Teachers’ acts of legitimation in second language education in Swedish upper secondary schools

Abstract

Drawing from an ethnographic project on the subject of Swedish as a second language (SSL¹) in three linguistically diverse upper secondary schools, the aim of this study was to investigate how three SSL teachers, one from each school, discursively constructed SSL, and whether and how they legitimated their role as SSL teachers, in relation to previously analyzed academic and public discourses on the subject (Hedman and Magnusson 2018). The teachers both co-constructed discourses motivating SSL and contested discourses on SSL as a low status and inferior subject compared to the “real” Swedish subject, through expert authority (van Leeuwen 2008), with regard to scaffolding of advanced literacy. In addition, a role model authority embedded in a democratic professionalism was claimed, linked to the construction of the students as unprivileged. The paper contributes new knowledge on how teacher discourse and its legitimizing functions can be scrutinized in greater detail in order to gain a deepened understanding of the complexity of educational policy, where it is important to account for what educational practice is being scrutinized. Such visibilization of practice and teacher voice strengthens a participative framing for teachers on issues concerning their own practices.

Keywords:

• Academic literacy
• Educational policy
• Language minority students
• Language teaching
• Scaffolding
• Second language acquisition

Introduction

Drawing from a larger ethnographic research project on the role of the school subject of Swedish as a second language (SSL) in three linguistically diverse upper secondary schools in Sweden (Hedman and Magnusson 2018, 2019, 2020), this paper focuses on the perspective of teachers. SSL is a core school subject intended for students who speak a language other than Swedish, or for whom Swedish is a second language according to the terminology of Swedish language policy. In the curriculum, both academic subjects (SSL and the subject of Swedish) have a parallel design, which gives the same qualifications. At the upper secondary school level, students are free to choose which of the two subjects to study, while students in primary and secondary education are referred by the school to one of these subjects. The relatively low prestige of the subject of SSL as a parallel subject to Swedish is well documented (e.g., Swedish National Agency for Education 2017). One of our previous articles (Hedman and Magnusson 2018) encompassed an analysis of the competing academic discourses surrounding SSL in terms of equality-as-equal-opportunity or equality-as-uniformity (Westling Allodi 2007). It showed how the subject, on the one hand, is motivated and legitimized as pivotal to the learning of multilingual students. On the other hand, its legitimacy is contested, as it is considered stigmatizing (Hedman and Magnusson 2018). Thus, one study also investigated how upper secondary students in SSL perceive the subject, why they choose it, and why they continue to study it (Hedman and Magnusson 2020).

This analysis focuses on how three experienced SSL teachers, one from each school, discursively construct SSL, and whether and how they legitimate their role as SSL teachers, given the questioned status of SSL. We build on van Leeuwen (2007) acts of legitimation, which are responses to “the question ‘Why’: ‘Why should we do this?’ and ‘Why should we do this in this way?’” (p. 93).

Our claim is that thorough analysis of teacher discourse is needed regarding any educational provision for students learning in a new language, since these educational provisions tend to be situated in the crossfire of sociopolitical debate, and as it is necessary to scrutinize the perspectives of those who are directly affected. As we have argued elsewhere, such debates have been poorly informed by empirical data and it has not always been clear from research what type of educational provision is being discussed, and what the local conditions are (cf. Hedman and Magnusson 2020). Yet another problem has been that the legitimacy of SSL has been discussed through the lens of language educational provisions in other national contexts that are not easily comparable. The poor conditions under which the migrant students in Talmy’s studies were learning English in Hawaii (e.g., Talmy 2009, 2015), with unqualified teachers, content that failed to challenge learners, and an explicit pejorative classroom/school discourse regarding “L2 students”, serves as such an example. Such comparisons become problematic since analyses of what an educational provision is and what it means to its agents need to be related to local conditions (see Hedman and Magnusson 2020, for a further discussion).

Below, the design of SSL is further outlined, which is relatively unique from an international perspective, as well as its implementational problems. The aforementioned academic discourses surrounding the subject are also taken into account (see Hedman and Magnusson 2018), and are used to analyze the teachers’ discourses.

The school subject Swedish as a second language

The school subject SSL was introduced in 1995 as a parallel core subject to the Swedish subject (SWE) in school years 1–12. At present, the requirements to become a formally qualified SSL teacher in upper secondary school are 90 ECTS credits (Hedman and Magnusson 2018, 2019, 2020), that is, in addition to a graduate exam as a teacher. As a field of research, SSL can be further studied up to the level of PhD. In comprehensive schooling, SSL and SWE have the most lesson hours in the curriculum. An advanced academic content is guaranteed through more or less identical final goals in SSL and SWE. While the SWE and SSL syllabi both emphasize functional language competence (Swedish National Agency for Education 2013), one difference is the lower demand for “language correctness” in SSL. In upper secondary school, both subjects include literature, although historical perspectives on literature and language as well as literary analysis are more prevalent in SWE. SSL also includes an orientation towards the students’ multilingualism, including contrastive analyses between their “mother tongue” and other languages (Hedman and Magnusson 2019). In line with the management by the objectives of Swedish schools, syllabi are general and provide neither guidelines regarding methods of instruction nor detailed teaching content (Nusche et al. 2011). Teachers thus enjoy a high level of freedom in framing their subject, and realizations vary between schools and classrooms.

The many implementation problems from which SSL has suffered since its introduction have been attributed to its status as a relatively new subject, but also to its lower prestige and marginalized position in schools (Lindberg 2009; Lindberg and Hyltenstam 2013; see also Hedman and Magnusson 2018, 2020). A major issue is the small share of qualified SSL teachers in Swedish schools (42% of the teachers teaching SSL in upper secondary school; Swedish National Agency for Education 2019). This means that many of the teachers teaching SSL do not have appropriate qualifications, which may partly explain low instructional quality, low academic standards, and low expectations of the students found in a few case studies on the subject (e.g., Fridlund 2011; cf. Hedman and Magnusson 2018), contrary to the high requirements presented in the syllabi. At the same time, SSL, owing to its high goals, has been criticized for building on a monolingual norm, which discriminates against students learning at beginner levels (Elmeroth 2006). Moreover, there are reports (e.g., Swedish National Agency for Education 2017) on students in comprehensive schools being referred to SSL without clear bases for doing so; in upper secondary schools, students themselves decide whether to study SSL or SWE.

Discourses and conflicts in relation to SSL

In the previous study on SSL (Hedman and Magnusson 2018), in which discourses on SSL within the academic field were scrutinized in relation to the notion of equality, we identified two central and entwined discourses. In the first, “SSL discourse”, which is based on equality as equal opportunity (Westling Allodi 2007), the adaptation to the language educational needs of multilingual students learning in a second language is vital. Its proponents, who are often researchers in multilingualism and second language learning, argue for a long-term perspective on L2 development, as opposed to layman attitudes towards L2 learning as fast and implicit (e.g., Lindberg 2008, 2009; Lindberg and Hyltenstam 2013). At the same time, they critique the school system for its prevailing monolingual norm and “difference blindness”, the latter here referring to a neglect of linguistic diversity and an insensitivity to varied language learning needs among students. In the second, “inclusion discourse”, the physical separation of students in SSL is identified as a main problem per se, regardless of language learning aspects (equality as uniformity, Westling Allodi 2007). Here, SSL is viewed as entailing a deficiency perspective of students that leads to exclusion and stigmatization (Hedman and Magnusson 2018, p. 17; see also Bunar 2010), which has motivated some to suggest the subject be discontinued.

Hedman and Magnusson (2018) found, however, the empirical underpinnings of both discourses to be scarce, in particular with regard to educational practices and the perspectives of its stakeholders. Two follow-up studies, based in the same three linguistically and culturally diverse upper secondary schools as in the current paper, focused on the perspectives of 15 students in their last grade, regarding their motives for choosing and staying in SSL, given that these students decide for themselves which subject to study, and their experiences after doing so (Hedman and Magnusson 2020). The findings were reflective of the above-mentioned discourses, with regard to an ambivalence towards SSL among students in relation to their choice of SSL, connotations of pejorative societal discourses on immigration and L2 use, as well as SSL as stigmatizing. On the other, counter-images appeared, and the pedagogical scaffolding of advanced content as well as the multilingual dimensions of the curriculum stood out as main reasons for choosing and staying in SSL (see also similar findings in Bjuhr 2019, an interview study with upper secondary school students in SSL). Further, differences between schools were found in the extent to which they opened implementational space (Hornberger 2002) for multilingualism (Hedman and Magnusson 2019).

The discourses and conflicts in relation to SSL, highlight a range of assumptions and ideas about SSL and its raison d’être to which SSL teachers may be exposed and amidst which they navigate. Along with the fact that SSL is a relatively new subject lacking long traditions of practice, and the relative freedom of the curricula, this creates an impetus for studying the role of SSL from a teacher perspective (see, e.g., Creese 2005; also Slembrouck 2010, on the importance of considering local perspectives). This has, we argue, relevance both as an example of an educational provision for multilingual students and of teachers’ agency in contested and debated school contexts, and, as here, a specific school subject.

Aim and theoretical points of departure

We thus aim to investigate how three experienced SSL teachers in three linguistically diverse upper secondary schools discursively construct the SSL subject, and whether and how they legitimate their role as SSL teachers. The three questions are:

• How are students in SSL discursively constructed by the teachers?

• What acts of legitimation, if any, can be discerned in teachers’ discourses on SSL and its students?

• To what extent and how do teachers (re)produce and/or contest previously outlined academic and public discourses on SSL and its students?

According to Abdi and Basarati (2017, p. 88), “the most prominent act of legitimation is to discursively construct, justify, and institutionalize certain institutional facts, values, and orders” (cf. van Leeuwen 2007; also van Dijk 2008). The analysis builds on van Leeuwen’s acts of legitimation (2007, 2008). These are answers to the questions “why should we do this” and/or “how should it be done?” (van Leeuwen 2007, p. 93) within “the practices of specific institutional orders” (p. 92). We distinguish, first, authorization, referring to legitimation on the basis of laws and legislation, such as language and educational policy legislation, as well as ideology and tradition. Personal authority depends on investment in persons, e.g., when the authority and status of an individual serves as legitimation, that is, an implicit “because I/somebody say/s so” (van Leeuwen 2008, p. 106). Expert authority relies on the expertise of somebody, and role model authority refers to people who are presented as good examples worth following. The second category, moral evaluation, reflects a certain order of (moral) values, signaled by qualifiers such as “natural” or “normal” (van Leeuwen 2008, p. 110), under which linger a “submerged iceberg of moral values” (ibid.). Instrumental rationalization, finally, relies on “goals, uses, and effects” (van Leeuwen 2008, p. 113) of practices, ultimately resting on utilitarianism and pragmatism, whereas theoretical rationalization “legitimizes practices by reference to a natural order of things” (ibid.). We leave out mythopoesis on the basis of being inapplicable. Mythopoesis refers to legitimation through storytelling, e.g. moral and cautionary tales.

In our endeavor to consider teacher perspectives, analyses of such possible acts of legitimation are related to the teachers’ enactment of policy in relation to the content and aims of SSL. Singh, Heimans, and Glasswell (2014) define policy enactment as comprising the processes of “policy interpretation” and “translation by a diverse range of policy actors across a wide variety of situations and practices” (p. 826f.). Official policies are, thus, enacted in complex ways and at various levels, whereby teachers are “policy subjects” who do not implement or read policy as something “pre-determined” (Ball et al. 2011, p. 612). Teachers’ space for enacting policy is relational and context-dependent, as well as shaped by individual’s experiences and professional history (Pyhältö, Pietarinen, and Soini 2014). It may imply both embracing change and rejecting it (ibid.). The role of teachers as policy subjects is also most likely to be reinforced when teachers can interpret a loosely framed curriculum in relatively free ways (Sahlberg 2010).

Design of the study

The data included SSL lesson observations with fieldnotes and audio recordings as well as interviews with each teacher and fieldnotes from teacher-related topics that emerged in informal conversations with the teacher (written after the conversations).

Observations in the teachers’ SSL classrooms

The ethnographic fieldwork took place in three schools over the course of one and a half years and encompasses observations of 58 lessons in SSL. The schools were situated in suburban areas with low SES, in particular with regard to Pine and Birch, which also had the largest proportion of foreign-born students. Although Rowan was situated in a less unprivileged area regarding SES, many of its students still came from suburban areas with low SES (Hedman and Magnusson 2019, 2020).

We observed the teaching of mainly final grade students (grade 3, in Rowan and Pine), but also grade 2 students (Rowan and Birch), as well as teaching in so-called introductory classes (Birch), i.e., beginner programs for recently arrived migrant students. The SSL classes comprised approximately 20–25 students. During observations, we took fieldnotes, and did not use a specific protocol or checklist. Although the interviews are the focus for analysis in this article, fieldnotes and audio-recordings from the classroom observations influenced the topics discussed in the interviews, and the way we interpreted interview data.

The SSL teachers

The three teachers were formally qualified to teach SSL in upper secondary school: Sandra in Pine, Wera in Birch and Stephen in Rowan (all pseudonyms; the same teachers were referred to in Hedman and Magnusson 2019, 2020). Sandra and Stephen were also qualified to teach SWE at the same level, and Wera to teach another language subject. They had substantial experience in teaching SSL, at least five years, predominantly in linguistically diverse schools.

The teachers were interviewed once individually by both researchers in a semi-structured audio-recorded interview that exceeded more than one hour. These were carried out in the schools part way through the observed years at Birch, and after the observations at Pine and Rowan were completed, as we wished to have observed a substantial portion of their teaching before conducting the interview. The interviews centered around (1) the students, with regard to their language learning and language practices, and (2) the instruction planning, in relation to the syllabus, goals, and most important content, as well as the perceived differences between SSL and SWE. Typical prompts for section (1) were How much do you know about your students in terms of language proficiency when they begin SSL? How do students receive information on SSL? Do you relate to the students’ previous experiences and knowledge in other subjects? If so, how? Do students leave SSL for SWE? For what reasons? Typical prompts for section (2) were What is the goal of your teaching? What do you want your students to achieve primarily? What promotes language development, through what activities? How do you deal with student heterogeneity regarding language proficiency? Do you collaborate with other subject teachers? The Swedish teacher? If so, on what themes?

The three audio-recorded interviews were all transcribed, and the extracts in this paper were later translated from Swedish to English by the authors.

The analysis process

In the qualitative analysis process, the data was analyzed in an abductive manner (Alvesson and Sköldberg 1994), oscillating between data and theory in systematic ways (cf. the notion of trustworthiness in interpretive research (Schwartz-Shea and Yanow 2011). In our initial readings of the data, both authors analyzed all data separately and then together. The expressed characteristics attributed to students by the teachers were interpreted as their discursive constructions of the students. These were, for example, realized by the use of adjectives, such as “the students are disadvantaged”, or in processes describing what students generally, or what specific students, do or say, according to the teachers, e.g. “the freshmen … might feel that they shouldn’t study SSL”. These discursive constructions were further interpreted by us to function as legitimations in the form of “the way things are” (van Leeuwen 2007, p. 103). Our analyses of teachers’ discursive constructions of the students in SSL are discussed separately (see Findings), although these are linked to the theoretically induced categories of legitimation outlined below. Discursive student constructions were analyzed through the lens of discourses on SSL as referred to above (Hedman and Magnusson 2018) in terms of reproducing or contesting such discourses, and in relation to the goals of SSL and the teachers’ role as SSL teachers. Also, as in Hedman and Magnusson (2020), the analyses thus draw on a language ideological frame (Hult 2012) and the recognition of linguistic hierarchies, which may be reflected in status differences between languages in a societal context, and between first and second language speakers.

With an openness to data (see Schwartz-Shea and Yanow 2011), we also focused on possible acts of legitimation, as defined by van Leeuwen (2007, 2008), while coding the data, although not in a checklist manner. The initial analytical process of legitimations started with a bottom-up analysis of the data, to which the predefined categories as formulated by van Leeuwen were analyzed, although in a dynamic way between data and theory. We were sensitive to answers to the questions “Why should we [teachers] do this?” and/or “How should it be done?” (van Leeuwen 2007, p. 93), e.g., when teachers in our interpretation not only described what they did, but also why they did it. We were also sensitive to instances of “unspoken why” (van Leeuwen 2008, p. 94) and to our own role in the interaction (cf. Talmy 2011).

These analyses went beyond those of van Leeuwen (2007, 2008), who often analyzes legitimation types in single sentences, typically identified in certain lexicogrammatical realizations, for example, when pointing to the function of verbal processes to express impersonal and personal authority, or to the vital role of nouns such as law and regulation for impersonal authority legitimation (van Leeuwen 2007, p. 96; see Halliday 1985). In our analyses, we extended the analytical categories to include longer stretches of speech. These were thus not identified through lexicogrammatical realizations, but were interpreted as having legitimizing functions in accordance with van Leeuwen’s model (2008), such as Role model authority and moral evaluation, Expert authority and Instrumental and theoretical rationalization (see Findings). For example, a teacher’s reference to her personal experiences of learning Swedish as an additional language was, in this particular context, understood as a legitimation of her aptitude as an SSL teacher and of her competence for understanding the students’ lived experiences of language (Busch 2017), indicating a discursively constructed role model authority. Likewise, a teacher’s narration of how other teachers seek her advice on issues of second language development was interpreted as a legitimation of her role as an SSL teacher through expertise or expert authority. Although we were able to collaboratively identify the above-mentioned legitimation functions, not all categories in the model were found to be relevant. For example, the category of mythopoesis was not applicable.

Findings

In this section, the analyses are presented and discussed as Discursive constructions of the students in SSL, Role model authority and moral evaluation, Expert authority and Instrumental and theoretical rationalization based in the analytical procedure outlined above.

Discursive constructions of the students in SSL

Two of the teachers, Sandra and Wera, were found to construct their students in SSL as unprivileged. Sandra explicitly addressed her students as unprivileged and at a disadvantage in school and in Swedish society at large (excerpt 1). She characterized the students’ language and world knowledge in negative terms as “not good enough language”, that they “lack knowledge of the world” in line with the stance of Appreciation and Judgement (Martin and White 2005), and the students as mentally and physically confined to the area of their residence. Outlining language as a key factor for societal success, she expressed her ambition to help students to expand their repertoire of registers, “to know how to behave orally and in writing in different situations”, in order to be able to compete with others. She maintained that students born in Sweden in her classes also lacked these competitive language competencies (excerpt 1).

Excerpt 1. Teacher Sandra at Pine.

Sandra: I think, in today’s society, we are often judged on the basis of how we talk and write, and how that may close many doors, not being able to manage different ways of writing and speaking. I believe that in the context in which I work, they [the students] are already disadvantaged, and I know that there are great shortcomings in the language, although you are born here, grew up here […] I want them to be able to compete with all others and be as prepared for future studies and work life [as others].

Researcher: Why are they disadvantaged?

Sandra: I simply don’t think they have good enough language skills. But not just that they lack knowledge of the world, and of the labor market, and what it’s like in other schools. I find that many [from this municipality] tend to stay here. Sometimes, there are students who know [name of the shopping street in a nearby larger city], but, well, they don’t really have a clue about the surrounding world […] So it might be then, to know how to behave orally and in writing in different situations and contexts. I think that they should be prepared for life. (interview)

Sandra’s characterizations of her students in SSL also applied to students enrolled in SWE at Pine (excerpt 2). The majority of the SWE students were also multilingual but were, according to Sandra, in an even more unfortunate position. She contrasted them to the newly arrived students in SSL whom she found to be better prepared for schooling and who were highly motivated to achieve. The learning situation within SSL was thus in fact better than in SWE, according to Sandra.

Excerpt 2. Teacher Sandra at Pine.

Sandra: The majority of those who study ordinary Swedish are almost even more disadvantaged.

Researcher: In what way?

Sandra: They, um, many of my newly arrived students, they come from well-educated families. Some have already started upper secondary schooling in their countries of origin. They well, already keep track of the world, they read newspapers, keep up a bit and so, um, that’s how they are brought up, they have grown up this way, and they have parents who are academics. Many of those who were born here don’t have that, and their parents have also only stayed here. (interview)

By this twist, Sandra partly deconstructed a general image of SSL as a subject for low-achieving students (e.g., Fridlund 2011) in ascribing these types of negative attributes to local norms and a school culture at large that devalue hard study, and not to the SSL students per se. Rather, she presented SWE as a subject in which students do not get the teaching they need, thus overturning the conceptions of SWE as a better alternative, compared to SSL.

All three teachers focused on students’ language needs in the SSL classrooms, in terms of expanding linguistic repertoires. They stressed that students had to acquire a wide array of linguistic norms for their future success; otherwise, they would not achieve “the language they needed” (interview with Wera). Hence, all of the three teachers legitimized SSL with language-based motives, and, in this sense, they reconstructed the SSL discourse referred to above (see Hedman and Magnusson 2018). However, in comparison to Wera and Sandra, Stephen at Rowan did not talk much about his students’ socioeconomic backgrounds, but rather emphasized the students’ academic motivation and agency. Rowan was located in a more wealthy area and attracted students from suburban contexts with low SES, and, when asked, Stephen pointed out that the students’ active choice of going to a school further away might imply a higher degree of academic motivation.

Role model authority and moral evaluation

A role model authority was found to be related to the teachers’ overt discursive constructions of the students in SSL, as outlined above. As noted by van Leeuwen (2007), in authorization, the question of why something should be done is answered by “because I say so” (p. 94), and the kind of authority bestowed on a role model rests on the notion of a good example.

Sandra’s image of the SSL students, who “should be prepared for life” (excerpt 1), motivated her as a teacher in her claims of having the ability and being well-suited to relate to socioeconomically unprivileged students from non-academic families. In the classroom, this discourse was particularly prominent in a theme on language variation, in which Sandra spoke of her own origins in a similar multilingual and multiethnic suburban area, and of her transfer to a highly prestigious upper secondary inner-city school. She shared with the students her experiences of having to adjust to new socioeconomically coded behavior, involving new language practices, and by doing so approached the students on a personal level, in which, we argue, she also presented her authority as a role model. Sandra’s narration is also an expression of moral evaluation (van Leeuwen 2008, p. 110). It links to a discourse on life success submerged with a particular set of moral values (van Leeuwen 2008, p. 110), which relates, for example, to “the moral good” of getting acquainted with other – certain sets of – social contexts for better future prospects, and the unwanted behavior of only staying in the suburbs.

Wera expressed a role model authority as a migrant and as a learner of Swedish herself. She used multilingualism as a means for creating an “inclusive us” in the classroom and for empowering the students (see Hedman and Magnusson 2019), and thus was able, unlike Sandra, to claim a personal understanding of the situation of her students as migrants and learners of Swedish. She included issues of language and power in describing her students, exemplifying with students who positioned themselves as speakers of “immigrant Swedish” (Swe. “blattesvenska”, often used pejoratively). Wera acknowledged the students’ whole linguistic repertoire, this variety included, but urged them to be aware of how certain language choices ascribe them a certain set of identity positions on the basis of language ideological dimensions in this particular socio-political context (cf. Leeman 2018). In excerpt 3, Wera addressed a certain resistance in the students towards linguistic hierarchies, and the possible disparagement of their first learned languages for the benefit of Swedish, which was a feeling she could relate to from her own experiences (see excerpt 4). She, however, strongly emphasized Swedish as the language of power, leaving no option not to learn the language, “you have no choice”, but having to “own” the [Swedish] language for participatory citizenship (excerpt 3). Her authority here related to the positioning of her students as being in need of empowerment and consciousness raising, “that language is power”, in line with a critical language awareness approach (Leeman 2018).

Excerpt 3. Teacher Wera at Birch.

Well, I don’t know, in discussions with the students, I think I tell them that “you don’t have a choice, you have to study Swedish”. And that language is power. I tell them that you can choose not to speak Swedish but that you should own the language, you should own the language, and then you can refuse, if you like. (interview)

The embeddedness of this teacher discourse in her own lived experience (Busch 2017) of learning Swedish reinforced her legitimacy as SSL teacher and her authority as role model, as also apparent in excerpt 4. Again, in excerpt 4, we see how her work on consciousness raising and critical language awareness was embodied in her own lived experiences of language, and that she claimed similar feelings in an effort to be visible in her students. More than the other teachers, Wera claimed to understand feelings of resistance towards linguistic hierarchies, which she had also experienced. In excerpt 4, Wera shared her personal experience of typically being positioned differently as a French speaker – more well-educated and interesting – than as a speaker of her, in this context, less prestigious first learned language.

Excerpt 4. Teacher Wera at Birch.

But every time I hear that I’m good at French today, now when I’m older, that’s not something I really appreciate. It’s more like “well, they think French is much better than, for example, [language x]”. Or if people say like, well, they take a lot of interest in me because I’ve been to a French school. (interview)

Although Wera discussed her students in SSL as being unprivileged in Swedish society at large, she emphasized their responsibility for their own learning and future prospects. In excerpt 5, she even contested a student discourse on possible discrimination on the labor market, “don’t say that you didn’t get the job because you’re not Swedish”, and insisted on the students’ own possibilities for making their way, by applying a metaphor of “opening that door”.

Excerpt 5. Teacher Wera at Birch.

Wera: It’s just to bring it up and say that it [the upper secondary school] is not compulsory. But if you don’t go to upper secondary, what happens, what choice do you have?

Researcher: That’s right.

Wera: Choose, but you can’t blame anyone else, for example,” but I’m not Swedish.” Okay, okay, but don’t say that you didn’t get the job because you’re not Swedish, you exclude yourself. Maybe the door is closed but it is possible to open it. It is possible to open that door. But if you just say that you’re standing outside and say “well I can’t get in”, it’s up to you: just to push [the door] and walk in. (interview)

To different extents, all teachers recognized how some students contested the social categorization of “SSL student”, yielding an ambivalence towards SSL, particularly in students born in Sweden. Here, Wera, more than the other teachers, reconstructed the “inclusion discourse”. She related to her students’ own experiences in her contestation of who was to study SSL, as she called into question the fact that students born in Sweden do. She rhetorically asked: “When will they then ever be perceived as real Swedes?” (interview), relating to “value systems” and moral evaluation (van Leeuwen 2008, p. 106) on who is to be considered a Swede. According to Wera, it is a failure on behalf of the Swedish school system that students born in Sweden study SSL: “you have marked out that they don’t belong here, I think, or it’s their experience” (interview). In line with this, some of Wera’s students in SSL who were born in Sweden, dissociated themselves from being in the same class as students who had only been in Sweden a short time, asking whether they were studying “SFI” (Swedish For Immigrants, i.e., beginner education for adult newcomers to Sweden; cf. Talmy 2009; see excerpt 6).

Excerpt 6. Teacher Wera at Birch.

Wera: As soon as you explain something grammatical, despite that they [the students born in Sweden] need it, they tune out, or so, say: “is this SFI or?”

Researcher: A well, is it… they say so.

Wera: Yes […]

Researcher: And what do they mean by, um, SFI?

Wera: That you don’t know Swedish. (interview)

Wera’s questioning of the separation of students born in Sweden in SSL and SWE discloses a possible ethical dilemma entangled with being an SSL teacher, that is, teaching students who one does not think should be there. Her critique was directed mainly to the students’ previous schooling in compulsory school. However, as students in upper secondary school choose between SSL and SWE, all teachers informed the students explicitly that they were allowed to change from SSL to SWE if they wanted to, which also happened, in both directions, according to the teachers, although not during our fieldwork (Hedman and Magnusson 2020).

To sum, Sandra’s and Wera’s analyzed role model authority was based on their lived experiences of language and education, and on their constructions of the students, as well as of the SSL subject as an ethical project (see Felder 2018, p. 67) embedded in a democratic professionalism (Sachs 2001; also Hedman and Magnusson 2020). Expressions of role model authority were not salient in Stephen’s classroom discourse, nor in interviews with him. A democratic professionalism was, however, also part of Stephen’s teacher identity in the sense that he, like Sandra and Wera, emphasized equality in terms of equal epistemic access (Morrow 2007) to the language and literacy of power. This was salient through legitimations of an expert authority, which, in Stephen’s case, was related also to his role as a SWE teacher, which will be further discussed in the following section.

Expert authority

Contrary to discourses on SSL as a low prestige subject, the teachers asserted high local prestige. A possible reason may be the fact that it was not a minority subject in any of the schools, but catered to as many or more students than SWE did (cf. Hedman and Magnusson 2018, 2019, 2020).

The SSL teachers performed acts of legitimation in relation to their expert authority, broadly encompassing knowledge about language development in, and learning through, an additional language, as well as expertise in pedagogical scaffolding. The teachers stressed the impact of their teaching in SSL on students’ Swedish language and literacy development, implicitly through their expertise as SSL teachers. For example, Sandra confidently told us in the interview that those of her students who attended her classes did not fail the Swedish national tests, whereas those who skipped her classes, for various reasons, did fail. In this sense, she maintained expert authority, and also downplayed the role of uninstructed Swedish language learning outside of school as a means to pass this type of high stakes test (see Lindberg 2002 on myths regarding second language learning).

All three teachers emphasized advanced academic content in their teaching. This emphasis implies, we argue, strong legitimizing power, both of SSL as challenging and meaningful, equal to SWE, and of their expert authority in pedagogical scaffolding to keep these high educational standards. Sandra even said that she challenged the SSL students academically more than other teachers did.

Stephen, however, regretted the bias towards functional perspectives on fiction in SSL – “I think it’s sad that the SSL courses value literature only, well, functionally” – and, also, that the requirements for literary analyses are lower in SSL than in SWE: “There [in the literary reading in SSL] they are not challenged and I find that a bit odd. There [in the SSL syllabus] it could be, well the same type of content [as in SWE]”. In our fieldwork, we noted that Stephen presented more literary analyses in SSL, which was motivated on the basis of equal epistemic access to literary content, as well as by drawing on his expert authority as a SWE teacher: “[w]ell this is more my own, I think, as I also teach Swedish courses”. This act of teacher agency was thus based on, we argue, a democratic professionalism, where the high academic standards of SSL stood out as a key tenet. For example, addressing the above-mentioned initial reluctance towards SSL shown by some students (see also Hedman and Magnusson 2020), Stephen retold how he legitimated the subject by its advanced content (excerpt 7).

Excerpt 7: Teacher Stephen at Rowan.

Stephen: Well it’s the freshmen who are a bit unsure and might feel that” I shouldn’t study SSL, it’s worth less than Swedish”, but in the 11th and 12th year, that’s rarely heard.

Researcher: Interesting.

Stephen: And I think it’s important, well, that it’s not a stigma to choose Swedish as a second language, but that you can choose it for its challenging instruction […] also there.

Researcher: Hm.

Stephen: And I think they have got that, and that’s why I can say that they choose to stay because they notice that this isn’t easier than something else, some light version of Swedish to get better grades, but that they in fact are challenged and develop. (interview)

Apart from claiming expert authority, this was also a kind of rationalization based on the way things are (cf. theoretical rationalization below). SSL is thus legitimized, in contrast to its being perceived as a low-demanding second language subject and assumptions about SSL as the less prestigious alternative. Likewise, Sandra clarified to us that students at the upper secondary school level realize that it is not easier to study SSL as compared to SWE, as the final requirements in grade three in SSL are “really high […] everyone is supposed to continue to the university afterwards”. As part of her expert authority in pedagogical scaffolding, Sandra also referred to her teaching style as student-centered and relationally oriented, which was confirmed by our fieldwork. In Hedman and Magnusson (2020), her students expressed how much they appreciated her pedagogical scaffolding and negotiations of meaning that helped them understand. Sandra, however, explicitly said that she did not want to provide her students with answers but with pedagogical tools that helped them arrive at their own conclusion independently (excerpt 8).

Excerpt 8: Teacher Sandra at Pine.

Well, I don’t always have the attitude that I will need to explain this in different ways. I don’t want to give away any answers, but want them to find out for themselves. Then you might explain something in ten different ways. (interview)

The SSL teachers’ expert authority also became apparent in relation to other teachers in various collaborations. All three teachers valued language and content-based teaching across the curriculum as a means of enhancing learning in multilingual students. This type of collaborative work had been introduced in the schools through in-service training, in which the SSL teachers were perceived as “experts” on the teaching and learning of multilingual students, as evident in fieldwork and interviews. For example, Stephen had coordinated a project on reading comprehension across the curriculum, and Sandra, who held a position as a “first teacher”,² was regularly consulted by other teachers on the linguistic dimensions of their subjects. Wera expressed frustration with insufficient language and content-based approaches at Birch, but during our fieldwork she was constantly aware of what her students were working on in other subjects and was regularly consulted by other teachers.

Instrumental and theoretical rationalization

Finally, we discuss rationalization as an act of legitimation, where the factual goals and purpose of SSL are in focus, either in instrumental terms regarding “goal effects” of SSL, or in terms of theoretical rationalization, which refers to “the constructed, conventionalized, and perceived natural order” (Abdi and Basarati 2017, p. 90).

Stephen, as well as the other SSL teachers, frequently pointed out in class the goals of SSL in terms of access to the university (cf. epistemic access; Morrow 2007). All teachers emphasized the role of reading and oral comprehension as well as the production of academic texts in order to be able to enter tertiary education. In excerpt 9, Stephen particularly emphasized the role of text genre expertise, not only for future studies, but also for the students’ future careers.

Excerpt 9: Teacher Stephen at Rowan.

Researcher: So what would you say is most important for the students in SSL, what would you like to emphasize?

Stephen: Well, for the 12th year students, I think, well, reading is what I want to emphasize most of all.

Researcher: Mm.

Stephen: At all levels […] I think above all that the reading of factual texts is most important, as many will move on to the university. Then, to give them some tools to know how to treat this type of text. […] well, and the other thing is to treat different types, to write different text types so that they know what can be expected when they get a task in their future work life or studies, how something should be written. (interview)

The goal-oriented discourse in excerpt 9 reflects acts of instrumental rationalization, relating to the academic benefits of studying SSL (see also Bjuhr 2019). The recurrent theme of the high requirements in SSL in the data, particularly in terms of academic language, was a theorized rationale for SSL for all three teachers. In excerpt 10, Sandra draws on the underlying rationale of advanced academic content and language to explain a non-passing grade given to a student born in Sweden, who neither accepted that his Swedish would not fulfil the requirements, nor that newly arrived students could achieve higher grades than him.

Excerpt 10: Teacher Sandra at Pine.

Sandra: I talked to a student who failed his essay and oral presentation, and he couldn’t really accept that because he’s born in Sweden, he can talk. But he can’t understand that he uses everyday language in a formal text. He doesn’t buy that, so to speak, and he compares himself with another student who is rather quiet and has been in Sweden for six years, but still gets better grades than he does.

Researcher: Mm.

Sandra: He can’t understand this.

Researcher: What does he say about it?

Sandra: He thinks it’s strange.

Researcher: Mm.

Sandra: Without knowing what [the student’s] grade is, he doesn’t have any background information at all, but he and many others have some difficulty understanding the requirements.

Researcher: Mm.

Sandra: And then he might think, “well but he studies SSL”.

Researcher: But he knows Swedish.

Sandra: He knows Swedish, but it’s not enough. It’s not enough to speak English in an English course, either […] or, well, to calculate plus and minus in math. (interview)

In excerpt 10, Sandra clarifies why she thought some students born in Sweden need to study SSL, on the basis of the rationale saying that nobody’s first language is academic language proficiency (Gibbons 2006). This theorized rationale is here explained as a “perceived natural order” by Sandra, in line with how “[t]heoretical rationality legitimates practices by reference to a natural order of things” (van Leeuwen 2007, p. 101). Sandra did, however, not address the issue raised by Wera, who saw a risk for the SSL students born in Sweden to remain stigmatized as “non-Swedes” (see also Hedman and Magnusson 2020). In this sense, excerpt 10 illustrates the tension between the two academic discourses on SSL as outlined above, which here comes to a head.

Concluding discussion

The analyses of how the participants discursively construct the SSL subject and legitimate their role as SSL teachers indicate how teachers in general may navigate and legitimize a contested educational practice that they are co-constituting as policy agents at the local level. The analysis thus exemplifies how teachers enact agency, generated through their professional experiences (cf. Pyhältö, Pietarinen, and Soini 2014), by legitimizing their educational practices counter to impeaching discourses. To a great extent, the teachers contested previously outlined academic and public discourses on SSL as a low status and inferior subject taught by unqualified teachers to low-achieving students (e.g., Fridlund 2011; Swedish National Agency for Education 2017). Throughout the fieldwork, none of the teachers brought up the parallel organization of two Swedish subjects per se and/or the possible abolishment of SSL as a topic for discussion, which contrasted with the academic discourses previously referred to. The teachers legitimized the subject through their discursive constructions of students as favored by their instruction, and, predominantly, by foregrounding their expert authority. This teacher expertise was claimed in terms of pedagogical scaffolding qualities in a naturalized order of advanced literacy and academic content teaching. Previous analyses show that this teacher expertise was essential for why students chose to stay in SSL in the same schools (Hedman and Magnusson 2020; see also Bjuhr 2019). The expressed institutional order of advanced literacy thus provides contrastive insights (Hymes 1996) regarding SSL as a low status and inferior subject (cf. Hedman and Magnusson 2019, 2020), also in relation to other second language teaching practices, for example, those reported in Talmy (2009, 2015), with their prevailing order of unchallenging literacy teaching. This finding indicates the significance of taking the local stakeholder perspectives into account.

The claimed role model authority was embedded in a democratic professionalism or a democratic teacher stance “consistent with plural values associated with inclusion” (Felder 2018, p. 66). This democratic professionalism seemed to open space for teacher agency in the classroom, in line with the view of Singh, Heimans, and Glasswell (2014, p. 826) of policy enactment as a process of “policy interpretation”. Such policy interpretation may also entail policy frictions. One teacher’s rationale for why SSL comprised students with a migrant background born in Sweden – thus, with a questionable status as speakers of Swedish as a second language – was contested by another teacher as non-inclusive. The fact that the teachers reconstructed the SSL discourse, and also to some extent the inclusion discourse of possible stigmatization, points to the complex nature of education policy. This complexity surfaces in particular when analyzing language education policy discourses, as such discourses may cross, overlap, and coexist at all possible levels (cf. Fairclough and Fairclough 2012; Hedman and Magnusson 2019).

The findings are relevant to the field of language education, as well as education policy in other national contexts. This is particularly true for high stakes educational practices comprising migrant language learning students as “inherently tension-filled” (cf. Daugaard 2020). Such educational practices may become the eye of the storm, subjected to surrounding discourses on its value and legitimacy by academics, politicians, and others in the public sphere. Although the need for emic perspectives from whom it may concern within language education has been repeatedly emphasized (cf. Creese 2005; Slembrouck 2010), it is clear that such studies do not underpin public debate and high stakes policy making. This article contributes new knowledge by accounting for how teacher discourse and its legitimizing functions can be scrutinized in more detail to gain a better understanding of the complexity of educational policy. The article also provides a thorough account of what educational practice is under study. We argue that this type of visibilization of educational provisions and teacher voice contributes to a strengthened participative framing for teachers on issues concerning their own practices.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was supported by the Wellander Foundation.

Notes

1 The term second language (L2) is used in the Swedish education policy.

2 A first teacher is a head pedagogical coach in a subject at a school.