Jeg Gotta Like Spille Fortnite, Men I Never Win the Game: Implementing Multilingual Pedagogies in a Norwegian Primary School

Abstract

Teachers in Norway have been increasingly faced with the challenge of adapting their instruction methods to address the needs of minority-language students. The current body of research on the issue seems to indicate that multilingual practices are being introduced in Norwegian classrooms. However, they often rely on majority languages, such as English and Norwegian. Some teachers have been found to employ minority languages to support learners’ English writing in drafts. As a result, minority languages in Norwegian schools tend to be regarded as less valuable than Norwegian and English. However, more recent projects are being implemented in Norwegian schools to help teachers alter their ideologies of minority languages. This article adds to this body of research by presenting two teachers’ work with multilingual pedagogies, involving the active use of minority languages alongside Norwegian and English in student texts. The data were collected from: teacher reports, student materials, and mentorship meetings. The findings indicate that the teachers successfully implemented multilingual pedagogies by using language portraits, parallel translanguaging in multilingual posters and multimodal dictionaries, and complementary translanguaging in multilingual poetry. These multilingual practices enabled the students to showcase their linguistic identities and multilingual literacy practices. The implementation of multilingual pedagogies benefited from the long-term availability of scholarly input and guidance for teachers and the opportunity to share experiences in a professional network.

1. Introduction

According to Statistics Norway (2022), immigrants and Norwegians born to immigrant parents currently represent 18.9% of the Norwegian population. This contributes to a culturally and linguistically diverse society, where educators are challenged to adapt their teaching approaches to accommodate the needs of students who use two or more languages in everyday communication on a regular basis, and are thereby considered multilingual (cf. Franceschini 2009, pp. 33–34). This article investigates the implementation of multilingual practices to address the needs of multilingual students in mainstream Norwegian primary school classrooms. More specifically, it discusses how two primary school teachers have implemented multilingual practices and presents their reported experiences and needs throughout the implementation process.

The rights of multilingual students are stated in governmental documents such as the Education Act (Norwegian Directorate for Education and Training 1998) and, more recently, in the revised national curricula for both English and Norwegian (Norwegian Directorate for Education and Training 2019a, 2019b). More specifically, the Education Act states the following in Sections 2–8:

Pupils attending the primary and lower secondary school who have a mother tongue other than Norwegian or Sami have the right to adapted education in Norwegian until they are sufficiently proficient in Norwegian to follow the normal instruction of the school. If necessary, such pupils are also entitled to mother tongue instruction, bilingual subject teaching, or both. The mother tongue instruction may be provided at a school other than that normally attended by the pupil. When mother tongue instruction and bilingual subject teaching cannot be provided by suitable teaching staff, the municipality shall, as far as possible, provide for other instruction adapted to the pupils’ abilities.

In other words, students have the right to benefit from differentiated instruction in Norwegian until they are sufficiently proficient to follow ordinary education in Norwegian (Hilt 2016, pp. 13–14). Furthermore, mother tongue instruction and bilingual subject teaching would be provided “if necessary”, but the Education Act does not clarify what this condition entails (Hilt 2016, pp. 13–14). However, the more general phrase “other instruction adapted to pupils’ abilities” implicitly suggests the possibility of using a wider variety of linguistic resources, including mother tongues and home languages besides the target language of the classroom or school. In this article, a home language is understood as any language employed in the family domain, for various activities and purposes, and may thus include the concept mother tongue, which refers to the language(s) children learn from their parents. The concept majority language is employed to refer to languages that have an official status in a certain geographical area, for example, Norwegian in Norway.

In practice, this mandate for equal educational rights has been implemented in various ways, for example, by creating introductory schools, which aim to prepare students for mainstream schools (see Burner and Carlsen 2019) or introductory classes within mainstream schools (see Beiler 2019). In other cases, under the særskilt norsk opplæring ‘differentiated instruction in Norwegian’ initiative (henceforth SNO), individual teachers are employed to assist newly arriving students, either during lessons in class, or in pull-out sessions occurring in parallel with regular lessons. While these are important measures aimed at the educational well-being of minority-background students, they may be limited in their scope, since they do not address the complex educational needs of all multilingual students, some of whom may in fact be fluent and literate in Norwegian, irrespective of their home languages.

1.1. Theoretical Background

Multilingual students can benefit from the opportunity to freely draw on their linguistic repertoires to activate prior knowledge, express their identity, and meet their communicative needs in an academic setting (cf. García and Wei 2014, p. 22). First, multilingual students build their multilingual competence and literacy skills on previously acquired languages, as literacy abilities are transferable across languages, especially those sharing the same script (cf. Cenoz 2003). It may be argued that even where languages in a linguistic repertoire use different scripts, there is a certain degree of interaction and overlap between them in the brain, which is referred to as common underlying proficiency (Cummins 2000). Second, multilingual students construct their linguistic identities by reflecting on and using their entire range of linguistic repertoires (Cummins et al. 2005; García and Wei 2014; Beiler 2019). Third, multilinguals who have three or more languages often have an increased level of metalinguistic awareness (Cenoz 2003), which means that they can focus explicitly on linguistic forms. This ability, coupled with the comparison of linguistic forms across several languages, constitutes a learning strategy, which may contribute to further language acquisition (cf. Burner and Carlsen 2019). The use of this strategy is actually one of the competence aims in the new Norwegian National Curriculum for English (Norwegian Directorate for Education and Training 2019a), which implicitly supports the use of multiple languages for developing metalinguistic awareness. A discussion of different linguistic forms in the classroom might contribute to co-learning (Wei 2014), meaning that teachers allow themselves to be educated by their students, thus facilitating the negotiation of power relations in the classroom. García and Flores (2012, p. 238) advocated for the use of multilingual pedagogies, defined as organized and sustainable teaching practices based on translanguaging, which is here understood as a communicative practice whereby multilinguals employ a wide variety of linguistic resources without aiming to keep them separate (cf. García and Wei 2014, pp. 22–23). In this paper, the use of the term translanguaging is separated from the conceptualization of García and Otheguy (2014), who view the concept language as a political construct, and thus claim that a linguistic repertoire is a mass of linguistic features, which cannot be grouped into distinct languages. More specifically, it is argued here that translanguaging may be successfully applied in contexts of multiple language use, while still supporting the idea that the linguistic repertoires of multilingual students consist of different languages (cf. MacSwan 2017). The possibility of identifying different languages thus co-exists with the reality of using multiple languages in the same act of communication without keeping them separate. At the same time, recognizing distinctions between languages as systems does not question the existence of a common underlying proficiency (Cummins 2000) in the linguistic repertoires of multilingual learners. Hence, translanguaging is here understood similarly to pedagogical translanguaging, which scholars have recently employed to refer to multilingual practices in the classroom (see Cenoz and Gorter 2020; Cenoz et al. 2021; Prilutskaya 2021).

In certain ways, translanguaging is very useful as an umbrella term, covering a wide variety of multilingual phenomena, including code-switching and borrowing. Code-switching is referred to as the alternation between two or more languages in one sentence or between sentences (cf. Poplack 1980; Myers-Scotton 1993) and is often governed by linguistic patterns, such as the equivalence constraint (Poplack 1980), which stipulates that switching between languages typically occurs without violating the syntactic rules of either of them. In some cases, translanguaging is especially convenient as a higher-order term where it would be challenging to argue for the exact classification of a specific instance of multiple language use. For example, in the Norwegian sentence Det var nais ‘That was nice’, the English adjective nice has been adapted to the Norwegian spelling conventions, but it is debatable whether this should be classified as an instance of code-switching or borrowing. Such language use is typical in the everyday communication of multilinguals (cf. García and Wei 2014). In the classroom, this would translate into the use of two or more languages within one lesson, or even within one learning activity (Cenoz and Gorter 2011, p. 357).

As MacSwan (2017) points out, studies of multiple language use, including multilingual pedagogies, would benefit from incorporating conclusions from research on code-switching, which indicates that multiple language use is governed by norms and should, therefore, not be equated with lacking linguistic proficiency (MacSwan 2017, p. 169). More specifically, there are two main theoretical frameworks under the code-switching paradigm that may be useful for developing a multilingual pedagogy. First, code-switching has been classified as parallel and complementary (Sebba 2012), where parallelism refers to the use of different languages for the same content, while complementarity refers to the use of different languages for different content. Parallelism is, in fact, the translanguaging model for identity texts, where students present the same content in two texts—one in their home language, and the other in the target language (Cummins et al. 2005). By using two different languages for the same content, students are allowed to activate their previous knowledge, reflect on their identities as multilingual learners, and invest their identities in language learning (cf. Cummins et al. 2005; Krulatz et al. 2018). Second, code-switching has six communicative functions: referential, directive, expressive, phatic, metalinguistic, and poetic (Appel and Muysken 2005, pp. 118–19). Appel and Muysken (2005, pp. 118–19) explain the six functions as follows: (1) the referential function implies the use of another language if speakers do not remember a word in the language they were initially using, thus bridging a communication gap; (2) the directive function refers to the ability to both exclude and include conversation partners by alternating between languages; (3) the expressive function concerns the use of multiple linguistic resources to construct and present the identity of the speakers; (4) the phatic function has to do with using a certain linguistic resource for an enhanced effect, for example, providing the punch line of a joke in a different language for a more humorous effect; (5) the metalinguistic function implies the use of multilingual practices to attract admiration and is mainly employed by persons whose professional skills involve multilingual competence; (6) the poetic function refers to drawing on multilingual resources to create a rhetorical effect in literary texts. These functions may be applied within multilingual pedagogies to inform multilingual activities and prompt discussions about language use. To be more specific, many of these functions may be combined in various multilingual activities to incorporate the use of both home languages and majority languages alongside English.

1.2. Implementing Multilingual Pedagogies

Implementations of multilingual pedagogies in education reflect the softening of boundaries between languages in communication (Cenoz and Gorter 2013), meaning that multiple languages may successfully be used in one given setting or domain, without attempting to keep them separate. This is set in opposition to one domain-one language patterns of language use, which are often found in multilingual societies (cf. Fishman 1972, p. 144). Although the one domain-one language pattern may work well in some settings, applying it in the educational domain in Norway would situate Norwegian as the sole or main educational language, thus placing home languages outside of the educational sphere. This would, in turn, clearly contradict the principle of social justice, as referred to in García and Flores (2012, p. 242), whereby educators promote an inclusive and respectful attitude towards all languages and their speakers.

The new national curricula for Norwegian and English advocate for the implementation of multilingual pedagogies. First, the national curriculum presents multilingual competence as a valuable resource both at school and in society in general, irrespective of the languages involved, and encourages teachers and school leaders to adopt the same view (Norwegian Directorate for Education and Training 2019a, 2019b). Second, both curricula support this fundamental value with concrete competence aims, which require the implementation of multilingual teaching approaches. For example, after Year 4, or by the age of 10 to 11 years, students should have the ability to compare words and expressions in Norwegian to other languages. In English, they should discover and play with words that are the same or similar in English and other languages familiar to them. This in an important change from the previous curriculum, which limited the comparisons to English and the students’ native languages. The new Norwegian curriculum also includes another relevant aim, which refers to the students’ ability to explore and discuss linguistic variety and diversity in their social environments. In other words, teachers are expected to foster and guide active classroom work with a variety of languages in Norwegian and English lessons alike, or even in a cross-curricular manner, while also encouraging discussions about linguistic diversity, which can contribute to an increased level of metalinguistic awareness.

Multilingual pedagogies have both a stance and a design component (cf. García et al. 2017). This means that to embark on developing such pedagogies, teachers must first take a positive stance towards the use of multiple languages by allowing students to use them in both communicative and learning situations at school. Nevertheless, this positive stance on its own is not sufficient, so it must be complemented by the design element, whereby teachers purposefully plan and implement multilingual approaches at all stages of learning processes, based on the students’ needs and linguistic practices.

Such a multilingual teaching approach might include, for example, the use of language portraits, where the students showcase their linguistic repertoires in drawings. More specifically, they use different colors to represent their different languages on a body-like figure, and thereafter explain their representations (Busch 2012). This activity may contribute to the development of learners’ metalinguistic awareness, as they consciously reflect on when and how they use their different languages. A useful aspect of this process might be the potential to uncover patterns of domain-based distributions of languages in students’ experiences.

Teachers may also ask students to make comparisons between different languages to raise their metalinguistic awareness (cf. García and Flores 2012; also see Section 1.1 and Section 1.3). This may include discussions of the pragmatic functions of multilingual practices, as language choices in a text may serve to guide readers (see Section 3.1.3). Further, multilingual materials produced by students may be used to create schoolscapes, which Gorter (2017) defines as the linguistic landscapes of educational spaces such as schools. In other words, schoolscapes represent the totality of signs made publicly visible in a school and thus have the potential to make the students’ language repertoires visible in the classroom and other educational spaces (cf. Gorter 2017). This validates the importance of all represented languages and empowers their speakers by indicating that their identities are accepted and valued in a particular educational space.

1.3. Multilingual Practices in Norwegian Schools

As multilingual practices have recently started to be employed in Norwegian classrooms (see Beiler 2019; Krulatz and Iversen 2019), scholars have identified several challenging aspects. To begin, Iversen (2017) pointed out that in certain classrooms, some students used their minority languages as learning strategies (see Burner and Carlsen 2019) to enhance their learning process. This is undoubtedly beneficial, but the problematic aspect is that their teachers seemed to take on a rather passive role, as they merely encouraged multilingual practices if the students used them, but did not intentionally foster them. In general, teachers may shy away from implementing new teaching methods if they feel that they do not have the necessary training for such endeavors (cf. Šurkalović 2014), or they might be reluctant to challenge the status of the majority language in the classroom, even if they have the skills to do so (cf. Iversen 2017).

Burner and Carlsen (2019) presented similar findings in their study. Although, the teachers in their study had appropriate qualifications and reported positive attitudes towards multilingual practices, they did not employ home languages systematically in instruction. More specifically, the teachers sporadically asked the students for words in their home languages, especially in Norwegian lessons, but were reluctant to establish a consistent practice due to their perceived lack of time (Burner and Carlsen 2019, p. 11). In fact, the teachers admitted to intentionally prioritizing Norwegian, even in English lessons, to prepare the students for mainstream schools (Burner and Carlsen 2019, pp. 8–10). In terms of visibility, the researchers indicated that home languages did not seem to be part of the schoolscape, as only English and Norwegian were represented on the classroom walls, on two separate posters (Burner and Carlsen 2019, p. 11). Finally, the teachers expressed concern about the fact that students may use their shared home languages to form sub-groups, which would compete with school structures (Burner and Carlsen 2019, pp. 11–12).

In another important study, Beiler (2019) unveiled how multilingual learners make use of their multilingual repertoires for draft writing in two introductory English classes. In this study, the teachers explicitly encouraged the students to use their home languages and other languages they had acquired before their arrival in Norway. The students typically translated and alternated between languages to support their drafts and thereafter wrote the final piece in English. Interestingly, they seldom used their home languages, and seemed to prefer other languages. There seemed to be a strong contrast between their rich language portraits, and the extent to which they used their represented languages in draft writing. One student explained that she rendered her literacy skills in her home language as insufficient for school texts, even if she used it in informal conversations on social media (Beiler 2019, p. 21). According to Beiler (2019), teachers should legitimize students’ out-of-school literacy practices as useful resources for the development of in-school literacy. Furthermore, Beiler (2019) also suggested that translanguaging should be used as a rhetorical mechanism in finalized texts, which corresponds to Appel and Muysken’s poetic function of code-switching (Appel and Muysken 2005, p. 119).

Krulatz et al. (2018) showed that minority languages may remain invisible even when teachers actively engage with multilingual approaches in language teaching. The researchers conducted a project within the Kompetanse for Mangfold ‘Competence for Diversity’ initiative, which consisted of a one-year collaboration with two rural schools from Mid-Norway. The project focused on the use of identity texts (cf. Cummins et al. 2005) to increase metalinguistic awareness and respect for cultural diversity among students and teachers. Although this aim was achieved, the researchers reported that the identity texts produced by students only showed the use of English and Norwegian, with no visible presence of minority languages (Krulatz et al. 2018, p. 566), which reproduced domain-based patterns of language distribution in the educational domain, where Norwegian and English are typically taught, used and, thus, recognized as the only languages of the school.

1.4. Relevance and Aims of the Present Study

This paper contributes to the growing body of research on multilingual practices in Norwegian schools by presenting results that complement some of the main findings of previous research in the field, namely: teachers’ passive role concerning home language use (Iversen 2017), or their reluctance to adopt a systematic approach for including home languages in the classroom (Burner and Carlsen 2019), and the general lack of home language use in student texts (Krulatz et al. 2018), especially finalized written texts (Beiler 2019). In contrast to previous work, this study discusses systematic approaches of teachers who actively encourage and support home language use both in learning activities related to, for example, vocabulary work, as well as in final pieces of creative writing, such as multilingual poetry.

The research questions addressed are as follows:

RQ1: What kind of multilingual practices do teachers employ in their work with multilingual learners in the classroom? What might be the aims and outcomes of such multilingual practices?

RQ2: Do the practices teachers employ encourage students to use their home languages actively? If so, how are home languages represented in student texts?

RQ3: What are teachers’ reported experiences and needs at the different stages of implementing multilingual pedagogies?

These questions are answered based on various types of data collected during a larger project within the Competence for Diversity initiative, funded by the Norwegian Directorate for Education and Training (see Section 2). Section 3 presents the findings in detail, and Section 4 discusses them in connection with relevant theoretical frameworks and previous studies. Finally, Section 5 presents the conclusions and identifies several challenges and main factors that may influence the implementation of sustainable multilingual pedagogies in Norwegian mainstream classrooms.

2. Materials and Methods

This article is based on data collected during a professional development project involving primary school teachers of English and Norwegian, which aimed to equip teachers to perform in linguistically and culturally diverse classrooms through collaborations between schools and universities. This professional development project was conducted in a similar way to that presented by Krulatz et al. (2018). As Krulatz et al. (2018) point out, the local implementation of Competence for Diversity programs seems to vary across the country, as they are based on the requests from local schools. Typically, schools carry out an assessment of their needs, which the partner university uses to create a plan and eventually provide tailored training for teachers (Krulatz et al. 2018). As a result of such collaborations, teachers benefit from scholarly input and mentorship and often create new professional networks within or across schools.

2.1. Background of the Study

The project referred to here was conducted in the period 2020–2021 as a collaboration between the Education Agency (henceforth UDE) and a university in Eastern Norway. It started in September 2020 and was completed in April 2021 and initially involved eight schools in an urban area in Eastern Norway, 19 in-service primary and secondary-school teachers, and seven faculty members who acted as lecturers and teacher mentors. This project addressed a rather general need expressed by the school representatives, namely, to equip teachers to work with student groups characterized by linguistic, cultural, and religious diversity. The present author was one of the teacher mentors, and she held one of the lectures and led two mentorship groups in collaboration with a faculty colleague.

The program included four lectures in plenum and three mentorship meetings in small groups, consisting of two to six teachers per group. The attendance rate varied considerably during the project, due to the ongoing COVID-19 pandemic. The lectures were organized as 2 to 2.5 h sessions and were carried out in September and October 2020 and February and April 2021. The mentorship meetings were organized as 1.5 h seminars in alternation with the lectures, and started in October, after the first lecture. The lectures provided theoretical input and practical activities for use in multicultural classrooms on the following topics requested by the school representatives: religion in a multicultural context, multilingualism as a resource, transitions from everyday language to school language, and practical tools for multilingual teaching approaches. It should be noted that the latter topic was discussed in detail during the second and third lectures, as well as continuously in the mentorship meetings led by the present author. The mentorship meetings provided opportunities for teachers to clarify the theoretical input from lectures, and discuss how certain activities could be applied in their classes.

The general aim of the project was to guide and support teachers as they designed and employed new methods for teaching in a multicultural environment. After each lecture, the teachers received assignments to work on and share with the professional community at their respective schools. They then wrote brief reports to present their work, which they sent to the mentors before each mentorship meeting. During the meetings, they shared and discussed their experiences with the mentors and the other teachers in the group. Three main challenges were identified while the project was carried out. First, no mentorship meeting was organized after the final lecture. Second, many of the teachers participated passively. Third, two of the mentorship meetings and two of the lectures were held online due to the COVID-19 pandemic. This was also the reason why many teachers found it challenging to work with the provided assignments and attend all the lectures and mentorship meetings.

2.2. Participants

This article is based on the experiences of two primary school teachers, here referred to as T1 and T2, from a mainstream school in Eastern Norway, which offers education in grades one to seven, and where roughly 50% of the students have an immigrant background. Out of the 11 teachers in the mentorship groups led by the present author, only T1 and T2 participated in all lectures and mentorship meetings, handed in reports, and were actively involved throughout the program. T1, a female teacher, and T2, a male teacher, each had approximately 30 years of teaching experience at the primary level. Both teachers held a graduate degree in teaching, organized as a bachelor’s degree followed by additional courses. T1 taught Norwegian, Mathematics, Social Studies, and KRLE (Christianity, Religion, Philosophies of Life, and Ethics). T2 taught Norwegian, English, Music, Social Studies, Food and Health, and KRLE. Although T1 did not teach English, her work has been included here because it involves the use of English alongside Norwegian and minority languages.

During this project, T1 and T2 taught in primary-school classes, where most immigrant students come from Albania, Poland, Ethiopia, Pakistan, Denmark, Turkey, Somalia, and the Philippinesand Korea. The teachers worked on the integration of mother tongues in multilingual education in a cross-curricular manner. This means that the multilingual activities they implemented, except for the work on multilingual dictionaries, were not limited to either English or Norwegian. Instead, the activities were based on the inclusion of the students’ mother tongues alongside Norwegian and English. Although Norwegian seemed to be the teachers’ preferred language of communication and starting point in most activities, English proved to be a steadfast companion in all activities, in various ways.

2.3. Data Collection Methods

The methodology employed in this project was based on linguistic ethnography, an approach that aims to study the participants’ perspectives in the wider socio-cultural context where they occur (Copland and Creese 2015). An ethnographic perspective allows for qualitative introspection into the experiences of participants (cf. Heath and Street 2008), including the investigation of language practices in context over time (Heller 2008, p. 250). An important aspect of using an ethnographical methodology is that it values the combination of various data, such as interviews, written reports, and student materials, as a balanced approach that contributes to the overall validity of the research enterprise (cf. Copland and Creese 2015).

This study employed four main types of linguistic ethnographic data collected from:

-

21 student texts

-

Four teacher reports

-

One informal group interview during the first mentorship meeting

-

One formal semi-structured focus-group interview during the last mentorship meeting.

The 21 digital copies of student texts consisted of seven language portraits, four vocabulary charts, two multilingual dictionaries, and eight multilingual poems. The language portraits and multilingual poems were produced by 10 students. Seven of them also participated in recordings made by T2, where they presented their portraits or read their poems out loud. These recordings were transcribed and used to confirm the students’ linguistic repertoires. To ensure their anonymity, the students are referred to as Sn (n = 1–10) and the plural form of the third person, they, is used for both singular and plural references. In this study, biological gender is not a salient variable. The present author received written consent from the students’ parents or caregivers to use the student texts and recordings for research purposes. To ensure data protection, the files were stored on a password-protected computer to which only the present author had access. The project was approved by the Norwegian Centre for Research Data (henceforth NSD), with registration number 442887.

Furthermore, the multilingual dictionaries and poems were transcribed, and the language portraits and vocabulary charts were described based on their visual content together with the student recordings and extra-textual information provided by teachers in their reports, or orally during the mentorship meetings. The student texts were first grouped based on their types and shared characteristics, and then analyzed based on their linguistic composition by integrating MacSwan’s (2017) understanding of the existence of separate languages that may nevertheless share certain linguistic features, such as cognate words shared by North Germanic languages (see Section 3.1.3). To be more specific, the languages employed in the texts were identified, and their presence was interpretated and discussed by means of theoretical frameworks involving the functions of code-switching (see Appel and Muysken 2005; Sebba 2012) and applied linguistic perspectives on multilingual practices (see Cummins et al. 2005; Krulatz et al. 2018). This analysis was complemented by teachers’ insights and reflections presented in their reports.

The four teacher reports and digital copies of the 21 student texts were sent to the researcher by T1 and T2 prior to the mentorship meetings. Of the four reports, two were produced individually by each teacher, and the first and last were produced by T1 and T2 in collaboration. These reports provided various types of information, ranging from practical details about the school, such as the percentage of immigrant students, to descriptions of how multilingual practices were implemented and the teachers’ reflections on this process. The reports also contained contextual information about the student texts, such as the resources the students used and their reactions to the implementation of multilingual practices. For this reason, the reports were used as supportive material in the analysis of student texts.

An informal interview was employed to collect data during the first mentorship meeting of the project, which took place in person. This type of interview is an established ethnographic method (Copland and Creese 2015) and was selected for several reasons. First, it allows the participants to explore a wider area or topic (Copland and Creese 2015), where the researcher uses strategic questions rather than a pre-prepared interview guide (Agar 2008, p. 140). As a result, the researcher may avoid taking the formal role of an interviewer, thus allowing the participants to take charge of the discussion. This is particularly useful during the initial phase of a long-term project because it allows for an in-depth exploration of participants’ concerns, as the researcher can address other important questions at a later stage (Gobo 2008, p. 191). During the informal interview, the teachers reflected on their attitudes, knowledge, practices, and needs concerning multilingual practices within the larger frame of teaching approaches for multicultural classrooms. Their reflections were recorded in hand-written notes, which were thereafter processed through inductive coding to identify the most significant reoccurring themes (cf. Saldaña 2011).

During the last mentorship meeting, which took place on Zoom, a formal semi-structured focus-group interview was used to investigate the two teachers’ perspectives and needs after they had implemented multilingual practices in their classrooms. More specifically, an interview guide was prepared based on relevant theory and the themes identified during the coding of the informal interview notes. The teachers were also encouraged to provide input on additional topics related to their experiences with multilingual practices, as the interviewer’s purpose was to perceive the investigated matter from the interviewees’ perspectives (cf. Copland and Creese 2015). At this point, a formal type of interview was considered ideal because the teachers and mentors had already established a relationship based on mutual trust and respect (see Sherman Heyl 2001, p. 369). External factors, such as school announcements about infection control measures and teachers’ limited time availability, might have had an impact on the discussion. This interview was recorded with a physical voice recorder placed at the interviewer’s location to ensure the privacy of the participants. The interview was transcribed, and the data were processed through inductive and abductive coding (cf. Saldaña 2011) by focusing on theory-based themes, as well as themes generated by the participants. The present author received written consent from the other mentor and both teachers to use the recorded interview, reports, verbal contributions from mentorship meetings, and student texts for research purposes. To ensure data protection, the files were stored on a password-protected computer to which only the present author had access. The project was approved by the Norwegian Centre for Research Data (henceforth NSD), with registration number 442887.

The following section addresses the research questions by presenting the various multilingual practices employed by the teachers during the project and their experiences and needs identified at the start and end of the project. The results also make reference to students’ involvement in the process, as well as implications for their families and the professional community of the school, insomuch as it was affected by the teachers’ work with multilingualism.

3. Results

3.1. Multilingual Practices and Materials and Their Value in and beyond the Classroom

3.1.1. Language Portraits—A Springboard for Metalinguistic Awareness and Co-Learning

In this study, the participants’ work with language portraits seems to have had a domino effect by reaching learners, the school’s professional community, and the learners’ families. First, T2 worked with language portraits in his class. Subsequently, T1 and T2 presented this activity to their colleagues, who thereafter also employed them in their work with students from 1st grade to 7th grade. The teachers’ work with this activity had three main outcomes: (1) developing the students’ metalinguistic awareness and validating their linguistic identities, (2) facilitating innovative reflections on the concept language, and (3) establishing co-learning pathways within and beyond the classroom. These aspects are presented in detail below, together with the background of the implementation process.

After the second lecture of the teacher training program, T2 decided to use language portraits with his class to learn more about the languages the students used at home, and what other languages they could use, irrespective of their proficiency level. The teacher’s motivation was inspired by an assignment received during the program, but the use of language portraits for this purpose was his own decision. T2 decided to carry out this lesson in an interdisciplinary setting. More specifically, the students explored and reflected on their linguistic practices including minority languages, so instruction was not formally restricted to either English or Norwegian as overarching subjects, but the teacher used Norwegian as the main language of instruction.

By modelling the activity and allowing the students to define their own understanding of knowing a language, T2 fostered metalinguistic awareness development among his students and validated their linguistic identities in the classroom. First, T2 explained his motivation to the students and then proceeded to model the activity for them by creating his own language portrait. While creating their portraits, some of the students considered that they needed to speak a language fluently to represent it, while others believed one word was enough. Notably, both approaches were approved by the teacher. During the days following this activity, the teacher had informal, individual discussions with the students, where they explained why they chose to represent their languages with a certain color and in a certain place on their drawing of the body-like figure. Some of the students presented their portraits independently, while others needed guiding questions from the teacher. This paved the way to developing their metalinguistic awareness by asking them to reflect on the languages they knew, where and when they used them and with whom, and how they learned them. Afterwards, the portraits were displayed on the classroom wall, thus validating the students’ linguistic identities by making them visible in their social space.

According to T2’s report, the students engaged well with this activity, and believed it was an exciting exercise. Indeed, their engagement is clearly noticeable in some of the portraits, which present complex representations of multiple languages. For example, S5’s representation of languages included seven linguistic varieties: Norwegian (mother tongue), Turkish (mother tongue), Danish, English, Nynorsk (New Norwegian), Swedish, and American English (Amerikansk in original). Interestingly, the reference to Amerikansk here indicates that S5 has an awareness of distinctions between British and American English. Other mother tongues represented in the students’ portraits are Urdu, Amharic, and Polish, which are typically associated with the home or family domain. Unsurprisingly, Norwegian and English appear in all language portraits and seem to be the only languages used in two domains, namely, education and hobbies.

In their final report, T1 and T2 wrote that “they [the students] were motivated, proud, and showed mastery” during the implementation of language portraits at the school level. More specifically, the students appeared proud of their backgrounds, which the teachers considered to be a positive aspect. At the same time, this activity fostered the students’ reflections on their linguistic identities and contributed to creative conceptualizations of the term language among them. To illustrate, 5th-grade students discussed the potential for considering gaming and martial arts languages, and 7th-graders talked about multiethnolects, and more specifically, kebabnorsk, a local ethnolect used in and around Oslo by teenagers and young adults with a multilingual background. Such discussions revealed the students’ advanced abilities for sociolinguistic and metalinguistic reflection, which seemed to be directly proportional with their age.

Another important outcome of using language portraits was the creation of new pathways for co-learning (see Section 1.1) both within and outside the school. The teachers reported that they gained new knowledge about each other and learned more about the students’ backgrounds, linguistic repertoires, and communication practices. At the same time, the students also used this opportunity to get to know each other better. Beyond the classroom setting, 1st graders’ work on language portraits involved the participation of parents, who talked to their children about their own linguistic repertoires. This may be considered a useful step for establishing a systematic long-term collaboration between the school and parents, aimed at forging a common understanding of the students’ linguistic needs and development.

3.1.2. Parallel Translanguaging in Vocabulary Work

The use of multilingual practices for vocabulary work was based on parallel translanguaging, which refers to presenting the same content in two or more languages (see Section 1.1). The findings presented in this section are based on two main types of materials employed by the teachers: multilingual posters and multimodal dictionaries. The main findings indicate that both types of materials successfully showcase the students’ multilingual literacy skills and that the students’ families are involved in the students’ learning processes in various ways. The findings are presented in more detail below, following descriptions of the implementation of multilingual materials in the classroom.

Multilingual Fairytale Posters

The main aim of using multilingual practices in T1’s class was to raise her students’ awareness about the presence of multiple languages in their class while working with fairy-tale vocabulary. She aimed to activate and include the linguistic competence of all students, while allowing them to experience linguistic diversity firsthand. For this purpose, she used multilingual posters with key words based on the fairytale Little Red Riding Hood. This study analyzed four multilingual posters in Urdu, Albanian, Turkish, and English, respectively.

The fairytale was presented to the class orally in a parallel manner, by using Norwegian, English, and Turkish in sequence to convey the same content. Turkish was chosen since it was represented in some of the students’ repertoires, so a Turkish-speaking teacher employed at the school assisted with this. Afterwards, the students were asked to retell the story to their learning partners by using keywords in Norwegian provided by the teacher. This was followed by students activating their multilingual schemata by identifying key concepts in their home languages, such as Urdu, Turkish, Kurdish, Albanian, Korean, and English. T1 then placed the words written by the students on post-it notes into a chart where she provided the keywords in Norwegian in the left-hand side column. Interestingly, Kurdish and Albanian seemed to be missing from the image of the chart provided by T1. The absence of Kurdish may be related to the fact that in her report, Kurdish is presented as an alternative to Turkish.1 It is also possible that these two languages belonged to the repertoires of students who also spoke Turkish, but it is unclear if this was indeed the case.

The use of differentiation in this task played a major role in organizing the work on multilingual posters. In other words, the teacher employed collaborative groupwork organized in homogenous groups based on shared home languages. Additionally, minority-language students created multimodal posters with keywords in their home languages, accompanied by drawings and translations to Norwegian, while students with a Norwegian background created a comic strip based on the fairytale. One of the students, who had Albanian in their repertoire, managed to perform both tasks. After they created the posters, the students presented them to the rest of the class, who repeated the words out loud in each minority language.

A significant finding based on the multilingual posters is that the students’ linguistic identities seemed to be reflected by visually marking certain languages. For example, on the English–Norwegian poster, the English words and drawings were produced in plain gray pencil on a white background, while the Norwegian words were written in black ink on pink cut-outs glued onto the poster. As a result, the Norwegian words immediately caught the viewer’s attention and may mark Norwegian as a more important language for the students who created this poster. Similarly, the Turkish words were larger and took up considerably more space than the Norwegian ones on the Turkish poster. Additionally, this poster is different from the others in that the students wrote the Turkish title Kırmızı Şapka ‘Red Hood’ twice at the top of this poster, without the Norwegian equivalent. Additionally, it appeared that this title is the students’ interpretation since the conventional Turkish title would be Kırmızı Başlıklı Kız. This could indicate that (some of) these students had previously encountered this fairytale in Turkish, possibly in more informal settings. In general, these details seemed to indirectly present Turkish as an especially important language in these students’ repertoires. On the Albanian and Urdu posters, the visual representations of the languages seemed to be more balanced, which could indicate a relatively equal status of the languages employed in the respective posters. More specifically, on the Albanian–Norwegian poster, the words in the two languages were similar in size and written in the same color, starting with the Norwegian words. In the Urdu–Norwegian one, there were no significant differences in size between the languages, but the Urdu words were placed above the Norwegian ones, with the latter being underlined.

The work on multilingual posters allowed the students to showcase their literacy skills in the finalized pieces. For example, the Turkish and Urdu posters employed Turkish diacritics and the Urdu alphabet, respectively, which indicates the existence of advanced literacy skills in Turkish and Urdu. At the same time, T1 became aware of the students’ advanced multilingual literacy skills, as her report stated that the Urdu-speaking children showed a high level of proficiency in their mother tongue, both orally and in writing. It may be noted that the presence of mother tongues in the finalized pieces, such as multilingual posters, indicates that minority languages were accepted and valued as languages of communication in this educational setting.

The students who spoke minority languages, especially Albanian, Turkish, and Urdu, showed a high level of engagement with the task, while the ethnically Norwegian students seemed more passive. The minority-language students shared their experience with their families, which T1 perceived as an indication of their pride and satisfaction with the recognition of their linguistic background in the classroom. Some of the students experienced a certain degree of uncertainty with regard to finding the equivalent words in their home languages, but the students who spoke Turkish, Albanian, and Urdu seemed enthusiastic about working on their posters. They were proud to say the words in their mother tongue out loud in class, and have their classmates repeat after them, which may be a source of co-learning for both peers and teachers alike.

Multimodal Dictionaries

The teachers’ main purpose for using multimodal dictionaries in a cross-curricular manner with both of their classes was to support the students’ understanding and acquisition of idiomatic phrases while fostering their metalinguistic awareness. For this purpose, they created a template for a multilingual dictionary based on the topic Gi gass + Flaggregler ‘Speed up + flag rules’ from the SALTO student book (see Kolbjørnsen Bjerke et al. 2020, pp. 15–17), which they used with their classes. This was based on the theoretical input and assignment they had received during the third lecture of the training program.

The main principle for the template design was based on parallel translanguaging and partly also on the principle of direct language acquisition. In other words, visual representation was the starting point for the first two concepts, as it was provided in the first left-hand column. This was followed by a column dedicated to Norwegian, English, and mother tongue/dialect/synonym in this order, which allowed for the presentation of the same signified concepts by using signifiers in different languages.

Although Norwegian seemed to be prioritized, the presence of English and mother tongue was significant, as it provided a more inclusive representation of the students’ multilingual repertoires. The inclusion of dialect and synonym in addition to mother tongue was based on differentiation aimed at supporting the ethnic Norwegian students in exploring their entire linguistic repertories, which would typically consist of different linguistic varieties and styles, ranging from formal to informal ones.

This template guided the students in creating their own multimodal dictionaries, by drawing the last three images in the first column, filling in the last word in the English column, and providing corresponding words in their mother tongue (see

Table 1. Multilingual dictionary with Turkish.

Tegning (Drawing)	Norsk (Norwegian)	Engelsk (English)	Morsmål/Dialect/Synonym Mother Tongue/Dialect/Synonym Tagalog
Provided by teacher	trekke på seg	dress up	Giydir
Provided by teacher	løpet er kjørt	it is over	Bitti
Drawn by student	Gi gass!	Speed up!	hızlandirmak
Drawn by student	fullføre	finish	Bitiş
Drawn by student	lillebror	(by student: litel brother)	Küçük erkek kardeş

). In their final report, T1 and T2 mentioned that some of the students worked with their parents, while others used online translation tools to accomplish this task.

The teachers were generally impressed with the students’ achievements in working with these concepts. However, the report did not specify whether the students compared between the different languages in the classroom. This is interesting because such a parallel display of different languages would be ideal for identifying similarities and differences between various linguistic forms, and thus facilitating metalinguistic awareness as a learning strategy and competence aim (see Section 1.1 and Section 1.3). For example, comparisons could be drawn between languages within one table, by comparing between English, Norwegian, and the mother tongue, as well as by comparing the different mother tongues across tables.

Although this design was based on parallel translanguaging, it had a complementary dimension, which may have contributed to nuancing and clarifying semantic distinctions between signifiers. For example, the signifier finish may have different meanings as a noun or verb in English, but in this context its meaning was clarified by the presence of the Norwegian verb fullføre ‘complete’, which eliminated any potential confusion with, for example, the noun texture. This knowledge contributed to selecting the most appropriate word in the mother tongue for this specific meaning.

3.1.3. Complementary Translanguaging in Multilingual Poetry

T2 was inspired by translanguaging in poetry during the training program, where teachers were provided with examples and encouraged to write short poems about themselves in more than one language. Although this was not part of an assignment task for the teachers, T2 decided to implement it with his class in a slightly adapted version. In total, eight sample texts were made available for the present study. The poems are referred to as Pn, where n = 1–8. Seven of these contain a combination of Norwegian (see Example 1) and several other languages, such as Urdu, Polish, Swedish, and English. One text is a translation from Norwegian to Amharic (P5), and another text contains English, Danish, and Norwegian (see Example 2). Three samples are discussed in more detail below, with English translations produced by the present author, and original Norwegian texts provided in Appendix A (I).

T2’s work with multilingual poetry in the classroom revealed the employment of multilingual practices relying mostly on complementary translanguaging, which refers to the use of multiple languages to convey different content within the same text (see Section 1.1). In other words, the students were asked to interpret Norwegian rhyming couplets by switching between Norwegian and other languages from their repertoires, leading to code-switching both within (see Examples 1 and 3) and between sentences (see Example 2). They also had other options, for example, to write their own multilingual poems freely, without relating them to Norwegian texts (see Example 3), or to translate the entire poem from Norwegian into another language of their choice, leading to parallel translanguaging (see P5 in Appendix A (II)).

Interestingly, T2’s guidance for the students involved advising them to switch to another language in content rather than function words, which seems to echo the equivalence constraint (see Section 1.1). However, the students themselves chose which words to insert into other languages, and while most switches seemed to occur in content words, function words were also switched; for example, the coordinating conjunction og ‘and’ was switched to the Urdu ar in Example 1.

Så rart-norsk og urdu   Så agib å være flaggermus ar flakse rundt fra gar til gar og jana til sengs i Per Men er det noen som samage Hvordan den kan få sona når Den latekrie etter anglio.	So weird-Norwegian and Urdu   So weird to be a bat and flutter round from house to house and go to sleep in trees But is there anyone who understands How it can sleep when It hangs from its toes.
Example 1. Multilingual poem by S3.

Complementary translanguaging in multilingual poetry seemed to have an expressive function (see Section 1.1), as it allowed the students to express their linguistic repertoires by switching to languages of their choice to represent distinct content in the poems. In Example 1, S3 chose to use their mother tongue, Urdu, together with Norwegian. This poem could also be used for drawing comparisons between languages, for example, between the linguistic forms of the verb to sleep in English, Urdu, and Norwegian. In Example 2, S1 used Norwegian, their mother tongue, together with Danish and English, but not Nynorsk, which is also listed in their language portrait. This may indicate that these three languages play the most important roles in S1’s repertoire.

The visual marking of languages seemed to fulfil several different purposes in the five texts that show this feature. For example, the text in Example 1 originally contained Urdu words written in blue, which made them stand out, with the pragmatic function of helping the reader to identify more quickly which Norwegian words had been replaced by Urdu. In contrast, in Example 2, all languages were visually marked in the original text, with red for Norwegian (here underlined), orange for English (here italicized) and yellow for Danish (here in bold). On the one hand, this clearly indicates that S1 could distinguish between the languages in their repertoire. On the other hand, this clear separation might have the function of visually signaling to the reader the presence of three languages in one text, even before they start reading it. This kind of representation, however, would have benefited from further discussion in the classroom, since the linguistic boundaries between these three languages are not always easily distinguished. For example, Danish and Norwegian as North Germanic languages have significant areas of overlap, so i sin egen krop ‘in its own body’ might have been classified as Norwegian if the word for body had been written kropp. Similarly, words such as so and can are mutually intelligible in all three languages involved, which might be addressed by further work on cognates shared by, for example, Germanic languages.

Så rart–norsk, engelsk, dansk   Så rart å be spider with yarn i sin egen krop og spinne alle dage. But how can det skjule på så mange kilometer tråd in such one lille mave?	So weird–Norwegian, English, and Danish   So weird to be a spider With yarn in its own body and spin all day. But how can it hide So many miles of thread In such a little belly?
Example 2. Multilingual poem by S1.

It is also notable that transliteration could be classified as a translanguaging phenomenon, where out-of-school literacy was integrated in classroom writing, even when literacy skills in the mother tongue did not seem to be fully developed. For example, Urdu words were produced in a Latin script in S3’s text. Transliteration was also used in P3 and P5 (see Appendix A (II)), where Polish and Amharic, respectively, were also transliterated. In contrast, in P4 and P6, words in Polish and Urdu were written by using Polish diacritics and the Urdu script, respectively (see Appendix A (II)); this indicated the existence of a certain level of literacy skills in the mother tongues in question.

In a way, the free multilingual poem produced by S9 was similar to an identity text (see Section 1.2), because S9 chose a topic related to their own interests, namely, their frustration with changes occurring in video games, and drew on their linguistic resources both in a complementary and parallel way (see Example 3). This makes it a unique instance in this sample set. Since S9’s language portrait was not available for this study, it was impossible to fully analyze their choice to visually mark the Norwegian words by underlining them. However, if their visual marking followed the pattern of most texts (see Appendix A (II)), it is likely that Norwegian was their mother tongue. It is, however, clear that English and Norwegian were part of their repertoire, and that they probably used both languages in the gaming domain. Most of the translanguaging here was complementary because different languages were used for different concepts, but parallelism was also used for concepts such as game and sucks, which were expressed both in English and Norwegian. In this case, as in Example 2, it was possible to discuss areas of linguistic overlap and differences between the two languages, in words such as like and favorite.

In general, multilingual poetry was mainly based on complementary translanguaging, but it did allow for parallel-based comparisons between languages. While the complementarity aspect would typically require the readers to understand all the languages employed in a poem, here, this was not problematic since most students used a Norwegian text, which they interpreted by means of code-switching. Consequently, T2 was able to understand the texts simply by reading the interpreted version together with the Norwegian one, and thus engaged with translanguaging and co-learning (see Section 1.1 and Section 3.1.1).

Jeg gotta like spille fortnite men i never win the game og din so …2 tinx or gay … var the game that was m[y] most favorit spill but the changes is a haug of sukcs epic games er en haug som suger	I gotta like playing fortnite but I never win the game and your so… tinx or gay … was the game that was m[y] most favorite game but the changes is a pile of sucks epic games is a pile which sucks
Example 3. Free multilingual poem by S9.

The students expressed their linguistic identities by selecting and representing different languages from their repertoires, which they often also marked visually. Based on T2’s report, the students read their poems out loud in class and compared the rhythm and rhymes to the original Norwegian poems. This indicates that one of the outcomes of this activity was the discussion of multilingual resources for a literary stylistic effect, which would correspond to the poetic function of translanguaging (see Section 1.1). In terms of student engagement, T2 reported that a student who usually found it challenging to speak in class appeared motivated when it was their turn to read their poem out loud to their classmates.

3.2. Teachers’ Experiences and Needs at Different Stages of the Implementation Process

First, the teachers’ perspectives on multilingual practices at the end of the project appeared to be different from those they reported at the start of the project. At the start of the project, T1 and T2 generally showed interest in using a multilingual approach, which seems to be relatively common in Norwegian settings (see Iversen 2017; Burner and Carlsen 2019). However, their main reported need was to develop specific tools for new teaching practices tailored to the needs of linguistically and culturally diverse classrooms (see Šurkalović 2014). Interestingly, T1 stated that she was aware that strong literacy skills in the mother tongue represent a solid basis for further language acquisition (see Cenoz 2003), but she admitted that using translanguaging in the classroom for this purpose was a novel idea to her. Furthermore, both teachers expressed their intention to regularly work in ways that would validate linguistic, cultural, and religious diversity. However, they were unsure of how this could be carried out, and explained that the level of awareness, competence, and experience with such work was relatively low at their school. At the end of the project, the teachers expressed their intention to continue integrating translanguaging regularly into their lessons and expressed different needs. More specifically, T1 felt confident that she had gained sufficient knowledge and practical tools to apply a multilingual approach in her work with diversity. She mentioned that the main challenge was to find time to plan and implement such work consistently, in a cross-curricular manner. T2 claimed that to continue his work in a similar manner, he would need a support network to provide him with scholarly input and critical perspectives on his work.

Interestingly, long-term work with designing multilingual pedagogies might have changed T2’s attitude towards the use of the directive function of code-switching (see Section 1.1). In other words, at the first mentorship meeting, T2 expressed the concern that allowing the students to speak their home languages at school would involve the risk of excluding certain interlocutors from the conversation, which is a fear reported by other scholars in the Norwegian context (see Section 1.3; Burner and Carlsen 2019). However, at the last mentorship meeting, T2 stated that he did not, in fact, experience the exclusive use of the directive function of code-switching among his students during the project work.

Both teachers worked actively with the implementation of multilingual practices in the classroom, which stands in stark contrast to the passive role taken by the teachers in Iversen’s study (2017) (see Section 1.3). Their work with multilingual practices was characterized by using established didactic methods, such as schemata activation, modeling, task differentiation, and scaffolding, which is unsurprising considering their extensive experience with teaching (see Section 2). Furthermore, T2 allowed a wide understanding of knowing a language among students in their work with language portraits (see Section 3.1.1) and T1 fostered co-learning processes among students and between students and herself during the fairytale work (see Section 3.1.2).

In general, both T1 and T2 reported that they experienced a high level of student engagement during or after the multilingual activities. For example, T1 mentioned that one of the immigrant students in her class who had typically been silent had for the first time started to contribute to class orally after the multilingual activities. T1 was especially impressed because this student was in a unique learning situation where they cooperated with a peer whom they had not previously met. Although T1 was reluctant to consider this a direct result of the multilingual work, she did believe that an active use of the students’ mother tongues in the classroom was beneficial for the students because they understood that their teacher valued and validated them as important resources. This belief was also shared by T2, who gave an example by referring to a student in his class who seemed more engaged with classroom work when their mother tongue was involved (see Section 3.1.3).

When asked about their needs concerning future work with translanguaging in the classroom, the teachers had different answers. T1 mentioned that she intended to integrate a multilingual perspective into the activities she would typically use, by, for example, encouraging the students to create multilingual mind maps for different topics. She also believed it was beneficial to speak English every now and then during other classes. Furthermore, T1 mentioned that simply asking the students for equivalent words in their home languages was a very convenient way for her to validate and acknowledge these languages in the classroom. Interestingly, she believed that by implementing a multilingual approach in all subjects in the ways they had implemented it during the training program, the teachers would fulfil the requirements for SNO (see Section 1.3). In comparison, T2 mentioned that he planned to first take a break to allow himself more reflection time, and then share more of his experiences with his colleagues to attempt more collaborative work in their professional community. He also mentioned that he would consult one of the school leaders to determine the best future course of action. In fact, both teachers mentioned that the school leader in question was very interested in taking on a multilingual approach at their school and, thus, was very supportive of their work and engagement with the training program.

4. Discussion

The findings in this research project can be classified into three main categories based on the three research questions presented in Section 1.4: (RQ1) the types of multilingual practices employed by teachers and their aims and outcomes, (RQ2) the representation of home languages in student texts, and (RQ3) the teachers’ reported experiences and needs at the different stages of the implementation process. These are discussed in more detail in the subsections below.

4.1. Multilingual Practices Employed by Teachers and Their Aims and Outcomes

The teachers in this study employed language portraits, parallel translanguaging on multilingual fairytale posters and in dictionaries, and complementary translanguaging in multilingual poetry. Overall, the aims of such multilingual practices were to allow the students to reflect on their multilingual repertoires, use their repertoires to activate their schemata and develop literacy skills, and raise awareness about the existence of linguistic diversity both among students and teachers. The general outcomes of these multilingual practices included increased metalinguistic awareness among students, acknowledgment of minority languages in the classroom, development of multilingual literacy skills, and co-learning both within and outside of school. A challenging aspect of multilingual practices is that it may be difficult to adapt them to appropriately address the needs of ethnic Norwegian students, who may not necessarily have many different languages in their repertoires. Although this may be addressed by including different dialects, similar challenges may arise in other contexts of applying multilingual pedagogies.

Based on the findings presented in Section 3.1, it can be concluded that multilingual practices such as language portraits were a successful ingredient for the implementation of multilingual pedagogies. First, the activities facilitated the students’ development of metalinguistic awareness and metalinguistic reflection, while also providing them with a sense of mastery and pride concerning their identities (see Krulatz et al. 2018). Second, language portraits validated the students’ linguistic identities through student-teacher dialogues and in schoolscapes consisting of the students’ drawings (see Gorter 2017; Cummins et al. 2005). There is, however, a potential caveat worth mentioning here. If the one domain-one language associations ensuing in student-teacher dialogues remain unchallenged, they may tacitly legitimize the dominance of majority languages, such as English and Norwegian, in the educational sphere. Third, multilingual work based on language portraits led to co-learning among teachers, students, and parents alike (see Wei 2014). Co-learning also involved the parents of multilingual students in their work with multimodal dictionaries. In this case, co-learning had two main advantages: (1) it lent fluidity to power positions in the classroom, as teachers allowed themselves to learn from their students, and (2) it constituted a significant step for promoting an open dialogue between the school and home concerning the students’ linguistic development.

An important result of employing multilingual posters and poetry was that minority languages were recognized as valid carriers of literacy and literary expression in written texts by both teachers and learners (see Section 3.1.2 and Section 3.1.3). In other words, in their work with multilingual posters and poetry, the students used their mother tongues in final text products, with two main implications. First, the students were engaged and proud to activate and share their linguistic backgrounds with the class. Second, the students used their mother tongues in writing, irrespective of their level of literacy, which contributed to their multilingual literacy development. At the same time, their out-of-school practices were validated as resources for multilingual literacy development at school (cf. Beiler 2019).

Furthermore, the students’ work with multilingual poetry led to the use of expressive, poetic, and pragmatic functions of translanguaging (see Section 1.1). The students expressed their identities by selecting the languages they would use in their poems and explored the poetic function of translanguaging by experimenting with rhythm and rhyme, which Beiler (2019) highlighted as an important component of a multilingual approach to writing literary texts. By pragmatically using visual cues such as colors and underlining to mark different languages, they helped the reader to faster identify which Norwegian words were replaced by words in other languages (see Section 3.1.3).

Finally, the work with multimodal dictionaries was least present in the data in the present study, but it nonetheless indicated how parallel translanguaging could be used for developing metalinguistic awareness, while working with idiomatic vocabulary learning. In other words, the representation of words and idiomatic phrases in different languages in the same table or chart may be used as a starting point for comparisons between linguistic forms and identifying similarities and differences between various languages the students are familiar with (see Section 1.2 and Section 1.3). Surprisingly, such comparisons between languages seemed to be missing from the teachers’ accounts, even if they would have been facilitated by multilingual posters, dictionaries, and poetry. Thus, certain multilingual activities may not fulfil their full potential if they are not followed by reflective discussions, even if they otherwise lead to successfully showcasing linguistic diversity in educational settings.

4.2. Home Languages in Student Texts

One of the most significant findings of this study is that the students employed their mother tongues across the various multilingual practices, often in the same texts with Norwegian or English or both. It should be noted that here too, as in other studies (Beiler 2019; Burner and Carlsen 2019), the majority language, Norwegian, was prioritized, which is unsurprising in the Norwegian educational sphere. However, the notable aspect here is that, even when Norwegian was selected as the main language of interaction between teachers and students, English was always included, from discussing its importance in language portraits, to employing it in multilingual dictionaries. This stands in contrast to the findings of Burner and Carlsen (2019), where teachers seemed to prioritize Norwegian in English lessons.

The representation of home languages in multilingual poetry and posters revealed a variety of literacy skills, ranging from advanced literacy skills to basic literacy skills. In other words, some students who spoke Urdu and Turkish, for example, used the conventional scripts and diacritics of the respective languages, showing an advanced level of literacy, while students who spoke Amharic and Polish used transliteration, showing perhaps a more basic level of literacy in these languages. However, transliteration may be considered a useful translanguaging practice that allows learners to transfer their out-of-school communication resources to support the development of their literacy skills in the classroom (see Section 3.1.3). This representation was also valuable because it provided the teacher with insight into the students’ abilities and needs concerning their literacy skills in their mother tongues.

On multilingual posters and in poetry, code-switching between different languages was accompanied by visual marking, which implicitly indicated that the students could distinguish between the different languages in their repertoires. In most cases, visual marking was employed to highlight words or stretches of text written in the students’ mother tongues, with different purposes. As pointed out above, in multilingual poetry, which employed complementary translanguaging, visual marking may be considered a pragmatic mechanism to help the reader quickly identify and navigate unknown words in the interpreted texts by comparing them to the original Norwegian texts. However, visual marking may have a different function in the multilingual posters, which contained parallel translanguaging. Since two languages were typically used to convey the same content, the reader was not likely to encounter any comprehension issues, which eliminated the pragmatic value of visual marking. In the multilingual posters, mother tongues were visually marked by using a larger script and repeating the content in these languages (see Section Multilingual Fairytale Posters). Consequently, it may be inferred that visual marking in multilingual posters reveals the important role played by mother tongues in the students’ repertoires. From a theoretical perspective, a visual literacy framework would have been complementary for a multimodal analysis of the student texts.

4.3. Teachers’ Experiences and Needs

In terms of teachers’ experiences and needs, the most significant finding revealed that working with the design stage of multilingual pedagogies may have an overall positive effect on the stance component, as conceptualized by García et al. (2017). At the start of the project, the teachers were interested in adopting a multilingual approach (see Section 3.2). However, they admitted that they lacked the competence required, which is comparable to the findings of Šurkalović (2014), and voiced certain concerns regarding the use of minority languages as an exclusion mechanism, which was also pointed out by Iversen (2017) and Burner and Carlsen (2019). However, at the end of the project the teachers had a favorable stance towards the spontaneous use of multilingual practices in the classroom and planned to include them in their regular teaching. This may be explained by the fact that they worked with designing their own multilingual pedagogies over time, with available theoretical support and mentorship throughout this process.

5. Conclusions

This study set out to investigate the implementation of multilingual pedagogies in a long-term project that aimed to equip teachers to perform in linguistically and culturally diverse classrooms. In this project, the implementation of multilingual pedagogies consisted of three main types of multilingual practices centered around language portraits, parallel translanguaging for vocabulary work, and complementary translanguaging in multilingual poetry. The focus on linguistic repertoires in their entirety contributed to creating a space where the boundaries between different linguistic resources were softened.

The theoretical make-up of this study benefited structurally and terminologically from the deliberate integration of translanguaging conceptualizations with code-switching frameworks, which indicates that further research in the field may benefit from an integrated perspective of these two paradigms. In the present study, this allowed the association of types of multilingual practices with specific aims in the classroom, such as employing parallel translanguaging for multilingual vocabulary work with a potential focus on metalinguistic awareness, and complementary translanguaging for developing multilingual literacy skills. Similarly, certain functions of translanguaging, such as the directive, expressive, and poetic functions, were useful for describing both the teachers’ stance to multilingualism and the multilingual practices reflected in the students’ texts.

An important conclusion of this work is that the implementation of multilingual pedagogies was an individually tailored process, which at the same time depended on external support provided by organizations such as UDE and higher–education institutions. It is admirable that the two participants in this study managed to carry out this project, given the special conditions imposed by the COVID-19 pandemic. Their in-depth work was not merely the result of passively following all the steps of a professional development program. In fact, they invested time and effort in selecting the most relevant tools for their teaching context and adapting them to lay the foundations of a tailored multilingual pedagogy for their group of students. This study illustrates how teachers may create their own multilingual pedagogies based on their access to resources, networks, mentorship, leadership support, and personal dedication.

This study revealed that the use of mother tongues, home languages, English, and Norwegian together, in various patterns and types of texts contributed to building and promoting multilingual pedagogies. More specifically, it involved multilingual practices successfully employed for a variety of purposes, such as developing metalinguistic awareness, showcasing multilingual identities and literacy practices in the classroom, working with vocabulary learning, and encouraging students to express their linguistic identities. However, more work is needed on metalinguistic awareness in conjunction with both parallel and complementary translanguaging. Another important direction for future research is the exploration of co-learning possibilities beyond the classroom, by directly involving parents or caregivers of multilingual students in the development of multilingual pedagogies.