Attitudes and beliefs on multilingualism in education: voices from Sweden

ABSTRACT

Sweden is often commended for the inclusion of home languages in the formal education system: both mother tongue instruction (where a pupil’s home language is taught as an optional school subject) and study guidance (where a pupil is given content support in their home language or prior language of schooling) are offered. Still, while many national educational policies are supportive of multilingualism, their enactment on the ground is often problematic. The attitudes and beliefs of teacher educators, in-service teachers, and pre-service teachers are crucial here, yet few studies have investigated how these key actors in Sweden perceive their encounters with linguistic diversity. Furthermore, an understanding of the similarities and differences in the perspectives across these three cohorts is lacking. We have analysed interviews with five teacher educators, five in-service teachers, and eight pre-service teachers concerning their attitudes and beliefs on multilingualism. These interviews reveal orientations towards language and language use in teacher education and primary schools. Specifically, language is seen both as a problem and as a resource. Our results uncover tensions in the expressed attitudes and beliefs about multilingualism, as well as about multilingual pre-service teachers in teacher education and multilingual pupils in the Swedish school.

KEYWORDS:

• Swedish primary school
• teacher education
• multilingualism
• attitudes
• beliefs

Introduction

In Sweden, about one-third of all school pupils have a mother tongue other than Swedish (SNAE, 2021); in some schools, close to 100% of the pupils have mother tongues other than Swedish (Norberg Brorsson & Lainio, 2015). Although Swedish is the main medium of instruction, Swedish primary schools offer both mother tongue instruction, where a pupil’s home language is taught as an optional school subject, and study guidance in mother tongue, where a pupil is given content support in their home language or prior language of schooling during subject lessons or in separate groups (Ganuza & Hedman, 2015). Due to these measures supporting the inclusion of home languages in the formal education system, Sweden is often hailed as a role model for implementing the multilingual turn (Conteh & Meier, 2014; May, 2014, 2019). However, few studies have investigated how key policy actors in Swedish education perceive their encounters with linguistic diversity.

Teacher educators, in-service teachers, and pre-service teachers are crucial to the processes of policy enactment (Ball et al., 2012). In Sweden, they must interpret, translate, and respond to the Language Act and the Education Act, which both address language practices in the Swedish education system. The Language Act (2009) promotes and protects Swedish as the ‘principal’ language, while also acknowledging the multilingualism prevalent in Sweden with five officially recognised indigenous minority languages, Swedish Sign Language and approximately two hundred minority languages. The Language Act ensures the right for everyone to learn Swedish and foreign languages, and explicitly stipulates that those with a mother tongue other than Swedish ‘are to be given the opportunity to develop and use their mother tongue’ (2009: §14, 2; our translation). The Education Act (2010) specifies the intentions of the Language Act as (1) pupils’ right to receive education in the subject of Mother Tongue Instruction (MTI) (other than Swedish) and as (2) pupils’ right to receive study guidance in mother tongue (SGMT). Since 1995, the subject Swedish as a Second Language (SSL) has also been offered to multilingual pupils as an alternative to the subject of Swedish.

MTI is a subject with its own syllabus in the national curriculum, offering pupils at all levels of schooling language instruction in home languages other than Swedish. Nearly 29% of pupils in the Swedish compulsory school (Years 0–9, ages 6–15) were eligible for MTI in the 2020/21 academic year (SNAE, 2021, p. 12), and approximately 60% of these pupils studied MTI that year. Arabic was the most common language for both pupil eligibility (approximately 80,500 pupils) and participation (approximately 58,700) in MTI, with Somali coming second (21,900 eligible and almost 17,200 participating). SSL also has a separate syllabus in the national curriculum. Approximately 143,000 pupils (13% of all pupils in compulsory school) participated in SSL in 2020/21 (SNAE, 2021, p. 13). SGMT, on the other hand, is not a school subject in its own right. It is a form of temporary support in the pupil’s mother tongue or prior school language that the principal of the school can offer to pupils assessed as unable to follow mainstream, Swedish-medium instruction in different school subjects. Most multilingual pupils are educated within mainstream education; introductory classes for newly arrived pupils can be arranged, but only as transitory classes focusing on Swedish language skills.

Swedish primary teacher education is organised federally, with the Master of Arts in Primary Education (a four-year higher education degree) covering areas such as teacher professionalism, curriculum theory, assessment, leadership, conflict resolution, and the history of Swedish education. While Swedish, English, and Mathematics are key components to the teaching degree, training in multilingualism is not a required component (Paulsrud & Lundberg, 2021; Paulsrud & Zilliacus, 2018), with no common measures in place to prepare pre-service teachers for working with linguistic diversity in the classroom. Furthermore, teacher training courses that may prepare for teaching in linguistically heterogeneous classrooms are mainly part of Swedish as a second language studies, and not currently an obligatory part of the teaching degree.

Although Sweden supports the multilingual individual on an official level, a monolingual ideology prevails both in society (Hedman & Rosén, 2021; Vuorsola, 2019) and, more specifically, in the school (Rosén, 2017). The Curriculum for the compulsory school, preschool class and school-age educare (SNAE, 2018, the version in use at the time of our study) does not clearly represent Sweden’s long experience and expertise with MTI, SGMT, and SSL, despite the presence of about 180 languages in Swedish schools (SNAE, 2020). A monolingual ideology for a multilingual school is evident in the national curriculum, exemplified by implicit language hierarchies and little explicit acknowledgement of the increasing linguistic diversity in today’s schools (Paulsrud et al., 2020). Without clearly allocated space in the official curriculum, practices that promote and support linguistic diversity in the classroom become the responsibility of individual teachers as well as teacher educators. Hence, although some official measures are supportive of multilingualism, the monolingual mindset shines through in legally binding documents such as the national curriculum and the teacher education guidelines, creating difficulties for key policy actors as they navigate these potential tensions.

In this paper, we explore the views on multilingualism of five teacher educators, five in-service teachers, and eight pre-service teachers. We specifically focus on their attitudes and beliefs vis-à-vis multilingualism, with the following research questions:

1. Which attitudes and beliefs on multilingualism do the participating Swedish teacher educators, in-service teachers, and pre-service teachers express?

2. What similarities and differences can be identified across and between the three cohorts?

Our main aim is to uncover an understanding of the similarities and differences in the perspectives across the three groups. Before reporting the results of our study, we first present the theoretical underpinnings and key previous research.

Language ideologies, hierarchies, and orientations

A monolingual ideology, or a monolingual mindset (Ellis et al., 2010; see also Eisenchlas et al., 2015; Schalley, 2020), is characterised by the devaluation of multilingual practices and the promotion of monolingual practices. The values associated with different languages in a society can be conceived of as a hierarchy, often visualised as a pyramid (Piller, 2016). The national languages are found on top of the pyramid, as they are the most valued, followed by languages taught as foreign languages at school. At the bottom, representing the lowest valued languages, we find different kinds of ‘community languages: autochthonous, indigenous, immigrant, varieties and ethnolects’ (Ellis et al., 2010, p. 442). According to Piller, linguistic hierarchies are maintained by locating diversity on the ‘lower rungs of the pyramid’ (2016, p. 18). This holds true for Sweden, too. The role of the Swedish language as key to Swedish society is emphasised by policies (e.g. Language Act, 2009), as well as among teachers (Lundberg, 2019) and by the general public (Hedman & Rosén, 2021). Only English, highly valued and visible in society, closely follows the top-positioned national language Swedish (Hult, 2017).

Ruiz (1984) suggested, in analysing the ideological underpinnings of language policies, that we can describe ‘a complex of dispositions toward language and its role, and toward languages and their role in society’ (Ruiz, 1984, p. 16) as different orientations towards language(s). These are language-as-right, language-as-problem, and language-as-resource. The language-as-right orientation focuses on the civil rights of minoritized groups or individuals in society. The language-as-problem orientation imposes a deficit perspective of minoritized languages based on monolingual language ideology, and, in relation to language hierarchy, positions minoritized languages at the bottom. The language-as-resource orientation celebrates linguistic diversity and, in school contexts, recognises and encourages the use of pupils’ varying linguistic repertoires, thus echoing a multilingual turn in education. Ruiz’s orientations inform our analysis and discussion.

Attitudes and beliefs on multilingualism

Societal beliefs and policy orientations impact on the ‘on-the-ground’ contexts in which teachers are trained and carry out their work – i.e. teacher education and schools. Fives and Buehl (2016) argue that teachers’ attitudes and beliefs on multilingualism shape their practice and thereby influence educational outcomes, and that beliefs specifically ‘serve as helpful heuristics for teachers embedded in the complex, ever-changing contexts of classrooms and schools’ (p. 114). Given the two focus contexts of teacher education and schools, the attitudes and beliefs of teacher educators (as teachers in the teacher education context), in-service teachers (as teachers in the school context) and pre-service teachers (as a link between teacher education and the realities at school) are thus investigated in this paper. Attitudes include participants’ opinions and evaluations about the value of multilingualism in general, linguistic diversity in the classroom, and the role of linguistic resources in education (Pratkanis & Greenwald, 1989; Young, 2014). ‘Beliefs or assumptions about the nature of the world, its past history, and the way it ought to be’ (Martin & White, 2005, p. 95, our emphasis) – the knowledge base of attitudes (Pratkanis & Greenwald, 1989) – include how participants think multilingualism affects learning as well as what potential benefits or challenges they think linguistic diversity offers, i.e. whether multilingualism is seen as opening up possibilities in educational contexts or as a hindrance inhibiting learners. As argued by Pratkanis and Greenwald (1989), attitudes and beliefs predict social behaviour.

While several studies have investigated attitudes and beliefs of teacher educators, in-service teachers, or pre-service teachers both internationally (De Angelis, 2011; Gkaintartzi et al., 2015; Pulinx et al., 2017; Young, 2014, amongst others) and to a lesser extent nationally in Sweden (Lundberg, 2019; Paulsrud & Zilliacus, 2018), the attitudes and beliefs of these groups have not, to the best of our knowledge, been compared and contrasted. International research on teacher educators and, to some extent, pre-service teachers remains rather scarce (e.g. Flockton & Cunningham, 2021; Smith, 2004; Yang & Montgomery, 2013), although some studies have focused on the role teacher education may play in views of diversity (e.g. Halpern et al., 2021). Rannu et al. (2021) found that teachers with more recent education were more positive to multilingualism, suggesting the value of pre-service teacher training that reflects current trends in diversity.

In Sweden, ‘a Swedish linguistic norm is taken for granted’ in teacher education (Carlson, 2009, p. 46). Still, Lundberg (2019) maintains that in-service teachers’ attitudes and beliefs are rather positive towards multilingualism and multilingual pupils, although ‘sceptical views, often based on monolingual ideologies, are present and are likely to pose challenges for the implementation of pluralistic policies’ (p. 266) in schools. In their interview study conducted at four Swedish universities, Paulsrud and Zilliacus (2018) found that pre-service teachers do not feel properly prepared for teaching in the linguistically diverse classroom, indicating a deficiency in teacher education programmes. Teacher educators in turn express an awareness of the need for training in multilingualism, but also uncertainty and lack of consensus about where measures should be included in teacher education. Usually, the responsibility falls on teacher educators in languages, especially those in Swedish as a second language (Paulsrud & Lundberg, 2021; Paulsrud & Zilliacus, 2018).

Participants and data collection

In this study, we analyse semi-structured interviews with 18 participants: five teacher educators (all working with primary teacher education); five in-service teachers (all with some training in multilingualism; all working in schools with a high percentage of multilingual pupils, except IT04); and eight pre-service primary teachers (in three group interviews with two or three participants each, as the participants requested).¹ Table 1 presents an overview of the participants, with TE.01-TE.05 for the five teacher educators; IT.01- IT.05 for the five in-service teachers; and PT.01, PT.02, and PT.03 for the three groups of pre-service teachers.

Table 1. Overview of participants.

Interview cohort	Self-identified linguistic repertoire
Teacher educators
TE01 Mathematics	Swedish, English
TE02 Arts	Swedish, English
TE03 Ethnology	Swedish, English
TE04 Swedish	Swedish, Finnish, English (Migrant background)
TE05 Language education	Swedish, English
In-service teachers
IT01 Primary teacher for age 7 (Year 1)	Swedish, English, Sámi
IT02 Special education teacher for ages 6–12	Swedish
IT03 Primary teacher for age 7 (Year 1)	Swedish
IT04 Primary teacher for age 8 (Year 2)	Swedish
IT05 Primary teacher for age 9–10 (Years 3 and 4)	Swedish, Norwegian, Sámi (Migrant background)
Pre-service teachers
PT01: Group 1
PT01.1	Swedish, English
PT01.2	Swedish, Arabic, English (Migrant background)
PT01.3	Swedish, English
PT02: Group 2
PT02.1	Swedish, English, French
PT02.2	Swedish, English, Spanish
PT03: Group 3
PT03.1	Swedish, English (Migrant background)
PT03.2	Swedish, English
PT03.3	Swedish, English

Our aim was a diverse selection of participants. Despite the necessity of a sample of convenience, the participants do represent a broad geographical spread as well as varied backgrounds (e.g. languages spoken and migrant background). An equal gender distribution was not possible, given the overrepresentation of females in the education sphere, but was not deemed relevant for this study. Although we had both male and female participants, we use she for all, for expediency and anonymity.

The interviews focused on multilingualism generally and in education, although each cohort also had specific questions on their roles. For example, the teacher educators responded to questions about how their courses and chosen literature prepared pre-service teachers to work with multilingualism in the Swedish school. The pre-service teachers were asked to consider examples of how different courses or assignments, including their work placement in schools, increased their knowledge about multilingualism. Both these groups also described the linguistic diversity among students and staff at their respective universities. The in-service teachers were asked about their experiences with multilingualism in the classroom and their own teacher education.

The interviews were each about one hour and conducted in Swedish, either in person, at a school or university, or on Zoom. We adhered to the research ethics of the Swedish Research Council (2017), including providing information on the study prior to the interview, outlining the participants’ rights, and obtaining written informed consent of each interviewee.

Data analysis

Our analysis of the interviews was inspired by data-driven approaches and qualitative content analysis (Mayring, 2014), with an aim to move from the identification of semantic codes to the unpacking of latent themes revealing attitudes and beliefs (Braun & Clarke, 2006). The audio recordings of the interviews were transcribed in full and then analysed in the research software ATLAS.ti, a qualitative data analysis tool chosen for its features that allowed us to systematically explore both what attitudes and beliefs the participants expressed and how they did this.

We performed the following data analysis steps: First, we identified, mainly from previous research, 51 concepts related to attitudes and beliefs on multilingualism in the Swedish education context (e.g. diversity, mother tongue, norms, culture, curriculum), and auto-coded these. The auto-codes informed but did not restrict the next step of manual coding, which uncovered positive and negative attitudes and beliefs towards multilingualism and categorised them. For each manual code, we considered five analysis aspects: focus, topic, domain, attitude or belief, and polarity, as exemplified in Table 2.

Table 2. Manual coding.

Aspect coded	Example of manual coding multi.resources.school_bel_neg
Focus: General subject area (e.g. Swedish, English, multilingualism, national minority languages)	multi = multilingualism
Topic: Specific focal point (e.g. agency, deficit, diversity, hierarchy)	resources
Domain: Who or where? (e.g. the interlocutor, such as the multilingual pupil; the organisation, such as teacher education; or the physical location, such as the home)	school
Attitude or belief	bel = belief
Polarity (positive or negative)	neg = negative

The example of the manual code multi.resources.school_bel_neg denotes a negatively loaded utterance expressing a belief that schools are not equipped with sufficient resources for working with multilingual pupils.

Using combinations of the five aspects, 169 different manual code types were assigned to the data. Additionally, we coded manually for tension, to indicate instances where utterances indicated multilingualism-related tensions between cohorts, or between institutional contexts. The 113 manual codes for tension coincided with other coding (47 instances) or were single codes (66 instances). In the next step, we created 49 combined codes (called smart codes in ATLAS.ti), which allowed us to create code groups that comprised more than one auto-code and/or manual code from above. An overview of the total number of coded quotes for each cohort, including auto-codes, manual codes and smart codes, is shown in Table 3. We cross-validated the coding carefully, thereby demonstrating that we were very consistent in our analyses across coders.

Table 3. Overview of total number of coded quotes in ATLAS.ti.

Interview cohort	Teacher educators	In-service teachers	Pre-service teachers
Number of interviews	Five individual	Five individual	Three group
Number of total codes	2 873	2 391	1 221
Number of total codes on average per interview	575	478	407

Once the coding was complete, the 13 interviews were analysed both individually and in comparison. In a recursive process, we analysed the data openly. The various codes were analysed quantitatively to help us note patterns as an initial analysis. Our focus then turned to a qualitative analysis of the coded quotes across the three cohorts.

Results

Here we present our findings according to the positive and negative attitudes and beliefs expressed on multilingualism, as well as the tensions revealed between the three participant cohorts, and between teacher education and the realities of school. See the overview in Table 4, which is further elaborated on in the following subsections.

Table 4. Overview of total number of coded quotes in ATLAS.ti.

Manual codes	Interview cohort
	Teacher educators	In-service teachers	Pre-service teachers	Totals
positive attitudes	18	38	7	63
negative attitudes	20	8	5	33
positive beliefs	25	65	6	96
negative beliefs	37	52	8	97
Totals – beliefs and attitudes, both polarities	100	163	26	289
Tensions	59	40	14	113

Positive attitudes and beliefs

Generally, the in-service teachers were more positive about multilingualism than the teacher educators. There are differences across the cohorts regarding assigning benefits of individual multilingualism to school vs. university learners, i.e. to multilingual pupils and pre-service teachers, possibly partly indexing the density of contacts between the cohorts, as seen in Table 5.

Table 5. Overview of positive coding (TE = Teacher educators; IT = In-service teachers; PT = Pre-service teachers).

Manual codes	Interview cohort (percentage of respective category in brackets)
	TE	IT	PT	Totals
Totals positive attitudes	18	38	7	63
- about multilingual pupils	0 (0%)	19 (50%)	2 (28.6%)	21 (33.3%)
- about multilingual pre-service teachers	1 (5.6%)	0 (0%)	0 (0%)	1 (1.6%)
Totals positive beliefs	25	65	6	96
- about multilingual pupils	0 (0%)	26 (40%)	1 (16.7%)	27 (28.1%)
- about multilingual pre-service teachers	1 (4%)	0 (0)	1 (16.7%)	2 (2.1%)
Overall totals positive (attitudes and beliefs)	43	103	13	159
- about multilingual pupils	0 (0%)	45 (43.7%)	3 (23.1%)	48 (30.2%)
- about multilingual pre-service teachers	2 (4.7%)	0 (0%)	1 (7.7%)	3 (1.9%)
Positive attitudes as percentage of overall positive total	41.9%	36.9%	53.8%	39.6%
Positive beliefs as percentage of overall positive total	58.1%	63.1%	46.2%	60.4%

Looking more closely at the in-service teachers’ 45 positives codes about multilingual pupils, we discerned differences in what is considered positive. Only eight instances indicated a positive attitude of multilingualism as having value in itself. Notably, all eight of these coded quotes are from IT.05, who herself has a multilingual background and is currently teaching in a school offering a national minority language as one medium of instruction. When asked about pupils speaking their mother tongues in school, she emphatically responded: ‘Yes, they should absolutely do this. Also, at university and in upper secondary school. Their whole lives’. (All translations are our own.) Her strong positive attitude points to a clear language-as-resource orientation.

Conversely, most of the in-service teachers’ positive beliefs signposted an instrumental view of multilingualism as a vehicle to learn the majority language Swedish (11) or for learning in general (6). According to IT.03, ‘If you know your mother tongue, it is also easier to learn Swedish … when you know the basics there’. Likewise, IT.04 maintained that pupils who speak a mother tongue other than Swedish at home should receive mother tongue instruction in school, saying, ‘Absolutely. Super important. The better you know your mother tongue, the better your Swedish will be’. Both IT.03, who has had in-service training in pedagogical translanguaging (the purposeful use of all linguistic resources for learning, see Juvonen & Källkvist, 2021), and IT.05, who teaches in a national minority language, had strong views on the instrumental value of multilingualism for learning in general, especially literacy skills. IT.05 explained her view that all literacy must come from the mother tongue: ‘The foundation comes from there … [literacy] is based on their mother tongue’. Still, she expressed concern as well that her multilingual pupils would be expected to attain the same level of Swedish as the monolingual classmates they would later meet in secondary school, as the national minority school only offered instruction up to Year 6. IT.03 also emphasised another kind of instrumental value, translating for parents:

IT.03: Because often if it is just a parent and [he/she] should say something, the children can help and interpret a bit and translate and say what mom said […]. So then we also notice that the children are very important, to be able to help their parents.

Coded with both a negative belief about the parents’ lack of Swedish proficiency and a positive belief about the functions of the multilingual pupils’ resources, she gives voice to both a language-as-problem orientation (for the parents) and a language-as-resource orientation (for the pupils).

Pre-service teachers had less to say about multilingualism, which may be due to their lack of experience in the classroom. Two from PT.02, however, have clearly positive attitudes about multilingualism as having a value in itself, one stating, ‘Multilingualism is very interesting and exciting and it is great fun when a person, a person has several languages’. The other simply said, ‘Multilingualism is awesome’, adding that she also found it interesting and fun to work with multilingual pupils, as she had in her practical work experience in her teacher education. Only one quote was coded for positive belief about pre-service teachers’ own multilingualism in the three PT groups: one pre-service teacher in PT.01 described her own linguistic resource of knowing and studying Arabic (her mother tongue) and Swedish, as going ‘very well’ or, as she said, ‘[I have] both languages in my hand’.

Negative attitudes and beliefs

Notwithstanding the positive attitudes and beliefs described above, in-service teachers viewed multilingual pupils’ multilingualism as a challenge, and teacher educators pronounced a deficit view about multilingual pre-service teachers’ language capacity in the majority language Swedish (see Table 6). The teacher educators consistently voiced more negative than positive attitudes and beliefs, while the opposite is true for the in-service teachers. Pre-service teachers were more balanced in their expressions of attitudes and beliefs, with equal total numbers for positive and negative evaluations (13 each) (see Table 4). The most frequent codes for negative attitudes across the cohorts focused on the pupils and pre-service teachers (see Table 2 for an overview of manual coding), with six instances of multi.challenge.multiPup_att_neg [the multilingual pupil is a challenge] as well as six for Swe.deficit.multiStu_att_neg [a deficit perspective of the Swedish language proficiency of multilingual pre-service teacher/Student]. For negative beliefs, there were nine instances of multi.deficit.multiPup_bel_neg [a deficit perspective of how multilingualism pupils learn]. Thirteen instances of multi.content.teachEd_bel_neg focused on beliefs about how teacher education offers preparation for working with linguistic diversity in the classroom. Negative beliefs about teacher education were noted across all three cohorts (see more in the following section on tensions).

Table 6. Overview of negative coding (TE = Teacher educators; IT = In-service teachers; PT = Pre-service teachers).

Manual codes	Interview cohort (percentage of respective category in brackets)
	TE	IT	PT	Totals
Totals negative attitudes	20	8	5	33
- about multilingual pupils	2 (10%)	6 (75%)	0 (0%)	8 (24.2%)
- about multilingual pre-service teachers	8 (40%)	0 (0%)	0 (0%)	8 (24.2%)
Totals negative beliefs	37	52	8	97
- about multilingual pupils	6 (16.2%)	21 (40.4%)	1 (12.5%)	28 (28.9%)
- about multilingual pre-service teachers	9 (24.3%)	0 (0)	0 (0%)	9 (9.3%)
Overall totals negative (attitudes and beliefs)	57	60	13	130
- about multilingual pupils	8 (14%)	27 (44%)	1 (7.7%)	36 (27.7%)
- about multilingual pre-service teachers	17 (29.8%)	0 (0%)	0 (0%)	17 (13.1%)
Negative attitudes as percentage of overall negative total	35.1%	13.3%	38.5%	25.4%
Negative beliefs as percentage of overall negative total	64.9%	86.7%	61.5%	74.6%

In Table 6, we see the overview of coding for negative attitudes and beliefs, with the total numbers of all manual codes together with the specific numbers for codes related to multilingual pupils and multilingual pre-service teachers, respectively. There is a difference across the participants concerning the focus of their negative attitudes and beliefs. For example, similar to the coding for positive beliefs, the inexperienced pre-service teachers have little to say about multilingual pupils or multilingual pre-service teachers, with only one negative belief about pupils coded.

While in-service teachers were generally more positive than negative about multilingualism (see Table 4), they experienced challenges in their work with multilingual pupils. IT.02 expressed negative attitudes by describing the common scenario of newly arrived pupils in the classroom.

IT.02: [They] can come at any time, ‘in two weeks you will get a new pupil’ … 

R: Who does not know Swedish at all?

IT.02: Who hardly knows any Swedish at all. In the middle of term, yeah, at any time. […] The thing is, then we probably do not have any resources.

Similarly, IT.04 expressed negative beliefs about multilingual pupils’ vocabulary skills, in part due to their perceived lack of exposure to typical Swedish activities, but also due to a perceived lack of literacy skills in both their mother tongue and in Swedish.

Teacher educators are the only cohort with negative attitudes and beliefs about the multilingual students studying to become teachers, with 17 negative codes combined. TE.04, who herself has a multilingual background, also attributed negative attitudes to some in-service teachers who question whether mother tongue instruction should be given space in schools. At the same time, TE.04 struggled with accepting linguistic diversity in teacher education, returning several times to her concerns with the Swedish proficiency of multilingual pre-service teachers: ‘ … but (pause) then (pause) as a teacher-teacher so (pause) we see we have a large (pause) population of multilingual individuals and students we have (pause) many do well, but many do not do well’. Her use of pauses as hesitation markers reveals a reluctance to acknowledge the linguistic resources of the pre-service teachers, instead focusing on how well they may manage their studies or not (i.e. a language-as-problem orientation).

While teacher educators acknowledged the realities of multilingualism, they were reluctant to accept accountability for teaching pre-service teachers about multilingualism, as TE.03 said: ‘I think that knowledge of multilingualism is very important, but it’s not really our responsibility’. At the same time, they claimed lack of knowledge about how and what other teacher educators include in their subject courses and did not indicate attempting to coordinate efforts within teacher education. This is indicative of the kinds of tensions revealed in the interviews, as seen in the next section.

Tensions

The most common tensions identified in the interviews centred on how multilingual pre-service teachers fare in teacher education and how prepared pre-services teachers are to meet the realities of the multilingual school. In the interviews with the teacher educators, descriptions of whether students with Swedish as a second language manage their studies were prominent. TE.01 described tensions created in the assessment of multilingual students whom she considered as having weaker Swedish. TE.04 discussed the tensions created among her colleagues when the pre-service teachers write their final degree theses, stating, ‘A number of students have not been in Sweden so long and they will not manage their final theses because they cannot write in Swedish’. She continued with a description of how she and her colleagues have ongoing discussions about whether these students can pass the thesis course or not, and commented:

TE.04: We cannot pass [the thesis] if someone cannot read or speak Swedish, […] then it is actually, then it is a burden for those who do not have a rich Swedish language or who are perhaps new Swedes and (pause) and then I think in my … then maybe I become normative but I think that (pause) I want teachers when they get out [of university] to be able to speak, listen and write in Swedish, I think that is because I want my grandchildren to learn good language.

Teacher educators accounted for over half of all tension codes (see Table 4). Of these, nine involved multilingual students and four multilingual pupils, see Table 7.

Table 7. Overview of tensions (TE = Teacher educators; IT = In-service teachers; PT = Pre-service teachers).

Manual codes	Interview cohort (percentage of respective category in brackets)
	TE	IT	PT	Totals
Totals tensions	59 (52.2%)	40 (35.4%)	14 (12.4%)	113 (100%)
-about the multilingual pupils	4 (6.8%)	5 (12.5%)	1 (7.1%)	10 (8.8%)
-about the multilingual pre-service teachers	9 (15.3%)	0 (0%)	0 (0%)	9 (8%)
-about teacher education	3 (5.1%)	2 (5%)	8 (57.1%)	13 (11.5%)

The nine coded quotes for tensions expressed by teacher educators in relation to pre-service teachers came from educators in the Arts, Mathematics, and Swedish. TE.02 described the tension of wanting to acknowledge the linguistic resources of pre-service teachers, but also seeing there are problems with their language:

TE.02: Because we need, we need these people in school. And yes, you can tell them to become a teacher in like mother tongue or something, but I do not really think that. I think they should be able to become – But it is clear that they must be able to speak Swedish well if they are to be able to become Swedish teachers.

TE.01, in the same vein, expressed a concern that she may have different expectations for the multilingual pre-service teachers, struggling with whether to accept a level of Swedish that she did not consider was up to par. On the one hand, she strove to accommodate their needs, while on the other hand, she felt that strong Swedish was needed for their future teaching.

Even though TE.05 claimed that a second language perspective is present in every syllabus for Swedish for every learner age, most of the pre-service teachers described their fears of being woefully unprepared for working with multilingual pupils. In PT.01, they stated that a multilingual perspective tends to be limited to the subject Swedish. Both PT.01 and PT.03 also wished there were more focus on the teaching methodology, or as one pre-service teacher in PT.01 described it: ‘The how is missing’. In PT.02, one pre-service teacher stated, ‘The knowledge I have from my university about multilingualism hasn’t helped me so much’. Instead, she felt she had to continue to search for resources, such as those found at the Swedish National Agency for Education. Her classmate in the same group concurred about the lack of preparation in their education: ‘These multilingual children disappeared a little, yes, but that's how I have experienced it. That they are not really part of the education’.

Several pre-service teachers noted that their practical work experience could be like a lottery: if they were in a school with multilingual pupils, they learned something new, but if they were in a school with ‘nothing’, as one pre-service teacher in PT.03 expressed it, then they would not be as prepared. This was echoed in TE.05’s statement: ‘The students themselves comment, it is that some students end up on [practical work experience] in schools that have no problems so to speak, so what can it be, a small rural school where everyone is called Svensson’. By Svensson, she meant schools in which all the pupils are ‘Swedish’, implying that schools with multilingual pupils are problematic and highlighting a language-as-problem orientation.

In-service teachers also indicated a lack of preparation in their own studies when they reflected on their teacher training. For instance, IT.01 mentioned learning to focus on ‘food and festivals’ in her education, which she subsequently realised is not at all what her multilingual pupils need. It would have been more useful to know how to meet the challenges of receiving newly arrived pupils ‘at any time’ or how to work with study guidance in mother tongue. IT.02 also emphasised the need for practical solutions. With her training in translanguaging, IT.03 seemed to promote multilingualism in her interview, but acknowledged that her pupils know each other’s languages better than she does: ‘They are more at ease and they speak and change languages and alternate, alternate languages and use from different languages as well’, thus confusing the capability of using different languages with the planned, purposeful use of the pupils’ linguistic resources for learning (see Juvonen & Källkvist, 2021).

Discussion

In the results above, we explored the attitudes and beliefs of teacher educators, in-service teachers, and pre-service teachers, differentiating between attitudes as participants’ opinions and evaluations about multilingualism and beliefs as how the participants think about multilingualism in relation to learning (RQ1). We now consider the similarities and differences identified across and between the three cohorts (RQ2), as well as in light of Ruiz’s (1984) language orientations, highlighting the existence of language hierarchies and the ideological underpinnings of both the attitudes and the beliefs of the participants.

Both teacher educators and in-service teachers make clear the role of Swedish as the only legitimate language for learning in teacher education and the school, in adherence to the monolingual mindset. For example, by not engaging in teaching that makes use of her pupils’ linguistic resources – despite the pupils’ own active use of translanguaging – IT.03 delegitimises their value as languages for learning. Nevertheless, a difference is evident in how Ruiz’s language orientations are reflected in the role of Swedish at the two levels of education. In primary school, a lack of Swedish proficiency is not seen as a problem. Rather, in-service teachers describe their beliefs that multilingual pupils use their mother tongues as a resource for learning Swedish and general subjects, similar to Lundberg’s (2019) findings of generally positive views of multilingualism among primary teachers. A pupil’s Swedish language development is accepted as an ongoing process, as well as a process that benefits from literacy skills in the pupil’s mother tongue. Even the in-service teacher who works with national minority language pupils feels that in the end, as Swedish is key to the pupils’ school success, the pupils need to ‘catch up’ to their monolingual peers from other schools in order to manage the education trajectory in Sweden.

Teacher educators, in contrast, reveal a language-as-problem orientation in relation to Swedish in teacher training courses. They expect pre-service teachers to have already attained high proficiency in Swedish, unlike the process view held by in-service teachers. The teacher educators’ evident struggles in accepting the perceived lower Swedish language proficiency of pre-service teachers confirm the prevalent linguistic hierarchy, reflecting the dominance of Swedish in teacher education as noted by Carlson (2009). Only the majority language is allocated space for learning – and, in turn, for future teaching. This is plain in their assessment quandaries when pre-service teachers do not match the Swedish norms expected in assignments such as theses, as well as in their reflections on the future positions that pre-service teachers may have in the Swedish school. Some teacher educators suggest that multilingual pre-service teachers may be best suited to working as MTI teachers or as SGMT tutors, as their level of Swedish is too problematic for the monolingual Swedish pupil, revealing negative attitudes towards the multilingualism of these pre-service teachers.

Contrary to the teacher educators and in-service teachers in this study, the pre-service teachers do not reveal any clear beliefs about mother tongues as a problem for learning. They do not express attitudes towards seeing multilingualism as valued in itself or as a resource for society (as argued in Schalley, 2020) either. This is most likely due to their limited experience in working with linguistically diverse classrooms, as well as the lack of instruction on multilingualism in their teacher training programmes, which could also lead to a potential absence of awareness. More experience in the multilingual classroom would facilitate greater awareness, as seen, for example, in Halpern et al. (2021). Instead, pre-service teachers have a generally positive attitude towards multilingualism, albeit without clear beliefs on whether multilingualism offers possibilities or hindrances in the classroom, or an understanding of the enactment of language-as-resource. In addition, the few mentions of language-as-right in the interviews all referred to the policies of MTI and SGMT, and not to general rights of one’s own language. Interestingly, Ruiz’s (1984) language-as-right orientation is thus not expressed as an attitude or a belief, but rather as a matter of fact, as part of the national policy of the Swedish Education Act (2010).

Finally, tensions are apparent: first, between the cohorts, expressed in the polarity of attitudes in particular (see Table 4), and second, between teacher education and the realities of the multilingual classroom in the Swedish school (as seen in Table 7). The tensions between cohorts are most evident in the differences seen in the teacher educators’ perspectives contrasted with those of the in-service teachers and pre-service teachers. The teacher educators grapple with being unprepared and uncertain about how to cope with multilingual pre-service teachers, whom they find problematic, possibly in parallel to a lack of knowledge about how to best prepare them for working with multilingual pupils. The pre-service teachers present a more positive attitude (see Table 5) but struggle with the lack of preparation for a multilingual reality they have glimpsed in their work placement (see also Paulsrud & Zilliacus, 2018). This clash of teacher education meeting the realities of the classroom represents the other evident tension. Both pre-service and in-service teachers lament the nearly non-existent presence of multilingual pedagogy in their training, which results in an education out of step with the multilingual Sweden of today. As Halpern et al. (2021) suggest in their study of culturally aware teacher education, pre-service teachers preparing for work with linguistically diverse pupils need consistent exposure and training over the course of their education. However, as our study makes clear, it is not only necessary to invest in the further education of teachers. Teacher educators must work on their attitudes as well, so that pre-service teachers can acquire a solid understanding of multilingualism as both a resource and a right during their teacher training.

Generally, we see a tendency across the cohorts towards not only more expressed beliefs than attitudes, but also towards an imbalance in the polarity between the two concepts, especially for in-service teachers who have more coded instances of negative beliefs (21) than negative attitudes (6) (Table 6) about multilingual pupils in particular. However, in-service teachers also expressed positive beliefs (26) and positive attitudes (19) towards multilingual pupils (Table 5). Thus, tensions are evident in in-service teachers’ own experiences of working with multilingual pupils. As beliefs are related to learning, they may reflect how the in-service teachers encounter challenges in Swedish schools today. Still, as more coded positive attitudes and beliefs in total are identified in in-service teachers’ utterances, these tensions may give way to positive practices recognising the linguistic resources of multilingualism in the classroom.

Conclusion

A growing number of multilingual pupils from diverse cultural backgrounds in schools globally presents new challenges and possibilities to teacher educators, in-service teachers, and pre-service teachers. These key actors on the ground ‘vote with their feet’ in translating, interpreting and enacting the policies. Nevertheless, as demonstrated here, the interviewed Swedish teacher educators acknowledge the need to prepare pre-service teachers for the multilingual realities of school but take no responsibility of acting upon their view. They are more concerned with the pre-service teachers’ proficiency in Swedish. The possibilities of developing a positive orientation towards multilingualism during teacher training are limited when teacher educators maintain a monolingual orientation and may shirk from their responsibility to prepare pre-service teachers for the multilingual classroom. Pre-service teachers also acknowledge the need for – and criticise the lack of – education in multilingualism. Finally, in-service teachers try to manage the realities of linguistic diversity in the classroom by instrumentalizing their positive attitudes towards multilingualism as vehicles to language and subject learning, despite the absence of in-service training.

Research has increasingly reported a gradual shift from a ‘monolingual mindset’ (Clyne, 2005) that has often coloured the enactment of language and education policy at school (Ellis et al., 2010; Hedman & Rosén, 2021) to a ‘multilingual turn’ in education (Conteh & Meier, 2014; May, 2014, 2019), and hence towards seeing the multilingual, multimodal and multicultural repertoires of individuals as resourceful prior knowledge for learning (Juvonen & Källkvist, 2021). Our results instead reveal that in several primary teacher education programmes and primary classrooms in Sweden, multilingualism may be tolerated but the goal remains Swedish. This study is focused on the Swedish context, but similar outcomes have been described in other contexts as well. Flockton and Cunningham (2021), for example, report on a similar disparity in the United Kingdom between the training needs expressed by pre-service teachers and what teacher educators believe pre-service teachers need in order to be prepared to support multilingual pupils in school.

Methodologically, our study offers a novel approach to understanding expressions of educators’ attitudes and beliefs about linguistic diversity in the Swedish education system. A systematic analysis grounded on five different aspects of the semantically identified themes (focus, topic, domain, attitude or belief, and polarity) goes beyond a standard thematic analysis, offering a more precise view on not only what, but also how, the educators’ views reflect their language orientations, based on the analysis of language use. The analysis also affords a more detailed picture of the differences between the cohorts’ expressed attitudes and beliefs than an analysis based on broader themes would have offered, not least due to differing attitudes and beliefs about both the multilingual pre-service teachers in teacher education and the multilingual pupils in the Swedish school. The polarity coding also reveals more than simply identifying attitudes and beliefs, as tensions become visible when we look beyond the concepts.

We acknowledge limitations due to various cohort sizes and targeted interview protocols. However, our aim was not to obtain generalisable results but rather to contribute an understanding of both attitudes and beliefs vis-à-vis multilingualism of key policy actors in Swedish education, filling a gap in the current research with an innovative comparison of the three groups. In conclusion, this study contributes to the greater conversation about how teacher education must address training in multilingual pedagogy. The results presented here may help teacher educators in Sweden and beyond to understand the needs of pre-service and in-service teachers who strive to provide quality education for all, in which language is seen as a resource for both the individual and society.

Disclosure statement

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

Additional information

Funding

This work was supported by Dalarna University, Sweden, under an Education and Learning grant; Stockholm University, Sweden, under the research project Multilingual and Intercultural Education in Sweden and Finland; and by the Centre for Language and Literature Education at Karlstad University, Sweden.

Notes

1 The interviews with the teacher educators and pre-service teachers were conducted as part of the international research project MINTED (Multilingual and Intercultural Education in Sweden and Finland) (need for external ethics application deemed not necessary after internal review, Stockholm University/University of Helsinki). The interviews with the in-service teachers were conducted as part of the international research project ABK (Attitudes, Beliefs and Knowledge on Multilingualism) (Ethics approval C2017/947, Karlstad University, Sweden).