Introduction

In 1977, Sweden became the first country in the world to offer publicly funded mother-tongue instruction (MTI) to children of immigrants within the regular school system, a policy that did not even exist for indigenous languages in many countries at the time. This so-called Home Language Reform became one of the cornerstones of Sweden’s overall shift from assimilationist to multicultural policies towards historical and immigrant minorities in the mid-1970s (Borevi, 2002; Wickström, 2015). Nearly half a century later, MTI is provided in over 160 languages in 286 of Sweden’s 290 municipalities todayFootnote 1, with a clear majority of eligible pupils consistently choosing this optional school subjectFootnote 2. Publicly funded MTI on this scale is unheard of even in big migration-receiving countries with the strongest multicultural policy records such as Australia and CanadaFootnote 3.

Sweden’s long history with this type of MTI praxis makes it a particularly valuable case for studying the diachronic evolution of MTI from a language planning and policy (LPP) perspective. Although Sweden’s enduring tradition of educational policies geared towards additive bilingualism and minority-language maintenance have attracted much academic attention both nationally and internationally (Boyd & Huss, 2000; Boyd, 2007; Cabau, 2014), until now only few studies have attempted to study its trajectory from an LPP perspective based on a large corpus of policy documents (Bajqinca, 2019 being the notable exception). The aim of the present study is to fill the research gap regarding the evolution of Sweden’s LPP orientations towards MTI based on a new diachronic corpus of key policy documents, and by applying the framework of language-planning orientations developed by Ruiz (1984). His three orientations − language as problem, language as right, and language as resource − have proven to be a fruitful approach to studying bilingual education, not the least the teaching of minority languages (Hult & Hornberger, 2016; Fránquiz et al., 2016). To my knowledge, Ruiz’s orientations have so far only been applied to MTI in Sweden in one study, which focused on a single minority language (Vuorsola, 2019).

The present paper examines the motivations that key policy documents have put forward as justifications for MTI between 1974 and 2022. By including the period 2017 − 2022, when several key policy documents were published, the present corpus also serves to complement Bajqinca’s (2019) earlier corpus, which did not include the developments after 2016. The following four research questions will be pursued.

Q1: How are the motivations for MTI formulated and thematised in key policy documents?

Q2: How can the attested motivations and themes be related to Ruiz’s three LPP orientations?

Q3: How has the prevalence and internal make-up of the LPP orientations evolved over time?

Q4: How was the LPP orientations’ evolution underpinned by political ideologies?

Given Sweden’s leading position in MTI and Ruiz’s (1984) portrayal of language as resource as the most effective orientation for promoting additive bilingualism, one might hypothesize that this orientation must have been responsible for Sweden’s enduring MTI policy. The paper will test this hypothesis and discuss the relationship between different orientations towards MTI. A further theoretical objective is to examine more closely the ideological underpinnings of his LPP orientations, especially the much-debated language as resource, which has been accused of marketizing minority languages (Petrovic, 2005; Ricento, 2005). The paper’s methodological contribution is its innovative quantitative approach to the evolution of the LPP orientations by breaking them down into so-called motivational units based on keyword searches and by grouping these into themes.

Theoretical background

Ruiz (1984:14) defines orientation in language planning as “a complex of dispositions toward language and its role, and toward languages and their role in society” and points out that “they are basic to language planning in that they delimit the ways we talk about language and language issues” and “determine what is thinkable about language in society.” He goes on to distinguish between three main orientations: language as problem, language as right, and language as resource.

Ruiz’s three orientations in language planning

According to Ruiz (1984:16 − 20), the language-as-problem orientation is, on the one hand, related to practical issues to be resolved through language planning that typically arise in nations’ developmental phases and tend to stem from language diversity. On the other hand, problems in this sense also entail (correctly or falsely) associating certain language issues and groups, such as linguistic minorities, with social problems that need to be remedied through policy. As an example, he mentions educational policies advocating transitional bilingualism, whereby minority languages (ML) are tolerated until pupils attain sufficient proficiency in the majority language, whose acquisition is the main linguistic focus of schooling. Thus, the minority pupils are supposed to overcome problems such as “intellectual limitations”, “linguistic deficiencies” and general “social backwardness” associated with maintaining their minoritised mother tongue (Ruiz, 1984:18).

In their comprehensive overview of Ruiz’s framework, Hult and Hornberger (2016:35) characterise the language-as-right orientation as one that “seeks to address linguistically-based inequities using compensatory legal mechanisms” and point out that it shares its compensatory nature with language as problem, the crucial difference being that the compensation has diametrically different grounds. They go on to specify that “language rights can be understood as what is legally codified about language use, often with special attention to the human and civil rights of minorities to use and maintain their languages” (Hult & Hornberger, 2016:36).

The final and historically newest orientation, language as resource, differs from the other two in several ways. Firstly, it is not compensatory in the sense that it does not aim to solve problems or right historical wrongs and present inequalities. Instead, it shifts the focus to positive aspects of multilingualism and linguistic diversity (cf. Hult & Hornberger, 2016). According to Ruiz (1984:24), this orientation has the potential to alleviate some of the conflicts that stem from the other two orientations. By foregrounding that language functions as a resource in society, this orientation can ease tensions between minority and majority, draw attention to the roles that different languages play in society, raise the prestige of MLs and promote cooperative language planning.

Ruiz (1984:26 − 27) goes on to highlight two dimensions in which MLs are valuable. Firstly, they have value for the larger society in terms of business, trade, diplomacy, national security and military preparedness. Secondly, the many positive effects that multilingualism entails for individual speakers is also a social value. Hult and Hornberger (2016:33, 38 − 39) use the terms “extrinsic” and “intrinsic” respectively for these two dimensions. The intrinsic value accrues to the minority communities and their individual members, whereas the extrinsic value benefits the whole society. As such, language is both a personal and national resource, as Ruiz (2010:159) clarified later. Ruiz (1984:27) concludes his original description of this orientation by stating that “[l]anguage planning efforts which start with the assumption that language is a resource to be managed, developed and conserved would tend to regard language-minority communities as important sources of expertise.”

The ideological underpinnings of the orientations

Regarding language as problem, (Ruiz, 1984:20) remarks that “since language problems are never merely language problems, […] this particular orientation toward language planning may be representative of a more general outlook on cultural and social diversity.” Hult and Hornberger (2016:33 − 34) comment that this orientation labels linguistic diversity, bilingualism and MLs as problematic, because they are viewed as threats to the majority language and to national unity. Accordingly, the proposed solution to the problem is subtractive bilingualism and full assimilation to the majority language, as typically advocated by nationalist ideologies. Instead, language as right is associated by Hult and Hornberger (2016:33) with the promotion of additive bilingualism and by positive attitudes to multilingualism and diversity.

In contrast to the other two orientations, the underpinnings of language as resource in terms of political ideology have been the topic of intense academic debate. The main points of criticism against language as resource can be summarised as follows (for an overview see Ruiz, 2010 and Hult & Hornberger, 2016:39 − 42). The way Ruiz (1984) originally presented this orientation downplayed both the intrinsic value of language (Petrovic, 2005) and the need for legal instruments in LPP (Kontra, Phillipson & Skutnabb-Kangas 1999, Ricento, 2005). At the same time, it relied heavily on the extrinsic value of language, thus facilitating the marketization of minority-language skills (Petrovic, 2005; Ricento, 2005). Furthermore, the latter reliance was labelled as “neoliberal” by Petrovic (2005) and the language-as-resource orientation was accused of complicity in the maintenance of the system that underlies the marginalisation of minority communities and languages to begin with (Ricento, 2005; Petrovic, 2005). Hult and Hornberger (2016) and Ruiz (2010) have responded to this criticism by reaffirming the intrinsic value of language and by stating that language as resource should not be viewed from a narrow utilitarian perspective. Moreover, Ruiz (2010:165) has toned down his previous scepticism towards language as right and has clarified that language as resource constitutes a precondition to language as right (Ruiz, 2010:165). Also Hult and Hornberger (2016:42) point out that these two orientations are not antagonistic but complementary.

More recently, Ruiz’s framework has also attracted criticism of a more fundamental type. Kaveh (2023:2) questions the whole premise behind fixed orientations that take languages as their starting point. Instead, she calls for a “humanization” of the orientations by putting the language users and agency first and thus focusing more on the micro level. Kaveh (2023:5) also joins Petrovic (2005) in emphasizing that power relations are missing as a central dimension in Ruiz’s framework. Similarly, Zúñiga (2016:252) calls for ideological clarity in the application of the framework.

Mother-tongue instruction in Sweden

A brief history of MTI in Sweden

Sweden’s multiculturalist integration policies of the 1970s with the slogan “equality, freedom of choice, cooperation” (Prop., 1975:26) emphasised that immigrant integration into majority society did not require abandoning one’s minority heritage. Following pilot projects with MLs (mainly Finnish) in the 1960s and early 1970s, the Home-Language Reform (Prop., 1975/76:118) came into force in 1977. Its stated objective was to achieve “active bilingualism” among minority pupils, which was later clarified as “first-language level in both languages” (SOU, 1983:57, 143). The reform stipulated that upon the fulfilment of certain criteria, municipalities had an obligation to offer MTI to pupils in compulsory and upper secondary schools and that they would receive earmarked funding for this.

The reform was appreciated by minorities and MTI was rapidly established in a large number of languages in many parts of Sweden. However, the policymakers’ optimism waned, when implementation problems surfaced (Hyltenstam & Milani, 2012:122). When Sweden was hit by a financial crisis in the beginning of the 1990s, austerity measures and neoliberal economic reforms were enacted (Kärrylä, 2021). At the same time, there was a surge in xenophobic opinion and violence. The costs and benefits of MTI soon became the topic of a critical debate in politics and media (see Wingstedt, 1998 for the media debate). Consequently, the availability and organisation of MTI was economised and teacher training was discontinued in an effort to cut public costs (Borevi, 2002:230 − 233). Soon, substantial local budget cuts to MTI ensued, when the earmarked funding for MTI was abolished. These changes in the early 1990s lead to both a devaluation of the status of MTI and a decrease in the number of teaching hours (from an average of two hours to under one hour) and of pupils (Hyltenstam & Milani, 2012:122 − 123).

Towards the end of the 1990s, the pendulum started swinging in the opposite direction, as certain improvements were made that signalled that policymakers nevertheless appreciated the existence of MTI. These have been rather small for immigrant MLs but substantial for the national MLs. In 1999, Sweden ratified the minority conventions of the Council of Europe, thereby granting five MLs official status as national minority languages (NMLs). The 2000s saw an important turn in making language policy more explicit in Sweden (see Hult, 2004 for an overview). This resulted in the new Language Act (SFS 2009:600) and the National Minorities and Minority Languages Act (SFS 2009:724), which both strengthened the legal basis for protecting MLs.

In the 2010s, the government took steps to improve the implementation of the NML policies following several rounds of criticism from the Council of Europe (Cabau, 2014:416). The eligibility criteria for participation in MTI were gradually relaxed for the NMLs and a new scheme for MTI teacher training was introduced after a hiatus of more than twenty years. Lastly, two public inquiries on MTI were commissioned by the Social Democrat-Green Party government (SOU, 2017:91; SOU, 2019:18). However, the recommendations of these two inquiries have hitherto not been enacted. Instead, the current right-wing government intends to commission a new inquiry into MTI (Tidö Agreement, 2022:54).

Previous research on MTI in Sweden

In this section, previous research on MTI in Sweden is organised under two headings. The first subsection summarises research that provides an overview of the objectives and perceived benefits of MTI from a purely educational perspective. The second subsection surveys research that discusses the rationale behind MTI from an LPP perspective, where Ruiz’s orientations in language planning are invoked implicitly or explicitly.

Objectives of MTI

Several studies have examined the rationale behind MTI (Hyltenstam & Tuomela, 1996; Wingstedt, 1998; Spetz, 2014). These studies focus on how the objectives or benefits of MTI are perceived by pupils, parents, school principals, municipal officials or the general public as well as how they are justified in research and policy documents. In their authoritative book on language education for immigrants and minorities in Sweden, Hyltenstam and Tuomela (1996:29 − 44) summarise the rationale behind the Home-Language Reform of 1977 under two headings: strengthening the pupils’ ethnic and cultural identity; and supporting their cognitive, linguistic and learning development. In her dissertation on language ideologies in Sweden, Wingstedt (1998: 135 − 157) investigated inter alia the beliefs and attitudes manifested in the arguments that were put forward in support of MTI in the 1990s by researchers as well as arguments among the general public for and against MTI in the media debate on the subject. She lists the positive functions emphasised by both the researchers and the pro-MTI public as facilitating language acquisition, linguistic and conceptual development, social psychological development, multicultural integration and being an asset in internationalisation. In a report on MTI for the Language Council of Sweden, Spetz (2014: 26 − 27) compared three successive MTI syllabi between 1980 and 2011 and identified the following three overarching functions of MTI: strengthening cultural identity, contributing to personal development and strengthening knowledge of the mother tongue.

Orientations towards MTI

A number of studies have investigated the background of the ground-breaking Home-Language Reform of 1977. In a historical comparative study of Denmark and Sweden, Salö et al. (2018:591) highlight the role played by the perceived dangers of “semilingualism” in the establishment of MTI in Sweden. Salö et al. (2018:597) point out that “[p]oliticians and advocates of MTI alike could argue that the inequality bred by semilingualism in the north [i.e. in the Torne Valley Finnish community] threatened an entire generation of immigrant children, and that this situation was incompatible with the Swedish notion of equality hailed at the time (Wickström, 2015).”

In the period when the reform was conceived and established, there was a dominant academic discourse in Sweden that warned against the dangers of semilingualism among minority children (Municio, 1987; Hyltenstam & Arnberg, 1988; Borevi, 2002; Boyd, 2007; Wickström, 2015; Salö, 2020; Salö & Karlander, 2022). The term semilingualism was introduced in the late 1960s by the Swedish linguist Hansegård (1968) who studied bilingualism in the Torne Valley (see Salö & Karlander, 2022 for a full account). The policymakers’ concerns regarding semilingualism revealed by these studies suggest that MTI was partly seen as a type of preventative measure in order avoid intellectual and psychosocial problems due to insufficient development of both the mother tongue and the second language. According to Borevi (2002:202 − 210), these concerns were so strong that the policymakers even considered making MTI obligatory. Although the researchers who shed light on the relevance of this semilingualism discourse for MTI do not state it explicitly, taken together their accounts suggest that a language-as-problem orientation played an important part in justifying the original establishment of MTI in Sweden in the 1970s. According to Boyd (2007:148), “there was no talk at this time [i.e. when MTI was established] of multilingualism ‘enriching’ or being a ‘resource’ in Swedish society until the 1990s.”

In her general review of language education policy and planning in Sweden, Cabau (2014:415 − 416) mentions a brochure for parents published by the the National Agency for Education (Skolverket, 2006), where the benefits of MTI are “expressed in terms of the development of a positive and multicultural identity, facilitation of contact with the family inside and outside the country, as a resource for pursuing studies and integrating working life in Sweden or in other countries, and in terms of increasing opportunities to gain new knowledge in different subjects.” Moreover, “the presence of multilingual individuals in the country is presented as a considerable asset at the societal level”. Although it is not pointed out explicitly by Cabau (2014), her review reveals that a language-as-resource orientation to MTI (both in intrinsic and extrinsic terms) was present, at least in education-policy circles, in Sweden in the 2000s.

The research overview section of the latest governmental inquiry report on MTI (SOU, 2019:18, 197 − 201) touches upon what in the Swedish-language literature has come to be termed as the “instrumentalisation of MTI”, whereby the focus shifts to MTI’s benefits for achieving other goals than mother-tongue development (see also Cabau, 2014 and Ganuza & Hedman, 2018:19 for some critical remarks on the role of academia in this trend). The inquiry report puts this succinctly as follows (my translation).

“[T]he public domain recognises the importance of the mother tongue for culture and identity but nonetheless poses the question: What’s in it for us? What benefit does society have, in this case expressed as better Swedish and better school performance, from the funds and time invested in the subject mother tongue?” (SOU 2019:18, 201).

Although not explicitly stated in the inquiry, this type of instrumentality could be interpreted as an extrinsic resource orientation, if the extrinsic benefits that accrue to the whole society are considered greater or primary relative to those that accrue to the individual speakerFootnote 4.

There is one recent study that explicitly refers to Ruiz’s (1984) LPP orientation framework regarding MLs in Swedish schools. Vuorsola (2019) examined societal support for the educational provisions of Finnish as a NML in the Swedish school system. He studied four key national policy documents from the period 2009 − 2011 as well as the educational implementation of Sweden’s NML policies. His analysis shows that the most utilised among Ruiz’s LPP orientations is language as right on the national level (Vuorsola, 2019:372 − 380).

The research surveyed above has the considerable limitation of only referring to a small number of documents and/or a short time period. Thus, these studies only provide glimpses into the evolution of the rationale behind MTI. However, there is one study that has systematically applied the same method of analysis to a large corpus of documents pertaining to several decades. In a recent doctoral dissertation, Bajqinca (2019) conducted a critical discourse analysis regarding the status and positioning of MTI in policy documents pertaining to four periods from 1957 to 2017Footnote 5. According to his analysis, the period prior to any type of MTI in schools (1957 − 1965) was characterised by the discourse of a homogeneous nation held together by the Swedish language and accordingly, multilingualism was seen as a harmful complication. Although Bajqinca (2019) does not explicitly refer to Ruiz’s framework, this account suggests a language-as-problem orientation. In the following period 1966 − 1988, when the assimilationist approach to minorities was gradually replaced by multiculturalism, bilingual education received greater attention and the mother tongue began being considered an asset and MTI an instrument of equitable education. This points to a shift towards the orientations language-as-right and language-as-resource in intrinsic terms. Bajqinca (2019:182, 185) goes on to show that the period 1989 − 1999 was, instead, marked by negative attitudes towards MTI and a “devaluation” of MTI as well as by greater focus on Swedish as the main instrument of integration and employability based on the new market-liberal ideology. Furthermore, this was coupled with a new conservative appreciation of Swedish cultural heritage in a globalised world (cf. Oakes, 2005 on “nationalist revival” in Sweden in the same period). At the same time, the mother tongue was portrayed as mainly belonging to the individual’s home sphere. These developments can be interpreted as a neoliberal shift to extrinsic aspects of language as resource (cf. Petrovic, 2005). Bajqinca (2019:184) analyses the final period 2000 − 2017 as one dominated by new language-policy developments that granted official status to Swedish as the country’s principal language and highlighted human rights in relation to language as well as Sweden’s historical heritage. This suggests a language-as-right orientation in this period in addition to elements that can be linked to the extrinsic value of NMLs as national resources that merit conservation (cf. Ruiz, 1984). According to Bajqinca, this shift in focus has practically reduced MTI to an instrument of school success while at the same time reducing the maintenance of the mother tongue to a responsibility for the individual rather than the state. Within Ruiz’s framework, these trends can be interpreted as extrinsic value being reserved for the NMLs, while immigrant MLs are associated mainly with intrinsic value.

Table 1 summarises the results from these previous studies and provides an overview of the diachronic evolution of LPP orientations towards MTI, as put together based on my application of Ruiz’s framework to the findings of these studiesFootnote 6.

Table 1 How previous studies on MTI can be interpreted according to Ruiz’s (1984) LPP orientation framework
Full size table

Taken together, these previous studies suggest that Sweden’s orientation to MTI has shifted over time from language as problem to a combination of language as right and language as intrinsic resource, followed by the later addition of language as extrinsic resource and a comeback for language as right. The table also reveals some dissonance between Salö et al. (2018) and Bajqinca (2019), as they indirectly point to a problem orientation in two different periods without discussing possible links to preceding or subsequent periods. There also seems to be disagreement between Boyd (2007) and Bajqinca (2019) regarding when the extrinsic resource orientation first emerged.

Materials and methods

Material

This paper focuses on documents pertaining to what (Schmidt, 2008:310) calls “coordinative discourse”, which is at play among policy actors in the formation, development and justification of policy ideas. This type of documents constitute either policy output (government bills, laws, public ordinances, curricula and syllabi for schools) or pertain to the final part of the policy input stage that is geared towards potential policy output (governmental inquiry reports). Hence, the intention is to zoom in on the part of the policy process where official discourses are justified and legitimised.

Firstly, 50 potentially relevant policy documents geared towards policy output concerning language policy, language education, minorities and integration policy were compiled into a preliminary corpus. Each document in this corpus was studied in order to ascertain its objective, area of focus, general tone and wider language-political significance. The document’s summary and potentially relevant chapters were examined in their entirety in order to identify relevant passages. Subsequently, a content analysis was carried out in order to identify those documents that actually contained formulations that could be interpreted as motivations for MTI. Here, the following keywords were used in the search: mother tongue, home language, language, bilingual, multilingual, minority, immigrant, multicultural, and diversityFootnote 7. The analysis of relevant passages was done through an approach where the whole passage (often a paragraph), the document’s type and objective as well as the passage’s heading and position in the document were together taken as the discursive context for the formulation of motivations. 16 of the documents turned out not to contain any of the specified keywords or any passages related to MTI. Therefore, these were excluded from the corpus. Among the remaining 34 documents, eight documents contained fewer than four different types of motivation. These were excluded from the final corpus, as their peripheral treatment of the subject did not qualify them as key documents, and in order not to eschew the planned statistical analysesFootnote 8. This selection process resulted in a final corpus consisting of 26 key policy documents, which are summarised in Table 2 (see Appendix 1 for a more detailed overview).

Table 2 Overview of the distribution of the 26 examined policy documents according to type
Full size table

As the second column of Table 2 shows, the four periods’ lengths range from nine to thirteen years and they contain different numbers of documents between four and eight. As the document type columns show, input and output documents are equally prevalent in the corpus and their distribution differs only slightly from one period to another.

Method of analysis

The most obvious type of motivation contained a formulation that directly addressed the purpose of MTI such as “MTI helps achieving the goal of active bilingualism”. However, many documents contained indirect motivations where the formulations could be situated on any of the following levels in a chain of argumentation: language > mother tongue > educational support for MLs > bilingualism > direct motivation for educational support for MLs. To give an example, a chapter discussing MTI could contain a formulation such as “Language is a tool for thinking”. In this particular discursive context, the general term “language” was taken to refer to the “mother tongue”. Thus, a wide net that could catch both direct and indirect motivational formulations was cast in order to provide as comprehensive an account as possible of how the motivations were framed discursively.

Next, each formulation that constituted a motivation for MTI was summarised by a few keywords. This resulted in 53 types of motivations. Subsequently, some of these formulations were deemed semantically so close that they were collapsed into one single “motivational unit (MU)”. This procedure resulted in 45 distinct MUs. Finally, the attested units were coded for each document in an Excel file providing an overview of which MU occurs in which policy document.

Periodization of the documents

Since MTI was established as an important instrument of Sweden’s multicultural policies, the periodization of the policy documents follows major developments in Sweden’s multicultural policies regarding immigrant integration, minorities and language. The periodization principle that was determined a priori was to divide the timeline into separate sections that cover the introduction and evaluation of relevant policies respectively. Upon putting together the corpus, this principle resulted in two pairs of periods focusing on policies mainly geared towards immigrant and national minorities respectively. The first period 1974 − 1983 encompasses the official establishment of multicultural policies starting with the public inquiry report (SOU, 1974:69) that directly preceded the foundational government bills of Swedish multiculturalism (Prop., 1975:26; Prop., 1975/76:118). The next period 1984 − 1996 starts with the public inquiry report (SOU, 1984:55), which led to the government bill of 1985 (Prop., 1985/86:98). These policy documents are widely recognised as laying the foundation for a critical re-evaluation and revision of Sweden’s multicultural policies (Borevi, 2013:145 − 149), which was concluded with the bill of 1997 (Prop., 1997/98:16). Thus, the second period involves the evaluation of the policies introduced in the first period. After these first two periods, where the language policies focused on immigrant minorities without any anchoring in language laws, the final two periods capture developments in multicultural national minority policies and explicit language policies enshrined in new laws. The third period 1997 − 2007 covers the establishment of official national-minority policies that commenced with two public inquiry reports (SOU, 1997:192&193), which recommended that Sweden ratify the minority conventions of the Council of Europe. The final period 2008 − 2022 includes both an evaluation of the national minority policies introduced in the preceding period, and the emergence of new explicit language policies, beginning with the public inquiry for a new language policy (SOU, 2008:26), which ultimately resulted in the Language Act of 2009.

Results and discussion

Overview of the motivational units according to themes and LPP orientations

The aim of this section is to provide answers to the first two research questions that were presented in the introduction. Based on semantic similarity, the 45 MUs were categorised into the following eight themes that indicate what is to be facilitated by MTI (see Appendix 2 for a detailed list): bilingualism, equal rights and opportunities, psychosocial development and wellbeing, intercultural competence, capacity for learning, ethnocultural heritage, multicultural integration, and cultural contribution to society. The delineation and naming of the themes was largely guided by the objectives of MTI highlighted in previous research and by concepts of particular importance for Sweden’s brand of multiculturalism.

Table 3 Theme distribution among 254 occurrences of 45 motivational units 1974 − 2022*
Full size table

In the second column of Table 3, we can see that the number of MUs in each theme ranges from three to ten. The second from last column captures the themes’ occurrence through the proportion of their MUs in the corpus. The last column indicates to which of Ruiz’s (1984) orientations in language planning each theme’s MUs correspond (see the last column and the colour coding in Appendix 2 for the correspondence for each MU). The themes in Table 3 are organised in chronological order according to the mean year of the introduction of each theme’s MUs (see the third column).

As can be seen in detail in Appendix 2, the theme “bilingualism” includes three MUs regarding the facilitation of active bilingualism; language development; and second-language (L2) acquisition. All correspond to the language-as-resource orientation in intrinsic terms, as they primarily focus on benefits for the individual speakerFootnote 9. From the perspective of Sweden’s multicultural policies, this is a key theme for achieving the language-policy goal of “first-language level in both languages” (SOU, 1983:57, p. 143) in bilingual pupils. Another theme of particular importance in the multicultural integration reform of 1975 is “equal rights and opportunities” as it is directly linked to one of the reform’s main pillars, “equality”, as well as to the language-as-right orientation. This theme consists also of three MUs: right to one’s MT; equal opportunity to maintain and develop one’s MT; and right to equitable education.

“Psychosocial development and wellbeingFootnote 10” is made up of nine MUs associated with emotional development; self-esteem; harmonious development; confident identity; and managing minority strain; as well as cultural identity; personality development; understanding of self, group belonging and life situation; and feeling of belonging. The first five of these MUs are linked to the semilingualism discourse of the 1960 and 1970 s and the psychosocial consequences associated with it. By referring to problems in minority children that are to be alleviated through MTI, these MUs clearly correspond to the language-as-problem orientation. Instead, the remaining four MUs in this theme invoke language as a resource in the intrinsic sense without overtly associating the special minority circumstances with psychosocial problems.

The theme “intercultural competenceFootnote 11” consists of the following four MUs: understanding for the norms and values of different cultures; belonging to two cultures and its perception as an enriching experience, as well as the freedom to choose linguistic and cultural identity. Most of these MUs correspond to the language-as-resource orientation in intrinsic terms except for “making multicultural belonging an enriching experience” which displays a language-as-problem orientation. Although some of these MUs were mentioned separately in earlier studies (Hyltenstam & Tuomela, 1996; SOU, 2019:18), in this theme they constitute an innovative way of grouping several motivations under the same heading. Moreover, this unified theme creates a link to the other two main pillars of the multicultural integration reform of 1975, namely “freedom of choice” regarding cultural identity, and “cooperation” between immigrant communities and the majority society.

Another theme that involves an innovative grouping of MUs is “capacity for learning”, which includes four MUs that highlight the role of the mother tongue and MTI in the development of minority pupils’ mental capacities that contribute to successful learning: intellectual development; academic performance; cognitive development; and a means for learning. The first two MUs invoke the importance of preventing potential problems associated with semilingualism and its negative consequences in terms of failure in school. Thus, they easily lend themselves to an association with prevalent discourses on social problems among immigrant children, which corresponds to a language-as-problem orientation. Instead, the other two MUs portray the mother tongue as a resource in intrinsic terms.

“Ethnocultural heritage” is the largest theme with ten MUs that involve developing, maintaining and transmitting one’s cultural heritage and ethnic group identity including the use of the mother tongue in varied contexts. As the MUs in this theme focus on the importance of the mother tongue for the minority group and its members, they all display a language-as-resource orientation in intrinsic terms.

The theme “multicultural integration” is comprised of the following four MUs: preventing socio-economic problems; improving educational conditions in segregated areas; facilitating integration in school; and the fulfilment of integration-policy goals. The minority pupils’ integration into the majority society and its institutions invoke extrinsic benefits for society in the latter two MUs, while the former two rather invoke the risks of failed integration based on a language-as-problem orientation.

The final theme is “cultural contribution to society” with eight MUs that are related to being an asset for society; linguistic and cultural diversity; the school’s internationalisation; creating understanding for and solidarity with different cultures; affirming multiculturalism as well as promoting and revitalising NMLs. The role of minority languages in internationalisation was touched upon in earlier research as an asset for the whole society (Wingstedt, 1998; Cabau, 2014), but not as part of a larger unified theme. All MUs in this theme display a language-as-resource orientation in the extrinsic sense as they invoke the value that different languages have for Swedish society in general.

The evolution of the rationale behind MTI

This section answers the paper’s third research question regarding the evolution of the LPP orientations over time.

Qualitative trends within themes and LPP orientations

In the theme “psychosocial development and wellbeing”, the MU “managing minority strain” is only mentioned once in 1980 and MUs regarding the pupils’ emotional, harmonious and personality development disappear altogether after 2006 (see Appendix 2 for the MUs’ periods of attestation). Instead, “cultural identity”, “self-esteem” and “feeling of belonging” become more prevalent. This suggests a shift from a child-developmental perspective to psychosocial wellbeing. As for “capacity for learning”, the early MU “intellectual development” seems to be replaced by “cognitive development” after 1992. The changes in these two themes are likely owing to the academic discreditation of the semilingualism in the 1980s (Salö & Karlander, 2022). Instead, “means for learning”, which is the only MU attested in all four periods, becomes the most prevalent MU in “capacity for learning” after 1998. The other MU that becomes more prevalent after 2004 is “academic performance”, although it was attested already in the first period. Thus, we observe a parallel shift from a child-developmental perspective to an instrumental perspective that emphasises the importance of the mother tongue for learning and school success (cf. Cabau, 2014; SOU 2019:18; Bajqinca, 2019), rather than as a prerequisite for normal intellectual development. Regarding these two themes with strong ties to the semilingualism discourse, the overview in Appendix 2 reveals a strong tendency for the MUs that were introduced early to display a problem orientation while the later MUs rather adopt an intrinsic resource orientation.

Within the theme “bilingualism”, the MUs advocating the facilitation of “bilingualism” and “language development” are equally prevalent in all four periods. Although “L2 acquisition” is present in all four periods, it is conspicuously more prevalent in the third period. This follows the previously commented pattern of instrumentalisation of MTI (Cabau, 2014; Bajqinca, 2019; SOU, 2019:18).

As for the theme “equal rights and opportunities”, the MU “right to equitable education” disappears altogether after 1980 and “equal opportunity for mother-tongue maintenance” is only mentioned once after 1982, namely in the Language Act of 2009. The only MU that is present in all periods is “right to one’s MT”, although it is only mentioned in connection with the NMLs in the final period.

The most important change in the theme “ethnocultural heritage” is owing to the increased focus on NMLs after 2000. This is visible in the emergence of several new MUs after 2004 such as “community participation”, “group identity”, “transmission of culture” and “reflecting on the origin and status of the language”, which are exclusively used to motivate support for NMLs and which account for the marked expansion in this theme’s MUs, which also means a strengthening of the intrinsic resource orientation. While the first period focused on preserving the heritage of all minorities, the final period focuses mainly on the national minorities.

As could be seen in Table 3, the MUs of the theme “cultural contribution to society” have the latest mean year of introduction, which entails a strengthening of the extrinsic resource orientation in later periods. Accordingly, MUs such as “asset for society” and “the school’s internationalisation” emerge in the 1990s and last until the mid-2000s. The end of the third period is marked by another development that boosts this theme, namely Sweden’s new national minority policies. Thus, the main focus of the theme shifts from immigration-related issues to the protection, promotion and revitalisation of NMLs.

Within the language-as-problem orientation the occurrences for MUs that mention developmental or social problems regarding minority position (i.e. “managing minority strain” and “making multicultural belonging an enriching experience”) decrease after the early 1990s, while MUs referring to the academic performance and educational conditions in segregated areas become more prevalent after the year 2000. This suggests a shift within this orientation from early concerns related to semilingualism to later concerns regarding the educational risks of segregation.

Quantitative trends in the relative weights of the LPP orientations

Based on the years of emergence provided in the middle column in Appendix 2, 27 of the 45 MUs (60%) were introduced already in the first period 1974 − 1983. Nevertheless, new MUs continued to emerge in later periods; ten new units 1984 − 1996, seven new units 1997 − 2007 and one new unit 2008 − 2022. When the emergence of new MUs after the first period is examined, the fastest expanding LPP orientations turn out to be extrinsic resource with nine of its ten MUs being introduced after 1991, which is in line with what previous research had implied (Cabau, 2014; SOU 2019:18; Bajqinca, 2019). Intrinsic resource has also expanded, albeit less conspicuously, with six of its 21 MUs emerging after 1991. In contrast, the language-as-right orientation had all four of its MUs and the language-as-problem orientation all but one of its ten MUs present already in the initial period. The exception in the latter case is “improving socioeconomic conditions in segregated areas”, which was introduced in 2004.

The presentation of the results in Fig. 1 focuses on changes in the relative weight of the LPP orientations, i.e. with how many different MUs they are represented in relation to the other orientations’ MUs. The rationale behind this method is that each policy document has its own unique impact on policy and political discourse in a particular point in time. This impact is best captured by highlighting how dominant each orientation is in each document. Firstly, for every document it was computed what percentage of all MUs pertained to each attested orientation. Secondly, a period mean of these percentages was calculated based on all documents that are included in each period. The N values in parentheses at the bottom of Fig. 1 indicate the number of total MU occurrences for each period. Since Fig. 1 only includes period means and does not capture the variation among each period’s different documents, the periods’ orientation distributions were also compared through statistical tests in order to obtain more reliable comparisons. The results are presented in Appendix 3 and the following discussion relies on both Fig. 1 and Appendix 3.

Fig. 1
figure 1

Changes in the relative weight of the LPP orientations based on the distribution of their period means

Full size image

Figure 1 shows that all four LPP orientations were present already in the first period and have been sustained throughout the entire time span. This result is in line with Zúñiga’s (2016:341) statement that the different orientations are not sequential or discrete paradigms, as well as de Jong et al.’s (2016:208) conclusion that the orientations can coexist in the same time and place. However, the extrinsic resource orientation had a very meagre presence in the first period with only one MU “creating understanding for and solidarity with different peoples and cultures”. Hence, Boyd’s (2007:148) claim that MTI was not portrayed as a resource for society (i.e. extrinsically) before the 1990s does not seem to be entirely accurateFootnote 12. While Bajqinca (2019) had correctly pointed to the presence of MLs as intrinsic resources, like Boyd his focus on overarching discourses missed the presence of an extrinsic orientation in this period. Thus, the present study is probably the first one in the literature to point out that the extrinsic resource orientation goes back all the way to 1974.

Despite the continuity in the presence of all LPP orientations, considerable changes can be observed in Fig. 1, too. The only orientation without any statistically significant changes or tendencies throughout the four periods is language as intrinsic resource. Language as right experiences a significant dip after the first period, and its diminished weight persists unchanged in the subsequent periods. Thus, the analysis method in the present paper reveals that the comeback of the language-as-right orientation that Vuorsola (2019:372 − 380) identified regarding Finnish, and that Bajqinca (2019) suggested for NMLs more generally after 1999 was too weak to have a significant impact on Sweden’s overall orientation towards MTI.

Figure 1 suggests a gradual decline for language as problem, but this trend only reaches a significant tendency when the first and last periods are compared. This is confirmation for Salö et al.’s (2018) analysis that indirectly pointed to the presence of this orientation well into the 1970s, something that Bajqinca (2019) only attested for the period before 1967. However, the present study is probably first LPP study to reveal that the problem orientation continued to exist right until the last period with the MUs “self-esteem” and “academic performance” attested as late as 2019.

The only orientation that has increased its relative weight over time is language as extrinsic resource, as already been suggested in previous research (Cabau, 2014; SOU 2019:18; Bajqinca, 2019). Despite the fluctuations suggested by the means in Fig. 1, the statistical tests revealed that only the last two periods differ significantly from the first period and that the difference between the first and second periods only reaches a significant tendency. This means that Bajqinca’s (2019) claim of a shift to an extrinsic resource orientation already in the 1990s was premature.

Summary of the evolution of the LPP orientations

The results regarding the evolution of the motivations for MTI have shown that there is both continuity in the persistent presence of the different LPP orientations, and change regarding their relative dominance over time. With its tangible and consistent presence, language as intrinsic resource has been the backbone of Sweden’s LPP orientation towards MTI, while language as right has become less prevalent after the mid-1980s and language as extrinsic resource more prevalent after the late 1990s. Furthermore, the present paper has revealed that the extrinsic resource orientation actually goes back to the very beginning of Sweden’s multicultural reforms, albeit with a weak initial presence. Contrary to previous research, this study has shown that the language-as-problem orientation was neither confined to the period 1957 − 1966 (Bajqinca, 2019) nor unique for the early 1970s, when the Home-Language Reform was being conceived (Salö et al., 2018). In fact, this orientation has been present 1974 − 2019, i.e. nearly during the entire examined period, albeit with a gradual weakening accompanied by an internal focus shift from the mental threat of semilingualism to the socioeconomic threat of segregation. Again, in contrast to what had been implied in previous research (Vuorsola, 2019; Bajqinca, 2019) the analysis in this paper has illustrated that the language-as-right orientation has not made a significant comeback recently, because its revival after 1999 was largely restricted to the national minority languages. Hence, the overall evolution of the orientations shows that MTI was initially established based on a relatively even mix of compensatory (right and problem) and non-compensatory (intrinsic and extrinsic resource) orientations and that the non-compensatory orientations have become clearly dominant over time.

Theoretical implications

This section aims to answer the final research question regarding the ideological underpinnings of the LPP orientations. The results in Fig. 1 confirm the hypothesis that language as resource has been the dominant LPP orientation towards MTI in Sweden, when both its intrinsic and extrinsic senses are evaluated together. After all, this orientation is supposed to be the “antithesis” of language as problem, which is associated with assimilationism and subtractive bilingualism (Hult & Hornberger, 2016:33, 38). In this sense, MTI in Sweden could be viewed as a prime example of the language-as-resource orientation. However, beyond this general observation that seems to confirm how Ruiz (1984, 2010) and Hult and Hornberger (2016) regard the ideological underpinnings of the LPP orientations, there are also some new messages that should be taken home from the present paper, which are presented in the following sections.

Language as problem is not always detrimental to additive bilingualism

In Ruiz’s framework (Ruiz, 1984, 2010; Hult & Hornberger, 2016), language as problem is viewed as detrimental to additive bilingualism and ML maintenance. However, as discussed in several Swedish studies and most clearly articulated in Salö and Karlander (2022), the perceived problem of semilingualism nevertheless served to “shoehorn” MTI into Sweden’s education system. By explicitly linking the semilingualism discourse to the language-as-problem orientation, the present study has shown that this orientation need not always be detrimental but can be beneficial under special circumstances as those in Sweden. In fact, Salö and Karlander (2022:122) observe that “[f]or some, it seems, semilingualism may arise if mother-tongue instruction is provided, but for others, it may arise if mother-tongue instruction is not provided.” Considering Bajqinca’s finding that bilingualism was perceived as a problem in the period 1957 − 1966, what is unique about the evolution of Sweden’s MTI orientations is a reversal in the logic of the problem orientation. What the Swedish semilingualism discourse practically accomplished was to sustain the earlier problem orientation by reversing the role of MTI from aggravating the problem to mitigating it. Furthermore, the present study has argued that this orientation maintained its beneficial role to date, albeit at a lower level, thanks to the new perceived threat of academic failure in segregated areas. In the case of Sweden, the problem orientation appears to be ideologically underpinned by a social liberal concern for minority children in the period 1974 − 2022, in stark contrast to nationalist ideologies that typically underpin this orientation elsewhere (as in Sweden in the 1950 and 1960 s).

Language as extrinsic resource is not solely underpinned by neoliberal ideology

As mentioned earlier, Ruiz’s (1984) reliance on the extrinsic value of language as resource has been criticised for marketizing ML skills (Petrovic, 2005; Ricento, 2005). Bajqinca (2019:182 − 184) offers a similar analysis for MTI in Sweden regarding the period after 1989, when he discusses the discursive reduction of immigrant MLs to the home sphere and the portrayal of ML maintenance as the responsibility of the individual rather than the state as a consequence of Sweden’s market-liberal turn. This raises the question if the diachronic increase in the dominance of language as extrinsic resource in the present study is related to the increasing dominance of neoliberalism in Swedish politics since the 1990s (Kärrylä, 2021; Möller, 2019:269). One of the main findings seems to confirm this claim. Observing the statistically significant changes in LPP orientations, the increased dominance of the language-as-extrinsic-resource orientation has mainly come at the expense of the language-as-right orientation. This appears to parallel the shift in Sweden’s dominant political ideology from social liberalism to neoliberalism (Möller, 2019:266 − 267), according to which equality is defined more narrowly and economic rationality is foregrounded.

However, the precise motivations for MTI provided with the extrinsic resource orientation need to be considered here as well. Examining the ten MUs with this orientation, four of them regard the facilitation of multicultural integration, the affirmation of multiculturalism in society and creating international cultural understanding and solidarity. These are hardly about commodification, marketization or economic rationality in the neoliberal sense, but rather social liberal in character. Further four MUs concern linguistic and cultural diversity as well as the protection, promotion and revitalisation of the NMLs. Again, these do not really rhyme with neoliberalism. Rather, they resonate with Ruiz’s (1984) argument that languages are resources to be conserved. Regarding the period leading up to the Language Act of 2009, which made Swedish the de jure principal language of Sweden, Oakes (2005) speaks of a “nationalist revival”. However, this process did not result in the revaluation of Sweden’s national heritage in a strictly monocultural or monolingual sense, as the new national minority laws ended up compelling policymakers to include the five national minorities and their languages in Sweden’s cultural heritage. Since the conservation of the national heritage is an important pillar of conservative ideology, the latter four MUs could be argued to offer a conservative underpinning for language as extrinsic resource, albeit one with liberal undertones. In fact, only the remaining two MUs, “asset for society” and “internationalisation of the school”, invoke the economic utility of MTI for Sweden’s international contacts in a way that could be associated with neoliberalism. Moreover, in the present corpus the significant rise in the extrinsic resource orientation is attested after 1997, rather than in the period 1989 − 1999 in Bajqinca (2019), which fails to support the timing of his market-liberalism hypothesis.

All in all, contrary to Petrovic’s (2005) and Bajqinca’s suggestions (2019), the present study does not support the view that it is mainly neoliberalism that underpins language as (extrinsic) resource. Instead, a closer examination of the timing of the surge of this orientation and the precise motivations behind it reveal that it was underpinned in Sweden by a mixture of liberal (both social liberal and neoliberal) and conservative ideologiesFootnote 13. As clarified by Ruiz (2010), there are several inroads to language as (extrinsic) resource (see also de Jong et al., 2016:202 on the multidimensionality of language as resource). Although he admits that this orientation does easily lend itself to a narrow utilitarian interpretation, it should nevertheless be viewed as broader and multifaceted (Ruiz, 2010; Hult & Hornberger, 2016). Hence, the present paper corroborates the view that language as (extrinsic) resource is not inherently neoliberal, as it illustrates that this orientation can be underpinned by social liberalism, neoliberalism as well as conservatism depending on the circumstances.

Ruiz’s LPP framework is still relevant but needs an ideological dimension

Ruiz published his seminal paper in 1984, shortly before the critical and ethnographic turn in LPP research (Tollefson & Pérez-Milans, 2018:7 − 8). Although his framework includes some critical elements, it was mainly conceived with the structuralist mind-set of the early LPP studies, which focused on the macro level, especially on the role of the state (Tollefson & Pérez-Milans, 2018:4 − 5). As implied by Kaveh (2023), these aspects make Ruiz’s framework somewhat ill-suited for the kind of micro-oriented studies that have become popular in LPP research of late. Does this mean that Ruiz’s framework is no longer relevant or useful, four decades after its inception?

In my opinion, the outpouring of appreciation for the framework as evidenced by the special volumes and journal issues dedicated to Ruiz following his death in 2015 (Fránquiz et al., 2016; Hornberger, 2017) testify to its continued relevance today. The present study joins in the assessment that there is a continued need for such LPP approaches that link issues that lie at the very heart of language (its multifaceted nature and social value) to macro-level language policy. However, as several critics have remarked before (Ricento, 2005; Petrovic, 2005; Zúñiga, 2016; de Jong, 2016; Kaveh, 2023), Ruiz’s orientations lack a clear ideological dimension. Therefore, Ruiz’s framework is most useful when it is complemented with an analysis of power relations. The present paper has tried to address this issue by emphasising the orientations’ links to political ideologies, and that these are not as straightforward as implied by Ruiz’s original paper. The preceding discussion has demonstrated that depending on the particular circumstances, language as problem can serve both nationalist and social liberal agendas, while there are social liberal, neoliberal and conservative inroads to language as extrinsic resource.

What can we then learn from the ideological complementation of Ruiz’s orientations that was applied in the present paper? The policy actors behind the key documents that were examined in this study (commissioners and experts in governmental inquiries, the National Agency for Education as well as politicians and civil servants) have essentially held a steady course − in the sense of MTI’s enduring existence − despite the economically and politically turbulent conditions of the past five decades. They have weighed concerns for minority children and national minorities against the needs and resources of the whole society, and have adapted to new national and supranational legislation affecting MTI. From this perspective, the interests of minority groups have clearly been taken into advisement regarding MTI in Sweden. At the same time, MTI was downscaled after 1991, and only limited measures have been undertaken to improve it in later periods despite numerous academic studies, governmental inquiry reports as well as several rounds of national and supranational evaluations that have pointed to problems in its implementation and come up with possible solutions. Thus, it is equally clear that the educational needs and interests of the minority communities, which generally have low socioeconomic status in Swedish society, have been given low priority by the politicians.

In summary, the attested combination of LPP orientations in Sweden has been successful in justifying the initial establishment of MTI as well as its continued (downscaled) existence but not its full implementation or improvement. How can this somewhat paradoxical outcome be interpreted through the additional lens of political ideologies discussed in this paper? From the mid-1980s onward, the compensatory orientations language as right and language as problem have been on a downward trend. Perhaps, it was the pressing need to compensate minority groups that had originally compelled multiculturally minded policy actors to cross the crucial threshold towards establishing large-scale MTI (as implied by the “shoe-horning” metaphor in Salö & Karlander, 2022). Following the same logic, the subsequent weak implementation of MTI can be understood in light of the weakening of Swedish multiculturalism (in the face of rising neoliberalism and nationalism) and of the simultaneous downward trend in the two compensatory orientations. In other words, a strong language-as-resource orientation after the mid-1980s may have been necessary for maintaining MTI but it has not been sufficient for the full implementation or further improvement of MTI given the new dominant political ideologies of the last decades.

Conclusions

This paper has examined the evolution of Sweden’s language-planning orientations towards mother-tongue instruction based on a new diachronic corpus of key policy documents stretching from 1974 to 2022. The innovative three-tiered methodology with motivational units, motivational themes and language-planning orientations facilitated the analysis of the data from several different angles as well as a statistical comparison of the four periods, which the five examined decades were divided into. As suggested in previous research, the results show that the language-as-resource orientation has played a central role in justifying both the establishment and the maintenance of mother-tongue instruction in Sweden. This main finding confirms Ruiz’s (1984) original assertion that the language-as-resource orientation is crucial for additive bilingualism and minority-language maintenance, as it frames language in positive terms as a valuable societal resource beyond the compensatory approach that language as problem and language as right entail (cf. Hult & Hornberger, 2016).

At the same time, the applied periodization principle based on the establishment and subsequent evaluation of new policies regarding integration, minorities and language has revealed that major legal and policy reforms are closely connected to changes in language-planning orientations. It was demonstrated that language-as-right played an important role in the introduction of the Home-Language Reform of 1977, which legally obliged municipalities to offer mother-tongue instruction upon certain criteria. Similarly, the language-as-resource orientation in the extrinsic sense had a central place in the new policies regarding national minorities and their languages starting in 1999 as well as in the new explicit language policies, both of which were enshrined in law in 2009. This finding can be taken as a corroboration of the later statement by Ruiz (2010) and Hult and Hornberger (2016) that the orientations language-as-right and language-as-resource complement each other.

The present paper has also provided new and partly unexpected findings regarding the underpinnings of the orientations in terms of political ideology. Contrary to some claims in the literature, the examined case of mother-tongue instruction in Sweden revealed that language as extrinsic resource is not necessarily underpinned by neoliberalism alone, as there are also social liberal and conservative inroads to this orientation. Most surprisingly, the analysis in this paper has illustrated that the language-as-problem orientation need not always be detrimental to bilingualism and minority-language maintenance, as argued in previous research. Foremost, the case of the semilingualism discourse in Sweden in the 1960 and 1970 s but also the later discourse about the educational risks of segregation since 2000 can be viewed as prime examples of the beneficial role that the language-as-problem orientation can play under special circumstances.

Thus, this study contributes to language planning and policy theory by providing a more explicit and nuanced discussion of the connections between Ruiz’s (1984) three orientations in language planning and different political ideologies. From a practical point of view, the main findings in this paper show that although the language-as-resource orientation has played an indispensable role in supporting additive bilingualism in Sweden, the endurance of MTI for nearly five decades has not been all about language as resource. Both the language-as-right orientation and the language-as-problem orientation has complemented language as resource in crucial ways. Moreover, this paper argues that a strong language-as-resource orientation has not been sufficient for improving the implementation of MTI. Hence, in order to support additive bilingualism policymakers and minority communities should not solely rely on a language as resource as their language planning orientation but on a mix of several orientations that best suits their particular circumstances.