Performative functions of multilingual policy in second language education in Sweden

ABSTRACT

This paper targets the aim of ‘strengthening the students’ multilingualism’ in a second language subject in Sweden, Swedish as a second language (SSL), which represents a relatively rare L2 design internationally. The study investigates how ideological space for multilingualism with regard to this aim opens implementational space at the classroom level [Hornberger. 2002. “Multilingual Language Policies and the Continua of Biliteracy: An Ecological Approach.” Language Policy 1 (1): 27–51]. We draw on ethnographic fieldwork in three linguistically diverse upper secondary schools in Sweden, where SSL is a parallel subject to Swedish, free of choice, and is taught by qualified teachers. Ideological and implementational space was found to be primarily for empowering students and valorizing their multilingualism and less for learning functions and pedagogical multilingual practices. Some students emphasized multilingual aspects as being empowering and as a reason for choosing SSL, indexing the performative function [Fairclough. 2014. “Semiotic Aspects of Social Transformation and Learning.” In The Discourse Studies Reader. Main Currents in Theory and Analysis, edited by Johannes Angermuller, Dominique Maingueneau, and Ruth Wodak, 378–387. Amsterdam: John Benjamins] of macro-level policy in combination with teacher agency. We argue that in spite of the brevity of the formulated multilingual macro policy and its restricted ‘potential space’ [Johnson 2011. “Implementational and Ideological Spaces in Bilingual Education Language Policy, Practice, and Research.” In Bilingual Education and Bilingualism: Educational Linguistics in Practice: Applying the Local Globally and the Global Locally, edited by Frances M. Hult and Kendall A. King, 126–139. Bristol, GBR: Multilingual Matters], it still had a significant bearing at the local level. This finding reinforces the importance of researching established macro-multilingual policies in situ in education.

KEYWORDS:

• Second language instruction
• Swedish as a second language
• student perspectives
• equality in education
• contrastive insights

Introduction

In Sweden, 150–200 languages are spoken (Institute for Language and Folklore 2019).¹ Although Swedish is the ‘principal’ language, according to the Swedish Language Act, there is also legal support for the development and use of other languages. This paper focuses on multilingual policy and practice with regard to the school subject Swedish as a second language (SSL), which targets multilingual students who learn in and through an additional language (25% of students in compulsory school, and 32% in upper secondary school, Swedish National Agency for Education 2019).

SSL has been a parallel subject to Swedish (SWE) in compulsory school and upper secondary school since 1995, and is largely similar in terms of curricular content, goals and grading criteria. This study focuses on one important difference, namely the stated aim of SSL to ‘strengthen the students’ multilingualism’, as highlighted in this quote from a syllabus:

The subject contributes to strengthening students’ multilingualism and confidence in their own language ability […] The students shall also be given the opportunity to reflect on their multilingualism and their prerequisites to conquer and develop a functional and rich second language in Swedish society. […] Multilingualism is an asset for the individual as well as for society, and by comparing language knowledge and language experiences with others the students shall be given the opportunity to develop a better understanding of what function language has for communication, thinking and learning. (Swedish National Agency for Education 2013, our transl.)

Furthermore, according to a more specified SSL goal on metalinguistic awareness, the students are to develop the competence to compare Swedish with their (other) ‘mother tongues’ and other languages:

The ability to compare the Swedish language with one’s own mother tongue and other languages that the students have competencies in, and ability to reflect on one’s own language development (Swedish National Agency for Education 2013, our transl.)

To our knowledge, there are no previous studies on the multilingual aspects of the SSL syllabus or multilingual practices within this particular second language educational design (see, however, Rosén and Bagga-Gupta 2015, with regard to multilingual practices within Swedish adult second language education). The vast body of literature showing the benefits and empowering role of students having the ability to use their whole multilingual repertoire in educational contexts (e.g. Cummins 2000; Hornberger 2002) also indicates the importance of investigating multilingual policy within this type of second language subject. We argue for the importance of investigating how macro multilingual policies are enacted and negotiated at the local level (e.g. Canagarajah 2006; see also Flores and García 2017 for a critical review of bilingual policy), and, as here, from a student and teacher perspective (see Slembrouck 2010).

The study draws on fieldwork in SSL classrooms within three linguistically and culturally diverse upper secondary schools (Hedman and Magnusson 2019), Rowan, Birch and Pine,² where the SSL teachers displayed different approaches towards multilingualism within SSL. As we have discussed elsewhere (e.g. Hedman and Magnusson 2019), SSL represents a relatively unique L2 educational design from an international perspective, as the goals and content of SSL are largely similar to those of SWE, implying high academic content in SSL regarding literacy, and literary and linguistic content. Both SSL and SWE are core subjects that prepare students for upper secondary and tertiary education, with language and literature as their main content. In the upper secondary school level, SSL is voluntary, that is, the student decides whether to study SSL or SWE.³ SSL is similar to pull-out programs, in that SSL students are often (but not always) separated from those in SWE. However, due to its high academic content, SSL does not represent a ‘language first/content later’ approach, which has been a main critique of pull-out services (cf. Horst 2010; Hedman and Magnusson 2018). Nor can SSL be straightforwardly compared to sheltered instruction designs (Wright 2010), foreign language instruction or any kind of bilingual programs. SSL is, further, distinct from Swedish language introductory programs, which are framed for newly arrived students and are not regulated by national curricula (see Hedman and Magnusson 2019). In a previous article (Hedman and Magnusson 2018, 2019), we disentangled discourses on SSL within the academic field, among which were discerned, on the one hand, the ‘SSL discourse’, and, on the other, the ‘inclusion discourse’. Whereas the first promoted the need of a second language subject for students who learn an additional language and learn through that language, primarily, on basis of arguments related to language development, the latter opposed the separation of students in SWE and SSL, depicting SSL as discriminatory and stigmatizing. Similar viewpoints have been raised in public debate, where SSL has repeatedly been called into question. On the contrary, advocates of the SSL discourse claim the view that all students would benefit from the same language education is an expression of equality-as-uniformity (as opposed to equality-as-equal-opportunity, Westling Allodi 2007; see Hedman and Magnusson 2018, 2019), which reveals a kind of ‘difference blindness’, reflective of a monolingual Swedish hegemony (e.g. Lindberg 2009).

Contrastive language analyses, as in the quote from the SSL syllabus above, are found also in the Mother Tongue (MT) subject syllabus, reflecting a historical embeddedness of SSL in a Swedish pluralistic language education policy (see Hedman and Magnusson 2018, 2019). MT offers multilingual students non-mandatory tuition in their ‘mother tongue’, a language other than Swedish, provided that a parent/caregiver speaks the language, and that the language is spoken at home⁴ (Ministry of Education and Research 2010, 800). Furthermore, newly arrived students can receive ‘language guidance’ in their ‘mother tongue’ to scaffold content learning in Swedish (Swedish National Agency for Education 2015). Moreover, in the Swedish Language Act, it is stated that a person ‘with a mother tongue other than Swedish, sign language or national minority language shall be given the possibility to develop and use his/her mother tongue’ (Ministry of Culture 2009, 600, §14, our transl.). In this limited sense, there is state-mandated support for more languages than just Swedish. However, as policy documents are often heteroglossic and even contradictory (Hellberg 2012; Hult 2017), the ‘discourses in place’ (Hult 2015, 219) of the Language Act states, at the same time, that Swedish is the ‘principal language’, and, thus, Swedish is ascribed the uppermost position in a hierarchy of languages (on linguistic hierarchies in Swedish educational policy, see e.g. Hult 2012, 2017). Likewise, although the SSL syllabus comprises objectives on multilingualism, it affords more space for the analysis and development of Swedish.

Educational institutions have been found to be important sites for how paradoxes of multilingualism and linguistic friction are lived and negotiated in multifaceted ways (e.g. Jaspers 2015), although often on the basis of a ‘monolingual habitus’ (Gogolin 1994). For example, in a study of teacher attitudes towards multilingualism among Swedish primary school teachers, Lundberg (2019) identified a complex sets of beliefs. These comprised multilingualism as a right and the acknowledgment of multilingual practices in the classroom, parallel with opposing views, for example, that multilingualism was given too much space at the expense of students’ Swedish development. Such paradoxes illustrate how linguistic friction may be present in the same educational setting (Jaspers 2015). Further, as developed below, policy and practice need not necessarily be distinct (Spolsky 2004), as language practices per se constitute policy-making (cf. Bonacina 2010). For the purpose of this paper, we focus on how the syllabus formulations carve out an implementational space for multilingualism (Hornberger 2005). These formulations encompass contradictions: On the one hand, the overarching curricular aim of contributing to ‘strengthening the students’ multilingualism’ is both vague and demanding in its indeterminability, calling for a large undertaking on behalf of the SSL teacher if taken literally. On the other hand, the only multilingual goal that has a counterpart in the grading criteria, which must be assessed, relates to contrastive language analyses. The multilingual aspects of SSL may therefore also be seen as being restricted (cf. Wahlström 2018).

Aims and theoretical framing

The overarching aim of this paper is to explore a second language subject for which aspects of multilingualism are explicitly stated in the syllabus, and investigate how and to what extent this ideological space opens up space for multilingual practices. Johnson (2011, 129) outlines Hornberger’s (2002, 2005) notions of ideological and implementational space as follows:

Multilingual language policies that promote multilingualism as a resource […] open ideological space for multilingualism and bilingual education but this space is, in a sense, only a potential space, because language educators and language users must take advantage of this space by implementing multilingual educational practices. In other words, these opened ideological spaces ‘carve out’ implementational spaces at classroom and community levels which language educators and users, in turn, must ‘fill up’ with multilingual educational practices. (Johnson 2011, 129)

We particularly acknowledge this potential space, or the ‘agentive spaces in which local actors implement, interpret, and perhaps resist policy initiatives in varying and unique ways’ (Hornberger and Johnson 2007, 509). Theoretically, we see teachers as ‘policy subjects’ who do not implement or read policy as something ‘pre-determined’ (Ball et al. 2011, 612). Policy reading may, as outlined above, be contradictory, and implies, in line with Pennycook (2006), that power is located not only at a macro policy level, but also at a micro level through discursive practices. However, as Johnson (2009, 140) points out, Pennycook’s move to the micro level does not necessarily imply agency, but, on the contrary, ‘positions discourse, and therefore discoursers’ who act out ‘larger power relationships over which they have no control’. Although acknowledging teachers’ agency, our analyses are thus sensitive to macro discourses as being possibly ‘performative’, with a shaping function that impacts material and semiotic reality (Fairclough 2014; also Hedman and Magnusson 2019), and that such discourses may have effects (Foucault 1980; Marshall 2009) and govern human behavior (see Johnson 2010, 62).

As outlined previously, our point of departure is that the habitus of the Swedish school is basically monolingual, but that a possible opposing multilingual discourse is voiced in the SSL syllabus, which teachers may or may not enact in educational practices, and that the teachers may also give voice to other or similar discourses on multilingualism.

More precisely, the aim of this paper is thus to analyze and discuss how multilingualism is discursively constructed in SSL by students and teachers. On the basis of fieldnotes from classroom observations, we also analyze whether and, if so, to what extent multilingual practices are in fact salient in the SSL classrooms and for what functions. The study is guided by the following questions:

• How do local agents, such as teachers and students, express and (re)construct SSL in terms of multilingualism?

• What space is there for teacher agency, and multilingual practices, in relation to the multilingualism aspects of the syllabus?

Different concepts are used to analyze multilingual practices and discourses on multilingualism. First, to analyze the functions of multilingual practices and to critically evaluate their reach and scope, we use Halliday’s (1993) theoretical lens of three suggested aspects of language and learning: ‘learning language’, ‘learning through language’, and ‘learning about language’. This framework is based on the proposal that learning is a semiotic process, and that experience ‘becomes knowledge’ through language (93–94). Children engage early on in a process of both learning language and learning through language, in which schooling and literacy in and of themselves imply reconstructions of meaning-making resources throughout the school years (see Halliday 1993; Christie 2012). The third aspect, learning about language, was suggested for pedagogical settings (Halliday 1993). With the reservation that the theory suggests that all learning is a meaning-making process, it allows us to consider the functions of multilingual practices. Additionally, the potentially empowering role of multilingual practices, as formulated by Cummins (2000), will be considered within SSL. This is akin to Halliday’s interest in linguistic and educational equity (Harman 2018) and to critical approaches to language teaching (e.g. Leeman 2018).

Secondly, additional theoretical concepts function to highlight multilingual practices from a language ideological perspective. Through performative theatrical language metaphors, Goffman (1959) demonstrated how multilingual practices may be part of legitimate frontstage classroom activities or hidden as illegitimate backstage discourse. Similar but not identical processes are present in the construction of ‘ideological representations of linguistic differences’ (Irvine and Gal 2000, 37). For example, erasure occurs when language ecologies are simplified and refers to ‘the process in which ideology [ … ] renders some persons or activities (or sociolinguistic phenomena) invisible’ (38). Both from the perspective of language planning and individuals’ perceptions of languages (Hult and Hornberger 2016), language may be considered a right, a problem or a resource (Ruiz 1984).

Methodology

As part of a larger research project on ideologies and practices within SSL as a subject, this study draws on ethnographic fieldwork done over a period of one and a half years in three upper secondary schools with high ratios of multilingual students. In the following, the selected data in these schools are outlined, as well as the data collection procedures and the analyses.

Data collection

In particular, the study draws on fieldnotes from our SSL upper secondary classroom observations (58 lessons in total), as well as transcripts from interviews with 15 SSL students and three SSL teachers, one from each school: Stephen from Rowan, Wera from Birch, and Sandra from Pine. Each had taught SSL for five years or more, and were formally qualified to teach SSL, implying at least 90 ECTS credits in Swedish as a second language. All schools have a majority of ‘foreign-born’ (the term used in Swedish statistics) students: from approximately 75% (Pine) to approximately 55% (Rowan) (Swedish National Agency for Education 2018). The pool of interviewees comprised students who had immigrated to Sweden during secondary school (3), during middle school (7) and students born in Sweden (5). None of the students studied the mother tongue subject at the time of the interviews. The Birch teacher also taught language introductory classes for newly arrived students not yet qualified to begin an upper secondary program.

The audio-recorded interviews were conducted in the schools and lasted for about an hour. They were later transcribed and coded in Nvivo software following an abductive analytical process (Schwartz-Shea and Dvora 2011). In this paper, the transcripts were subsequently translated into English.

Data analyses

For the purpose of this paper, we focus on the students’ narratives centering around their multilingualism. These arose spontaneously in response to our questions on why they had chosen SSL, and we did not ask specifically about their understanding of the curricular formulations on multilingualism. Likewise, in the teacher-interview transcripts, we attend to the teachers’ narratives on their students’ multilingualism, often in response to why they thought their students chose SSL, how they informed their students’ choices, and whether languages other than Swedish were used in the classrooms. As in the student interviews, multilingualism tended to be broached spontaneously, without us asking specifically about its role or the teachers’ interpretations of the syllabi in these respects. This is noteworthy, as our point of departure is that interviewing is a social practice and that collected interview data are socially co-constructed (Talmy 2010).

The issue of curricular formulations on multilingualism represents a ‘rich point’ (Agar 2006), since choosing SSL due to its multilingual curricular content departed from our expectations as researchers. Following that, we also analyzed the portions of our fieldnotes that concerned multilingualism. The analyses thus concern core issues of how multilingualism was valued, its role in stakeholders’ constructions of SSL, and whether it emerged as peripheral or central. For classroom settings, multilingual practices were analyzed through the lens of Goffman’s (1959) language ideological metaphors, either as occurring in the frontstaged legitimate classroom discourse, or as backstaged, implying a denial of their legitimacy, or even an indication of a linguistic erasure process (Irvine and Gal 2000). To approach the purposes of observed multilingual practices, their functions were analyzed in terms of Halliday’s (1993) aforementioned types of language learning, and as possibly relating to empowerment (Cummins 2000). Furthermore, in analyses of interviews and classroom discourse, we consider how ‘explicit talk about languages [ … ] indexes particular language ideologies’ (Martin-Jones 2015, 102), specifically, in terms of language as problem, right and resource (Ruiz 1984).

In analyses of discourse, we draw on Gee’s (2010) analytical toolkit, particularly regarding how language ‘builds things in the world’ (83ff), for example, relevance, significance and identity. By creating ‘connections’ through linguistic means, relevance may be constructed when a teacher explicitly connects SSL to multilingualism. The connection is not only made relevant but also significant when the teacher further connects multilingualism to the high stake choice of a core subject, thus constructing a choice between ‘SSL/the multilingual subject’ vs. ‘SWE/the monolingual subject’. Further, speakers take on roles and identities through language, for example, when a teacher performs an authoritative teacher identity vs. an inclusive identity, for example, through expressions of a shared identity as ‘multilingual’ (see Findings). In discourse, speakers also perform actions, such as promises, prohibitions or encouragement, for example, when a teacher encourages multilingualism by saying that multilingual practices are ‘wonderful’ or by making connections between richness and being multilingual. Such connections also serve to assign significance to multilingualism, that is, these categories may overlap.

Finally, the use of the term discourses in place (Hult 2015) refers in this paper to discourses in play at a certain time and place, and could thus refer to various levels, e.g. both de jure and de facto policy (see Hult 2015).

Findings

As previously reported in Hedman and Magnusson (2019), the three schools regularly informed new students about SSL at the beginning of each fall. In all schools, students were allowed to switch between SSL and SWE, which also happened regularly. In the following, the findings will be presented and discussed for each school, respectively, after which a concluding discussion follows.

Rowan

At Rowan, we followed some pedagogical themes in teacher Stephen’s SSL classes during our fieldwork, viz. language sociology and literature reading, for example, Sophocles’ Oedipus and Juli Otsuka’s novel The Buddha in the Attic.

In the interviews and in other conversations with Stephen, he explicitly acknowledged the importance of students being able to use their ‘mother tongues’ rather than Swedish for learning, thus voicing a resource perspective on multilingualism (Ruiz 1984). Stephen expressed that SSL was not only a subject for learning Swedish, but that an interest in multilingualism was a potential reason for students to choose SSL: ‘she may have chosen SSL because of her interest in her multilingualism’. Through a causative conjunction (because of), he marked a connection between studying SSL and multilingualism, thus creating a relevant connection to multilingualism (cf. Gee 2010). In addition, Stephen used the formulations of the SSL syllabus on multilingualism in the initial student information on SSL, thus legitimizing the choice of SSL through multilingualism. Again, by connecting these multilingual aspects to a legitimate choice for choosing SSL, one of the core subjects, he ascribed significance to multilingualism (see Gee 2010) and opened up ideological space for multilingualism within SSL at Rowan. In the interview, it also appeared that Stephen had been assigned the informal role of a ‘spokesperson’ on multilingualism at the school. For example, he was regularly consulted by the school management on first language study guidance for newly arrived students (cf. Swedish National Agency for Education 2015): ‘I usually, well, point out that they, well, need study support’. He also regretted that few students attended mother tongue instruction at the upper secondary level. Teacher Stephen was thus found to be positioned and positioning himself as an advocate for multilingualism, not only within SSL, but in the school at large. Thus, an identity of a spokesman who valued multilingualism as a resource was formed (Gee 2010, 114f.).

However, in terms of classroom interaction in SSL, Swedish was the dominant language, although other languages were occasionally spoken among the students. Teacher Stephen told us that he encouraged the students to use their first language(s) with each other and described occasions when students used several languages to treat curricular content. Through various words of encouragement, he created the opportunity for multilingual activity based on the students’ own choices (cf. Gee 2010), which possibly counteracted linguistic erasure processes within SSL at Rowan. The results of this encouragement were, however, only visible in occasional student-student interactions in our data. A potential contradiction (cf. Jaspers 2015), such as advocacy for multilingualism vs. a monolingual Swedish habitus in the classroom, could, according to Stephen, partly be explained by the fact that not all students wished, or were able to, use their ‘mother tongues’ in class (Excerpt 1). Perceptions among students that monolingual use of Swedish would be more beneficial for their development in Swedish also added to the linguistic friction in the classroom (Jaspers 2015). Stephen assumed that such connections (Gee 2010) between a strict monolingual Swedish classroom discourse and learning were more prevalent in the Swedish language subjects than in other subjects (Excerpt 1).

Excerpt 1. Interview with the SSL teacher Stephen at Rowan

Stephen:

[…] Well then in separate [teacher-student] talks, with them, I try to encourage them to use their first language. When it comes to learning in other subjects, it becomes, well, more current when they are supposed to read, read in chemistry or something. But there are, well, several students who don’t want that, but rather “no, now I shall learn Swedish, now I shall only use Swedish” um so that can be harder.

Researcher:

Just newly arrived students or … ?

Stephen:

Yes, yes, rather many are relatively newly arrived students. Then, it might also be that they don’t really know their first language well enough either.

Researcher:

hm.

Stephen:

Because, I know, there are students in the third [final] grade who are very active in their first language.

Researcher:

hm.

Stephen:

And they discuss in class with one another, a bit in Arabic or so, and you notice that they discuss math or something.

Researcher:

hm.

Stephen:

Or something that we work on in Swedish, and that’s, well, very wonderful, I think, that they, then you notice, well, that they also speak on an advanced level.

Researcher:

hm.

Stephen:

That they, they know Swedish well, so they could, well, do it in Swedish too, but that they still continue to use their first language.

Researcher:

Does this mainly apply to newly arrived students in some sense?

Stephen:

No, it’s, well, many, several of them, who have been here rather long, who … still do it.

In Excerpt 1, Stephen evaluates students’ ‘advanced’ first language use as a means for learning as ‘wonderful’, indicating recognition of their multilingual competence. Still, in spite of Stephen’s performed actions (Gee 2010) through his encouragement and the narrated instances of students learning through language other than Swedish, there was, in our data, limited space for multilingual practices in SSL. Three of Stephen’s students in their final year clarified that they usually talked Swedish in class, whereas they used Arabic during breaks. In the following interview (Excerpt 2), the researcher had previously asked whether the teachers recognized that the students ‘know other languages’. Adam’s memory of a language comparison he made implies an activity that relates to learning language and learning about language. Following that, Ahmed commented that he found a more pronounced interest in the students’ backgrounds within SSL, indicating a perceived resource perspective on behalf of the teachers (Excerpt 2):

Excerpt 2. Interview with students Ahmed, Mohamed and Adam at Rowan

Adam:

We had a small presentation about that.

Researcher:

Yes.

Ahmed:

Yes, we had.

Mohamed:

Yes.

Ahmed:

The first year.

Adam:

[slight laughter] ehm.

Ahmed:

Yes, they are a bit more interested in SSL, they are a bit more interested ehm in what we’ve studied before we came to Sweden

Researcher:

Yes.

Adam:

Yes, exactly.

Ahmed:

Our culture, our ways of thinking and such, also.

Researcher:

Yes.

Mohamed:

hm.

Ahmed:

Well yes, that’s true.

Researcher:

So you presented something about that [turning to Adam]?

Adam:

I simply, as far as I remember, compared Arabic with Swedish.

Researcher:

hm.

Adam:

and how you use slang, for example, there and here, and what a certain sentence could mean, how deeply people look into a sentence, sort of, we in Sweden and maybe for example in Syria and such.

Ahmed seems to ascribe the above-mentioned use of languages other than Swedish as additional functions, such as an acknowledgment of cultural identity (‘our’ culture) with potential empowering values (Cummins 2000). In the wording of ‘our’, the students also construct a shared identity distinct from others (cf. Gee 2010).

Another student in the same class, Dana, who was born in Sweden while her parents originated from an Arabic-speaking country, initially chose SWE, but soon changed to SSL. Like many of the students in SSL, an important reason for Dana to choose to continue studying SSL was the opportunities for pedagogical scaffolding (Hedman and Magnusson 2019), but she also emphasized the multilingual aspect of SSL. With regard to our question of why Dana had chosen SSL (considering that she was born in Sweden), Stephen acknowledged that Dana could just as well have studied SWE, but that her ‘interest in her multilingualism’ was a likely reason for her choice. In line with this, Dana told us, following an interview question asking how she felt about her choice of SSL in retrospect, that she found the language contrastive tasks to be the most important part of SSL (Excerpt 3).

Excerpt 3. Interview with student Dana at Rowan

Dana:

I remember in the first grade that we were all supposed to do a presentation where we compared our mother tongue with Swedish. And then it became, then when you looked it up and all well it’s so very different, and then, for example, that Arabic resembles Spanish, which I’ve studied since grade 6 [in comprehensive school], well, the actual sentence construction how that looks and that you have, you know, the pronouns on the verbs and such ehm and then I became so … I got a completely other technique when I write or so.

[…]

 

Researcher:

What is the most important thing you do in SSL, you think, now when you’ve been through it all?

Dana:

Hm the most important is well that you compare them, well, because that’s given me very much even in other languages, well, in English when we wrote our English seven now … So when we write such texts it becomes … you think more about how the sentence construction should look and such … so it’s most important to compare the languages so you know what happens, well, that you’re aware.

Dana created connections between the contrastive tasks and her choice of SSL, and established relevance (Gee 2010) between these and her learning of English. She added that she ‘liked’ the SSL syllabus more than the SWE syllabus, referring to the lack of multilingualism goals in the latter, and ‘if I had had that [Swedish] syllabus, I wouldn’t be as interested in the Swedish language at all’. In relating the language comparisons to her writing competence, she primarily emphasized perspectives of language learning and learning about languages. Her expressed experience confirmed our observations from Stephen’s classroom, where implementational space for multilingualism, above all, seemed related to the language comparison goals of the syllabus.

Birch

At Birch, we followed a theme on language sociology in SSL and lessons in an introductory class. Teacher Wera’s interaction with the students was not only aimed at negotiating meaning and promoting learning, but was saturated with an empowering stance of the students as multilingual (Cummins 2000) and a forceful resource perspective. This stance comprised an expressed demarcation between being monolingual and multilingual, and positive connotations between the latter and SSL (Excerpt 4).

Excerpt 4. Interview with the SSL teacher Wera at Birch

Wera:

But you learn Swedish as your second language because you are multilingual.

Researcher:

hm

Wera:

That you’re monolingual or multilingual are two completely different things and then you might need more time and different ways of working.

Researcher:

hm

Wera:

The brain must work in completely different ways if you have many languages … And there are those explanations. That it’s a means, well, that it’s not only compensatory, it’s an advantage to study SSL.

In Excerpt 4, Wera legitimized SSL through multilingualism, in a way that is non-stigmatizing, arguing that studying SSL is an ‘advantage’. That is, multilingualism was made relevant (Gee 2010) as a raison d’être of SSL. Frontstaged multilingual practices, as part of legitimate classroom discourse, were more prevalent in Wera’s language introductory classes for newly arrived students than in SSL. During these lessons, she consistently maintained a focus on the students’ other languages by, for example, asking for and using words or phrases in their other languages or asking them to explain something to their peers. She told us that she usually understood what was going on in class in different languages, as she, apart from English, Latin-based languages, and a non-European ‘mother tongue’, understood substantial parts of students’ talk in several non-European languages through long-term exposure in various classrooms. When asked how it felt when Wera addressed her in her mother tongue, a student in SSL told us that it felt ‘good’, thus confirming its empowering role. Both in interviews and classroom interactions, Wera explained that immigration and learning Swedish as an additional language was a lived experience that she shared with her students.

Although Swedish dominated the classroom discourse more in SSL, Wera kept pointing to the students’ resourcefulness and possession of linguistic capital within SSL (Excerpt 5), clearly indexing an orientation towards multilingualism as a resource (Hult and Hornberger 2016). In Excerpt 5, we exemplify Wera’s empowering stance in relation to a theme in school year 10 on language sociology and attitudes towards language variation. Wera emphasized ‘an SSL perspective’ on language sociology that included the students’ multilingualism: ‘What they do in Swedish is about the same thing but they don’t have the SSL perspective, you understand, they are not multilingual, but you are … that is big’ (fieldnotes). In the whole-class discussion on language variation and multilingualism, Wera intervened in a student’s reading of the curricular goals with questions reinforcing multilingualism as an asset, and suggested that the goals could be exceeded. In clarifying that this was her own goal, she reinforced her own agency in the interpretation of the curriculum (cf. Ball et al. 2011).

Excerpt 5. Student-teacher classroom interaction with Wera at Birch

Student:

[reads aloud from the SSL syllabus] You can do simple reflections upon your own language use and with some certainty use different strategies to develop your own learning

Wera:

Reflect on this; why does he swear, why does the teacher speak in this way? You’re an asset, here we have the world’s asset, that’s what I’ve always said, in here, there are countless numbers of languages. I hope, after this course, we can say something nice, maybe hello, love in each other’s languages, maybe write in Arabic. I’ve tried, it’s very difficult. These are the things we can do if we work hard. That’s my goal, it's not the goals of the syllabus.

In Excerpt 5, the students’ multilingualism was not simply praised (cf. Karrebæk 2013 on ‘trivial’ multilingualism), but was used to deepen the students’ understanding of the curricular content on language sociology, related to ‘understanding what function language has’. Wera also signaled that it may be time to learn language through the languages represented in the classroom, which is in accordance with the syllabus objectives of ‘strengthening the students’ multilingualism’.

Subsequently, Wera initiated a discussion on the motives for studying language variation, which developed into a section on the value of knowing one’s ‘mother tongue’ and several languages (Excerpt 6), related to the syllabus goal ‘reflect on their multilingualism’.

Excerpt 6. Student-teacher classroom interaction with Wera at Birch

Wera:

Well, why are you to learn about language variation in society? Perhaps study how social groups talk to each other … Why do we have language sociology in school?

Student (boy):

We shouldn’t feel lost.

Wera:

You will orientate yourself in society, you will not get lost … That’s actually very good. Well, girls, what do you say?

Student (girl):

[inaudible]

Wera:

To be able to communicate?

Student (girl):

I think that you, well, that it becomes easier to communicate with others. Don’t you get a stronger mother tongue when you know the grammar in your own language?

Wera:

Yes, that’s important in this class, for us as a group, because we have another mother tongue too.

Student (girl):

My dad has been rather strict with us using Turkish. That’s important to me, if you know the grammar you know it in the second language.

Wera:

If you get lost in Turkish, you get lost in Swedish – imagine, what a richness, we have several languages.

Standing out in Wera’s classroom discourse was her stance of an inclusive ‘us’ (‘for us as a group’, ‘we have another mother tongue too’), which included herself as a teacher and her personal investment, where she expressed a shared identity with her students. She also created connections between the study of language sociology and to ‘orientate in society’ and having ‘another mother tongue’ (cf. Gee 2010), which was evaluated as ‘important’ and as ‘a richness’. Wera employed her lived experiences (Busch 2017) of immigration and learning Swedish, and thus her experiences as a multilingual teacher to bring out aspects of multilingualism (cf. Colliander 2017 in which multilingual teachers referred to their own life trajectory as second language learners and ‘embodied cultural capital’ as assets in teaching). Wera acknowledged that her students might have more than one ‘mother tongue’, including Swedish, thus pointing to multilingualism as enriching in relation to ‘monolingual’ fellow students, in line with the formulations of the syllabus that multilingualism is ‘an asset’. She also performed an empowering and reversed act of ‘Othering’, overturning the ‘Othering’ of migrants in societal and public discourses (e.g. Rydell 2018), and reformulated the essence of ‘us as a group’ as opposed to the negative connotations that SSL and its students may have in public debate (cf. Hedman and Magnusson 2018). This is in accordance with critical approaches to language teaching, possibly promoting student agency (Leeman 2018, 350; cf. Cummins 2000).

As in Rowan, there were students at Birch who emphasized multilingual aspects as motives for studying SSL. Fjodar, who came to Sweden during middle school, strongly emphasized the opportunity of studying and ‘connecting with’ his ‘mother tongue’ in SSL (Excerpt 7). Compared to Dana (Excerpt 3), Fjodar highlighted dimensions of identity connected to the ‘mother tongue’ (cf. Gee 2010).

Excerpt 7. Interview with student Fjodar at Birch

Researcher:

And then we wonder why you chose Swedish as a second language?

Fjodar:

Ehm, well, Swedish as a second language and Swedish as a first language do differ to large extents, and in SSL it is part of the course to sort of connect with your mother tongue.

Researcher:

Yes, okay.

Fjodar:

Well, to be able to connect to that somehow, it is after all part of one’s identity to know from where you come, and I sort of value that and want to keep it.

Researcher:

Yes.

Fjodar:

That’s why I chose SSL.

Fjodar’s emphasis on the role of multilingualism in SSL resonates with Wera’s framing of the subject, including multilingualism as resource (cf. Hult and Hornberger 2016), as visible in Excerpts 5–6, and the relevance of language for identity (cf. Gee 2010).

Wera’s classroom work indexed an opened ideological space for multilingualism. Implementational space for all the multilingual goals of the syllabus was provided, but the empowering function of multilingualism stood out as most prominent, resonating with the formulation of multilingualism as ‘asset’.

Pine

Also at Pine, we followed a comprehensive thematic work on argumentative writing and on language sociology, in which students were introduced to central ideas on ‘cultural capital’ by Pierre Bourdieu (cf. Hedman and Magnusson 2019).

Unlike at the other schools, multilingualism was not part of the initial information about SSL at Pine, and none of the students mentioned the multilingual goals of SSL as a reason for studying the subject. When asked about differences between SSL and SWE, teacher Sandra said that in SSL ‘there is much focus on their own mother tongue and comparisons between Swedish and the mother tongue’. However, in our data, she did not include multilingualism in her classroom discourse beyond the mandatory language contrastive tasks, thus classified as learning about language, nor did she address any other aspects of multilingualism in the interviews. In the SSL lessons, languages other than Swedish were spoken occasionally among students, but not in the whole class frontstaged legitimate classroom discourse. Sandra claimed that many of her students, who were not newly arrived in Sweden, did not ‘practice any other language in parallel with Swedish’. She maintained that although more languages than Swedish were used in the students’ ‘extended families’, they did not speak these other languages at home, nor did they attend mother tongue classes. As a possible consequence, not all students in SSL seemed, in Sandra’s view, to be competent or ‘full’ speakers of their ‘mother tongue’. Sandra did thus not express a resource perspective on all students’ multilingualism as the Rowan and Birch teachers had, nor was implementational space opened to the same extent as in the other schools.

On the contrary, Brad, an SSL student at Pine who was born in Sweden, claimed to speak five languages ‘fluently’ or ‘even better’. According to him (Excerpt 8), these languages were spoken widely outside of school and at home in the area, which, to Brad, made the mother tongue subject unimportant and superfluous.

Excerpt 8. Interview with student Brad at Pine

Researcher:

Do you study the mother tongue subject now?

Brad:

No. I haven’t done it previously either.

Researcher:

OK … Is it common for people to not study the mother tongue subject here, you think?

Brad:

We speak the languages outside and at home so much, so I don’t think that all this about mother tongue languages and that sort of thing, it, it’s not really needed in society.

Researcher:

Mm … Do you feel like that?

Brad:

I feel like that … I don’t like it when it’s written in Arabic in certain stores in Stockholm and so on, it’s not supposed to be like that. You must learn Swedish and English, it’s as simple as that.

Researcher:

Where did you get that idea from?

Brad:

Uhm, because some people from here, in our neighborhood, expects the bus driver to speak Arabic … Why don’t you speak English, we ought to ask them.

His discursive construction of SSL did not embrace aspects of multilingualism, and, in Excerpt 8, he seems to downplay the significance of multilingualism, by denying the importance of ‘mother tongue languages’ and by evaluating multilingual practices in class and mother tongue education negatively (‘it is not really needed’, ‘I don’t like’; cf. Gee 2010 on how things are discursively construed as insignificant). In this sense, Brad thus closed the implementational space for multilingual use. In accordance with a linguistic hierarchy in Sweden (e.g. Hult 2012; see also Palm, Ganuza, and Hedman 2019), he expressed that he considered Swedish and English to be more important than other ‘mother tongues’, notwithstanding that Arabic was part of his and many other students’ language repertoire. In addition, Brad evaluated negatively some people’s alleged expectations to be served in Arabic in public spaces (Excerpt 8) and refrained from such seemingly unwarranted behavior. This reported form of ‘language-as-a-right-claim’ was thus dismissed by Brad.

Gabriel, who was born in Sweden, answered in the negative to the question of whether his knowledge of languages other than Swedish and English was used in SSL. He did say, however, that he would appreciate having a teacher with whom he could switch between languages, much as he did with his friends ‘all the time’. Two other SSL students, Sarita and Sarina, who immigrated to Sweden during their secondary school years and were interviewed together, asserted that they talked Arabic in SSL ‘sometimes but not very often’ and only between themselves; ‘either it’s whole class and there ‘Swedish is what counts’, or it’s only the two of us and then we can say it in Arabic’. Such statements, together with our classroom observations, reinforces an image of multilingual practices as less legitimate in SSL, since they were either ignored or positioned backstage (Goffman 1959). Another student, Ana, who came to Sweden during secondary school, reported that she did not get the same type of support in Arabic that she had previously experienced in secondary school, which had ‘helped her a lot’ (interview). Such statements may indicate a school ethos at large at Pine (cf. Allder 1993) characterized by a monolingual habitus and tendencies of linguistic erasure (Irvine and Gal 2000; Martin-Jones 2015).

Concluding discussion

The findings here contribute to insights into how SSL curricular macro language policy is enacted and how multilingualism is discursively constructed at the local level through the examples of these specific schools. The study reveals how widely local practices, emanating from the same macro policy, differ. With a view of teachers as ‘policy agents’ (Ball et al. 2011), the differences in how ideological and implementational space were opened are expected but still conspicuous: from tendencies of linguistic erasure, and possible deficiency perspectives at Pine, to an emphasis of multilingualism in everyday practices and interactions at Birch. The students’ own ideological spaces appear largely reflective of their teachers’ expressed language ideologies and constructed implementational space, with, for example, the downplaying of multilingualism by some students at Pine, and the emphasis on multilingualism and identity by a student at Birch. These patterns seem also indicative of the school’s culture or ethos at large (Allder 1993) in terms of the range of its monolingual habitus (Gogolin 1994). Importantly, however, is that the discourses in place in de facto policy (see Hult 2015) were, at the same time, found to be complex and even contradictory (cf. Lundberg 2019).

Despite the differences, the implementational space for multilingualism appears more or less restricted in all schools, particularly regarding learning through language (Halliday 1993), which only extended to ‘a potential space’ (Johnson 2011, 129), since we observed no systematic or planned activities in languages other than Swedish, and multilingual interaction mostly occurred backstage (Goffman 1959). Hence, none of the teachers used languages other than Swedish in a way that would challenge Swedish as the main academic language (cf. Jaspers 2015). Students were provided some opportunities to learn about language in the contrastive analyses, which encompassed some space for learning language in terms of lexicogrammar. On the other hand, multilingualism was neither at Birch nor Rowan constrained to ‘trivial’ use (Karrebæk 2013), given the analyzed functions. The absence of opportunities for learning through language could be due to the fact that the syllabus does not stipulate such functions, whereby the study illuminates limitations of the discourses in place within de jure policy (see Hult 2015). Teacher Wera at Birch was found to provide implementational space for the overarching curricular goals, i.e. regarding the strengthening of and reflections on multilingualism, and in her emphasis on multilingualism as an empowering ‘asset’. That is, the implementational space at Birch went beyond the necessary contrastive analyses that must be assessed.

In line with Hornberger (2005, 2), it is possible that the implementational space of the syllabus ‘serves as wedges to pry open ideological ones’. Our findings indicate that the multilingual aspects construct SSL as a potentially enriching and empowering subject, affording students’ the possibility to attend SSL by virtue of their multilingualism as opposed to ‘second language learners’ only, that is, affording students an added value. Teachers may also value these aspects, like teacher Wera, who considered SSL to be an ‘advantage’ for multilingual students. Although multilingualism as an ‘asset’ is stipulated in the syllabus, multilingualism for identity formation is not, but may be a byproduct in practice, indicating the significance of ethnographic methods for the exploration of language policy (Canagarajah 2006).

We conclude that the window opened for implementational space through the SSL syllabus is important to some students and their choice of SSL, which points to the ‘performative’ function of macro policy (Fairclough 2014; cf. Foucault 1980). We emphasize the role of established macro multilingual policies within education, and the significance of scrutinizing educational policy and practices in situ and in relation to the persons they concern (e.g. Johnson 2009 on ethnographic policy; Hedman and Magnusson 2019). Moreover, this study informs macro curricular and syllabus designs also in other national contexts with respect to their potential to open up ideological and implementational space for multilingualism. We do not, however, argue in favor of a minimum of space for multilingualism within macro policy, but rather, on the basis of this particular case, claim that in spite of this minimum, it may still matter at the local school level.

Acknowledgments

The authors are indebted to the anonymous reviewers for valuable comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Erik Wellander Foundation.

Notes on contributors

Christina Hedman

Christina Hedman holds a Ph.D. in research on bilingualism and is a Professor in Swedish as a second language at the Department of Language Education, Stockholm University, Sweden. Hedman's recent research has focused on multilingual development from education and policy perspectives.

Ulrika Magnusson

Ulrika Magnusson holds a Ph.D. in Swedish as a second language from the University of Gothenburg and is a Senior Lecturer in Swedish as a second language at Stockholm University. Ulrika Magnusson's research interests include literacy, learning and education from second language and multilingual perspectives.

Notes

1 Second language (L2) is used in accordance with official terminology in Swedish language education policy.

2 All names of schools and persons are pseudonyms. For reasons of anonymity, students’ languages are not specified.

3 In compulsory school, the school principal is formally responsible for deciding who should study SSL.

4 The same restrictions do not apply to national minority languages.