to 'think like a Finn'. This story also reveals how learning a new language can be a holistic process that affects the learner's identity and world view. In spite of these obstacles, Omid learned Finnish quickly. He noted in the interview that after seven months of studying Finnish he was able to conduct official tasks in Finnish easily and not only no longer needed an interpreter but had also taken on a brokering role (see Kraft 2020;Orellana and García 2014) in different communities. For example, he had worked as a Finnish-Persian language broker in a church for 2 years. Further, he had helped Persian speakers to understand and complete bureaucratic tasks in both his internship and studies. Thus, his language architecture now enabled him to cross borders and participate actively in a broad spectrum of life areas. His vocational education group contains several Persian speakers, who are encouraged by their teachers to translanguage for learning purposes, as shown in the next extract. In group work, we speak in Persian although we are trying to use Finnish more instead. However, if the topic is difficult, we explain it in Persian to each other. In these cases, the teachers understand us and express their satisfaction with our cooperation. They have told us that we are happy that you are helping each other.
In this small story, Omid reflects on how the possibility to use Persian with others helps them to understand difficult field-specific content. For instance, in a class on the safety of medical products, a Persian speaker asked the teacher to clarify some terms she hadn't understood ('dokumentointi' documentation, 'lääkevalmiste' pharmaceutical product), and after the teacher had explained them in Finnish, Omid translated them into Persian. Some of the teachers even offered course materials in Persian. The anatomy teacher told us that one of his students had found a course book in Persian. This enabled the teacher to take photocopies of the most important pages and hand them out to the Persian-speaking students. Based on our observations, Omid used translanguaging practices regularly at school and processed the content multilingually. According to Meyer et al. (2015), using multiple meaning making resources deepens one's understanding of the content and also develops language repertoire (see also Meyer et al. 2015). Yet, as example 6 illustrates, Finnish as the official language of instruction remained the target of learning: 'we are trying to use Finnish more instead'. The teachers legitimised its dominant role while also allowing and supporting translanguaging for content learning -a tension (parallel monolingualism and translanguaging) also analyzed by Johnson (2019, 343).
However, some of the teachers not merely allowed brokering practices but actively designed collaborative activities that built on the students' existing knowledge and promoted translanguaging. The following fieldnote was written during the first lesson of a course titled 'Work-related interaction'. In the learning café activity described in this example, students are asked to pick a partner and write down (in Finnish) what they understand to be the learning objectives of the course. The teacher had copied the objectives from the curriculum and placed different objectives on different desks around the classroom. The idea is that the students then move from one desk to another in pairs and on post-it notes write down how these different learning goals pertaining to different aspects of work-related interaction can be achieved in practice.
Opiskelijat saavat itse valita parit. Osin parit menevät kieliryhmittäin, koska yhteinen kieli on seuraavilla: Alina + Eva (venäjä) Azara + Javed (persia) Ava + Omid (dari) Darius + Layla (dari) *** Next in the program is a learning café. The teacher has printed out the learning objectives of the work-related interaction course, and the task is in pairs to write down on post-it notes how you can implement these learning objectives in practice. A new student is very keen to ask. One of the questions is if you have to write on the note what you think yourself or search online. The teacher says that we are not using the internet now but thinking about how we understand it ourselves.
The students choose a partner. Some pairs are formed on the basis of a shared language: Alina + Eva (Russian) Azara + Javed (Persian) Ava + Omid (Dari) Darius + Layla (Dari) The activity supports the students both in using their existing interactional skills and in identifying their needs for further development. The assignment entails negotiation and writing in one's own words, as the students are not allowed to search the Internet for help in wording their ideas. This gives the teacher reliable evidence on the students' level of understanding and compels the students to consider the learning objectives personally and practically rather than exclusively on an abstract level. As the students work in pairs, they spontaneously organise the social space so that they can work with someone with whom they share the same first language, thereby enabling them to use and build their linguistic architecture in a versatile manner. Omid and his partner use their first language, Dari, during this activity.
Omid was also able to use his multilingual resources in his practical training in elderly care. He had been working for several weeks when the new summer employees arrived. The permanent staff were busy guiding them, and since Omid shared the same Persian language resources with a few of them, he was able to help the permanent employees in this task. Below, his supervisor is telling his teacher about this in the evaluation discussion: Example 8: Evaluation discussion (originally in Finnish, English translation below) ohjaaja: mut sit se oli hyvä toisaalta niinku näki siinä sen Omidin että se niinku oli vähän niinku mun työkaveri periaatteessa siinä kun oli kuitenkin kokemusta jo tästä. niin se pysty myös sit niinku sanoo jollekin että sinä voit mennä vaikka tolle ja ne puhu sitä persian kieltä -opettaja: niin. joo ohjaaja: niin sitte hän pysty niinku ohjeistaa myös niitä ilman että mun tarvi välttis sanoo. että se oli ihan hyvä opettaja: no sehän oli ihan hyvä. *** supervisor: but on the other hand it was good because I was able to see Omid like he was kind of my coworker then as he already had experience of this [work] so he was able to say to someone that you can go there and they talked in Persian -teacher: yes. right supervisor: so then he was able to give them instructions so I didn't need to tell them. so it was quite good teacher: well it was quite good.
In this small story, the supervisor tells the teacher how Omid was able to take responsibility for guiding newcomers due to his Persian language skills, and thus was almost like her 'co-worker'.
This indicates that in this brokering situation, the student was positioned as an equal employee in a workplace where he was officially practicing the vocational skills needed in the job. In this small story, his multilingual practices were acknowledged and appreciated by the care centre supervisor ('it was quite good'), whose words were echoed by the teacher. Although Omid was not present during the evaluation discussion, this situation was potentially empowering for him, as it showed he was able to use what he had learned during his practical training and guide newcomers.

Discussion
In this study, we examined the translanguaging trajectories of two students in the process of becoming multilingual professionals in their own field. Their backgrounds, like life histories, educational paths, and language architectures, however, differed. Ali had participated in higher education, namely medical studies, but had not completed these before arriving in Finland as an adult; Omid had a more fragmented schooling path and had arrived in Finland as a teenager. Both were already multilingual on arrival in Finland, and both strategically leveraged their multilingual resources during their education.
To some extent, multilingualism in Finnish vocational education continues to be unconsidered: multilingualism is not mentioned in the aims of curricula unlike in other educational levels in Finland and, based on previous research, the pedagogical practices may be informed by monolingual ideologies (Mustonen 2021;Mustonen and Puranen 2021). The present results show that translanguaging did not hinder the two participants' Finnish language learning; on the contrary, it seemed to speed up the process of becoming an independent user of Finnish. Ali purposefully used his multilingual resources throughout his studies while Omid took on a language brokering role both in his student group and in various other life areas, such as in his church and when assisting his fellow countrymen in their dealings with different authorities.
Our results indicate that in vocational education, both participants were positioned as multilingual professionals by their teachers, some of whom acknowledged their previous content knowledge and multilingual resources and purposefully supported their value for learning. Importantly, translanguaging strengthened not only the participants' Finnish language proficiency but also their multilingualism: both were encouraged to strengthen and develop their language architectures as a whole. However, even if some teachers allowed and supported translanguaging, it was only occasionally designed as a pedagogical practice to promote participation, agency, and learning. Translanguaging was mainly implemented in providing them with assignments that entailed negotiation and thus enabled them to draw on and further build their multilingual resources spontaneously or, less frequently, by selecting teaching materials in the students' first language.
Further, the results indicate that processing content multilingually deepens the understanding of field-specific content (see also Meyer et al. 2015). Promoting translanguaging supported students' participation in meaning-making processes and positioned them as agents of their own learning. Thus, further understanding of how to support translanguaging strategically and systematically in learning processes is needed.
The value attributed to translanguaging is also impacted by hierarchical structures. As Meier (2017) points out, attitudes, such as the monolingual myth about learning -the idea that it is enough to learn the majority language -may still be guiding pedagogical practices. Furthermore, ideologically, economically, and politically informed language hierarchies, intertwined with other intersecting factors, such as ethnicity, can still be found in the formal education context. These power structures have an impact on whose language architectures and which parts of it are valued, as well as what kinds of assumptions are made on the basis of a student's background. Students from non-western countries may be perceived as deficient language learners, and their academic skills may be underestimated, as in Omid's case. Translanguaging as a pedagogical practice and language architecture as a way of framing language skills entails dismantling the power structures that make categorical assumptions about for example a Somali-speaking student (see Kurki 2018) or a student with an interrupted schooling path who is seeking asylum (see Mustonen 2021). The ideal would be to value the language architecture of all students irrespective of their background. Positioning multilingual speakers as equal participants and skillful learners supports their identities, skills, and goals for the future.
This study has its limitations. First, our position as western, white researchers has evidently affected our observations and analysis: most probably, there are translanguaging practices and aspects of multilingual identity we have not been able to recognise. Second, the data were multilingual, which is both an advantage and a challenge. Interviews conducted in Persian were translated into English, and only the translated data were analyzed. Thus, fine-grained analysis of word choices or tones of voice was not possible. Third, the analysis of translanguaging in this study included naming languages and attributing different roles to them, thereby reinforcing their separateness rather than seeing translanguaging practices as the use of joint communicative repertoires (see Johnson 2019; Gynne and Bagga-Gupta 2015). However, making multilingual students' divergent trajectories and language architectures visible, even by naming the languages in question, may be a first step in dismantling the traditional hierarchies between different linguistic resources. The language architecture framework highlights how integral students' language practices are to the development of their professional identities (Flores 2020). In today's society, and in the present two students' field of practice, multilingual skills are highly needed.
The main theoretical contribution of this article was in applying the concepts 'translanguaging trajectory' and 'language architecture' in a study of multilingual students' experiences over time and across spaces. The longitudinal approach taken enabled us to show how the two participants became multilingual professionals during their first few years in Finland. Further, the vocational education context remains under-researched in the field of applied linguistics despite the increasing numbers of migrant students entering vocational education. This study contributed to this research gap by exploring the divergent trajectories of two multilingual individuals. Further studies, for example action research and interventions, are needed to locate and develop translanguaging practices in vocational education.