Facilitating or compromising inclusion? Language policies at Swedish higher education institutions as workplaces

1.1 Group identity and climates for diversity and inclusion at workplaces

The globalization of the economy has made workplaces in many countries increasingly diverse, including HEIs. Mor Barak (2017: 189–190) clarifies that diversity at workplaces is not about individual uniqueness but rather about “belonging to groups” and “employment consequences as a result of one’s association within or outside certain social groups”. One of the most serious problems associated with workplace diversity is the issue of inclusion, which refers to employees’ sense of being involved in both formal and informal processes, including decision-making and even simple access to information (Mor Barak 2017; Mor Barak and Cherin 1998; Travis and Mor Barak 2010).

Mor Barak (2017) draws on social identity theory (Tajfel and Turner 1986) to explain the process that leads to exclusion. Crucially, people tend to classify themselves and others into in-group and out-group, usually based on social categories such as gender or race. In-group identification is only possible through comparison with an out-group, leading those that self-identify as group members to attempt to maintain positive group identity as superior to the out-group. Thus, the members of a group with higher social status may discriminate against and exclude those perceived to be out-group members, while members of the lower-status group are likely to make compensatory efforts to gain membership in the desired group (Tajfel and Turner 1986), which can affect their well-being negatively (Mor Barak 2017). Discrimination by the higher-status group may include “the distribution of resources and privileges … based on group membership”, instead of individual “characteristics such as level of education, commitment and job-related skills” (Mor Barak 2017: 204).

To understand employees’ perceptions of inclusion, Mor Barak (2017) introduces two constructs: climate for diversity and climate for inclusion. Climate for diversity refers to employees’ shared perceptions of the extent to which their organization emphasizes representation of diverse groups and treats everyone in the same way regardless of their background (Mor Barak 2017: 491; McKay and Avery 2015). Meanwhile, climate for inclusion refers to employees’ shared perceptions of the extent to which policies and practices recognize employees’ unique qualities, facilitate “a sense of belonging”, and encourage “involvement in organizational communication, decision-making processes and informal interactions” (Mor Barak 2017: 491). Climate for diversity may therefore be considered as a matter of representation and opportunity (e.g., for advancement), while climate for inclusion may be considered as participation and belonging. Mor Barak notes substantial research suggesting that organizations that successfully create strong and positive climates for both diversity and inclusion allow employees to be part of workgroups and teams, with an equal chance for successful work performance regardless of their background.

Mor Barak (2017) proposes that the most effective, consistent way to achieve inclusion in the situation of workforce diversity and ultimately a supportive climate for diversity and inclusion should be “from the very top” of management including policy making, “reinforced by behaviors at all levels of the organization” (366). She characterizes inclusive leadership as recognizing individuals according to their skills and competence, valuing different perspectives, creating goals shared by all, and actualizing everybody’s active participation in work activities.

1.2 Language as a diversity category for inclusion and exclusion in multilingual workplaces in Nordic countries

In multilingual workplaces, different language repertoires may contribute to language-based in-groups that have varying statuses. The statuses of the languages themselves are often connected to naturalized language ideologies, such as one-nation-one-language ideology and valorization of English for internationalization (Lønsmann 2014; Lønsmann and Kraft 2018; Lønsmann and Mortensen 2018). Dual or double monolingualism (Holmes 2020, 2023) also plays a role, dictating strict separation of two languages and regarding bilingual or multilingual competence as the ability to use each of two or more languages in a monolingual context. When monolingual contexts are valued in this way, language-based in-groups are especially likely to be created, with those who do not speak the language chosen in that context as out-group, threatening a positive climate for inclusion. However, the workplace’s overall multilingual context means that excluded out-groups will vary depending on the requirements of the specific context.

For example, Lønsmann (2014) found that employees in a company in Denmark who did not speak both Danish and English fluently experienced exclusion. Thus, bilingual speakers constituted the dominant in-group, but out-groups varied by situation: internationals who did not speak Danish were often excluded from informal networks and socializing, potentially undermining their sense of belonging and thereby the climate for inclusion, and Danes with weaker English skills were excluded from some work activities, threatening both climate for diversity and climate for inclusion. Similarly, Lønsmann and Kraft (2018) report that, while multilingual practices are crucial for workplaces with linguistic diversity, international companies in Denmark and Norway often favored local language native speakers. This resulted in the workplace hegemony of an in-group of local employees or other Scandinavian language speakers and exclusion of internationals with insufficient local language skills.

In a higher education context, speakers of the local language may likewise constitute a primary in-group, leading to exclusion of those who lack mastery of that language. In Sweden, Holmes (2020) reports exclusion experienced by three non-Swedish-speaking international academics. An international researcher noted that he could not join certain committees that he volunteered for because he did not understand Swedish. Even more alarmingly, emails from the building manager – relating to safety issues, for example – were in Swedish only, often resulting in his putting them aside for later. Meanwhile, Kirilova and Lønsmann (2020) found that although an international lecturer at a Danish university did not initially see himself being limited at work by not speaking Danish, upon reflection he acknowledged that speaking Danish was necessary for further advancement. The differential treatment of employees based on language skills is likely to negatively impact the climate for diversity, while the limits to involvement based on language skills is likely to damage the climate for inclusion.

While group distinctions and the exclusion of certain groups in multilingual workplaces largely involve naturalized ideologies and may be imposed from the top, we also observe voluntary, bottom-up inclusive strategies by those we see as members of the dominant group to include those who would otherwise be out-group (e.g., Holmes 2023; Negretti and Garcia-Yeste 2015; Salö 2015, 2022). Holmes (2023) reports that support staff composed emails in both Swedish and English to avoid excluding international faculty, although not receiving help or recognition for the extra work. Meanwhile, some participants in Salö (2015, 2022) and Negretti and Garcia-Yeste (2015) showcase flexible translanguaging between languages to include non-Swedish-speaking colleagues. Such approaches may allow the greatest inclusion, but require more effort by the in-group, and greater emphasis on including one potential out-group can result in exclusion of another.

1.3 Constraints on workplace inclusion at HEIs in local and national language policies

Workplace exclusion based on language as a diversity category at Swedish HEIs may be related to language policies (see Holmes 2020, 2023; Salö et al. 2022). Particularly, as Karlsson (2016) and Karlsson and Karlsson (2020) acknowledge, assigning Swedish as the natural language for administration may lead to the exclusion of non-Swedish-speaking international academics from administration, preventing the top-down, consistent facilitation of positive climates for diversity and inclusion recommended by Mor Barak (2017).

Such possible exclusion at HEIs may be presented as inevitable (Karlsson and Karlson 2020) for compliance with the Language Act (Ministry of Culture 2009: 600), which associates Swedish with national responsibilities (Boyd 2011) and expects governmental bodies to operate in Swedish. Referring to Språk för alla – förslag till språklag [language for all - suggestions for language law, our translation] (Prop. 2008/09:153: 29–30), an accompanying document to the Language Act, Karlsson and Karlsson (2020) argue that, while HEIs are exempt from the act in research and teaching to allow global competition, their administration needs to be done in Swedish as it belongs to the Language Act’s core areas.

However, a close reading of Språk för alla – förslag till språklag can yield an understanding different from Karlsson and Karlsson’s. First, the Language Act’s core areas include those subject to public access, such as “the political decision process, court proceedings, verdicts, minutes, decisions, regulations, activity reports and other similar documents” (Prop. 2008/09:153: 29, our translation). Thus, these core areas may not involve all administration at HEIs, but mostly written documents that should be accessible to the public. In addition, the document also specifies that communication within governmental bodies can take place in languages other than Swedish, for example in Swedish agencies abroad when some employees cannot speak Swedish. The document notes that “what can be expected concretely from individual authorities – e.g., universities and university colleges – may be decided partly based on the Language Act, and partly based on their special conditions” (30, our translation). Therefore, HEI policies that may constrain inclusion in administration through their monolingual approach may not be as firmly grounded as Karlsson and Karlsson (2020) suggest.

Moreover, besides the question of the extent to which HEIs’ language policies should be based on the Language Act, the act itself has been problematized for inconsistency that may compromise national democracy (Boyd 2011). The establishment of the act was centrally motivated by the discourse of ‘domain loss’, the concern that Swedish should be protected from the threat of English, particularly in academic domains (Boyd 2011; Cabau 2011; Salö 2014). However, following market logic, its accompanying documents permit the use of English in HEIs’ research and teaching, the key areas in which ‘domain loss’ is relevant. At the same time, its reinforcement of the status of Swedish in society overall may, as Boyd (2011) discusses, create “grounds for discrimination against people with first languages other than Swedish” (32). Moreover, the act’s approach to multilingual Swedish society undermines the role of multilingualism for enabling communication (Boyd 2011). These concerns about Swedish society in relation to the Language Act clearly apply to HEIs whose language policies incorporate the principles of the act.

1.4 Towards language policy for democracy of inclusion: Shohamy’s (2006) expanded approach

The above discussion suggests challenges for language policies to facilitate inclusion while protecting against English linguistic imperialism (Phillipson 1992). In this section we consider how an expanded approach to language policy may begin to address these challenges. While language policy often promotes dominant social groups’ interests, creating and maintaining social inequality (Tollefson 2006), language policies established by dominant social groups can be resisted in language practices and reclaimed to be democratic and inclusive (Shohamy 2006; Tollefson 2006). Discrepancies between declared language policy and de facto language policy mean that a language policy that seemingly intends to be inclusive, e.g., promoting multilingualism, can merely be ‘lip-service’ lacking specific plans (Shohamy 2006). On the other hand, actual language practice can be more aligned with bottom-up policy rather than declared language policy imposed by the nation or other higher authorities in neglect of real-life language situations (Baldauf 1994).

Building on Spolsky’s (2004; also see 2022) broad definition of language policy as comprising language ideologies, language practices, and language management, Shohamy (2006, 2015) suggests an expanded approach to language policy, to envision how language policy can serve for democracy of inclusion in a multilingual community. According to her, language policy includes “mechanisms, policies and practices as well as the set of negotiations, conversations and battles that take place among them” (2006, xv). Mechanisms are devices such as language laws, rules, language educations and tests, and language use in public spaces, which create, perpetuate, and spread language ideologies that influence language practices. Mechanisms are often used by those in authority and power, since power grants greater access to different channels, knowledge of legal systems, and financial resources. Nevertheless, the hope for inclusive language policy lies in the fact that mechanisms “can be used by all groups in society, top-down and bottom-up” (54).

According to Shohamy (2006, 2015), in a democratic society, all who are affected by language policies should be able to participate in creating policy for greatest inclusivity. Language policies declared top-down often claim to include people through uniformity but are discriminatory in nature, privileging one group’s variety over others. Instead, achieving democracy of inclusion in language policy requires intensive negotiation among people from different language backgrounds. In addition, a democratic approach reconceptualizes language beyond named language boundaries, instead considering language in terms of repertoires facilitating communication.

1.5 Research questions

Extending previous research that has suggested that HEIs’ language policies might contribute to the exclusion of some employees, the current study provides a critical analysis of how workplace inclusion in view of linguistic diversity is considered in language policy documents, or ‘declared language policy’ (Shohamy 2006). We investigate how the policies may facilitate or compromise positive climates for diversity and inclusion (Mor Barak 2017), addressing the following research questions:

1. To what extent do Swedish HEIs as workplaces with linguistic diversity emphasize the value of workplace inclusion in their language policies?

2. How do the policies likely facilitate or compromise inclusion? How do HEIs justify and motivate their policies relating to language choice, requirements, and support, which are relevant to inclusion?

2.1 Materials

The analyzed material comprised all language policy documents from Swedish HEIs that were posted publicly or that were sent to us upon our request, for a total of 21 policies: eleven from universities, nine from university colleges (högskolor), and one from a department at Stockholm University, the one such example that was available to show how a HEI’s policy is further developed for a specific work situation. Among the 20 HEIs, nineteen are public institutes; the one private institute, Chalmers University of Technology, is also financed by tax and governed by Swedish higher education regulations.

All documents are written in Swedish, with some HEIs also providing English translations. Although the first author has sufficient competence in Swedish to analyze the documents in Swedish, we used English translations for efficient analysis, as both of us have higher competence in English. To do so, we translated the Swedish-only documents, using Google Translate with edits by us. A Swedish native speaker who has worked as an academic at Swedish HEIs compared our final translations with the Swedish originals to ensure accuracy. We similarly double-checked HEIs’ English translations and edited one policy (Stockholm University) for analysis where the English translation did not match the original Swedish, since the Swedish version is considered the official version. We provide the Swedish versions of the extracts in our supplementary document available online.

The total length of the 21 documents in Swedish was 27,965 words (mean = 1,332), although length and amount of detail varied widely, e.g., with Uppsala University’s at 3,255 words compared to Mid Sweden University’s 343 words. Table 1 presents each document’s publication year, as well as the source of the English translation. Some HEIs provide two publication years when the documents were initially published in one year and revised in another.

2.2 Our approach and positionality

According to Tollefson (2006), research on language policy can be critical in three ways: being “critical of traditional, mainstream approaches”; “aim[ing] at social changes”; and/or being “influenced by critical theory” (42). Our approach to investigating HEIs’ language policies is critical mainly in the second sense, with the intention to suggest changes to problematic policies (Shohamy 2006). To that end, as described below in detail, we read the material thoroughly to gain a fully developed understanding, based on which we critically examined and questioned naturalized assumptions.

Our own positionality naturally influences what we are likely to observe in terms of such naturalized assumptions. The first author’s language repertoire includes Korean as the first language, English as the second, and Swedish as the third. She came to Sweden in 2017 and began to work at a college that year. The language situation at the workplace was intense for her as a newcomer, where full participation required Swedish competence. As she used English to work full-time, progress in learning Swedish was not as speedy or consistent as she expected, despite the strong wish to learn the language to be more included in the workplace. Her lived experiences of workplace language situations led to reflection on language-related workplace inclusion, giving her the initial impetus for carrying out this study.

The second author is a linguistically privileged speaker who grew up and works in the US speaking a variety of English that passes for ‘standard’ there. Research interests in language bias contributed to her interest in the project after hearing about the first author’s experiences, in spite of her own lack of experience with Swedish HEI language policy. Given that her previous research has mainly attempted to combat forms of linguistic imperialism in the US, she was challenged to consider how resisting the encroachment of English can be balanced with facilitating inclusion in international workplaces. She has studied and used other languages previously, most notably German, Arabic, and now Swedish.

2.3 Data analyses

As presented in the introduction, we used inclusion at diverse workplaces (Mor Barak 2017) and language policy for democracy of inclusion (Shohamy 2006) as our theoretical lens for examining language policies, focusing on the extent to which and in what ways the policies may support or limit employees’ equal opportunity to get involved in work activities and decision-making processes at different organizational levels, as well as to have access to work-related information and resources (Mor Barak and Cherin 1998). We first engaged in inductive analysis to familiarize ourselves with the policies and then moved to deductive analysis through the theoretical lens.

First, we imported all documents into our NVivo project file and carried out conventional, inductive content analysis (Hsieh and Shannon 2005) to gain a comprehensive understanding of the content, familiarize ourselves with the textual features, and extract the relevant content for our research topic. Through the inductive analysis we identified six content areas, of which “values”, “language choice”, “language requirements”, and “language support” were relevant to our study. “Values” was defined as the guiding principles the HEIs described as relevant to their policies. “Language choice” was used for all text explaining what language(s) would or could be used in a given situation, or how such a decision would be made. “Language requirements” was defined as statements about expectations or requirements for language skills, while “language support” included both support for achieving such skills and for any assistance provided where they might be lacking. We excluded from the data content irrelevant to academic staff as employees.

We then further identified subordinate areas in the four content areas. In this step, some content was double- or multiple-coded into different sub-areas. For example, “With good command of several languages, our students and staff become attractive and competitive nationally and internationally” by Umeå University was coded into both “promoting internationalization” and “multilingualism, linguistic diversity” under the content area “values”.

The inductive content analysis was followed by a more focused, deductive analysis with our lens of inclusion. This stage focused on the four main areas identified through the inductive analysis that were relevant to addressing our research questions. Specifically, we observed the extent to which the policy documents emphasize inclusion and participation as well as how they may facilitate or compromise inclusion in their discussion of language choice, support, and requirements in different areas such as internal webpages for employees, general faculty/staff meetings, decision-making meetings, and documentation of decisions. In doing so, we critically examined discourse patterns, with close analysis of how HEIs interrelate different propositions to motivate and/or normalize inclusion or exclusion in their policies (see Crombie 1985 for semantic interrelations of propositions). For example, we observed that University West seemed to normalize exclusion through using the proposition “Ministry of Culture (2009: 600) emphasizes the authorities’ responsibility for use and development of the Swedish language” (1) as the (implicit) justifying ground for “Employees who are not fluent in Swedish have poorer opportunities to participate in the workplace” (3), positioning the exclusion of certain employees as inevitable. Meanwhile, we saw that Royal Institute of Art (3) assigned a means-purpose relation between the proposition, “it is important to think about which language you choose to speak” (means) and the proposition, “in order not to exclude anyone from the community of the conversation” (purpose), to motivate inclusive language choice.

Our analytic commitment involving both full, inductive understanding and deductive, critical evaluation of the policy documents required contextualizing and scrutinizing the policy documents by means of intertextual reading of the relevant literature (Björkman 2014; Johnson 2015), including Swedish laws as well as the studies reviewed in 1.1–1.4.

3.1 To what extent do Swedish HEIs emphasize the value of inclusion in their language policies?

Our inductive content analysis of the policies’ opening statements commonly found explicit discussions of universities’ values that were presented as related to their language policies. The two most highlighted values were “promoting internationalization” (mentioned in 85 % of the policies) and “protecting Swedish in accordance with the Language Act” (95 %). Other frequently mentioned values were “multilingualism and linguistic diversity”, “correct and/or plain language use”, and “public responsibility”.

On the other hand, explicit mentions of inclusion and the related concept of workplace democracy as values, which can create strong climates for diversity and inclusion (Mor Barak 2017), appeared exceptionally only in University of Gothenburg’s policy as seen in Extract 1, although the context of the statement suggests that it is focused on underrepresented students and staff and students with disabilities.

Interestingly, the emphasis on following the Language Act means that the need for access to important official documents is a value that HEI policies generally address, but mainly for the public; they do not necessarily address the need for all employees’ equal access to the same information for workplace inclusion. As seen in Extract 2 below, HEI language policies stipulate that highly important information should be written in Swedish for public access as required by the Language Act.

Although producing important documents in Swedish for public access is important, it should also be noted that Swedish “may, in fact, have lost its importance as a means for inclusion” within HEIs as internationalized workplaces with linguistic diversity (Karlsson and Karlsson 2020: 82). Contrasting with their emphasis on equal information access for the public, various HEIs, according to their language policies, do not translate important documents open to public access into English or other languages (e.g., Royal Institute of Art), do in a limited way (e.g., Karlstad University, Linköping University, Malmö University, and Uppsala University), or often do not specify at all whether such documents would be translated (e.g., Lund University, Mid Sweden University, and University West). Consequently, non-Swedish-speaking staff may have no or limited access to their workplaces’ important information, jeopardizing the climate for inclusion.

Returning to the more commonly explicitly stated values, the two most prioritized values, “protecting Swedish in accordance with the Language Act” and “promoting internationalization” seem likely to constrain the inclusion of employees. Specifically, to address these values, the parallel language principle (Björkman 2014; Bolton and Kuteeva 2012; Hult and Källkvist 2016) is invoked, making high competence in both Swedish and English an essential condition for being included in the work community, similar to the situation found by Lønsmann (2014) and Lønsmann and Mortensen (2018) at international companies in Denmark. The pursuit of internationalization and adherence to the Language Act can result in rationalizing and normalizing the exclusion of employees who are not highly competent in both languages, whether non-Swedish-speaking international staff or domestic staff with limited English competence. For example, although in Extract 3 Malmö University highlights the importance of an inclusive work environment ‘regardless of language skills’, the policy addresses this problem of language skills by presupposing that employees already have or will quickly acquire them.

Both Swedish and international employees can be exposed to the possibility of exclusion or limited participation if they are not at least Swedish-English bilingual. While pursuing the value of internationalization, interpreted as requiring English (Hult and Källkvist 2016), for example, Swedish HEIs may create situations where local employees with limited English skills are excluded. Nevertheless, as acknowledged in policies at University West (Extract 4), Lund University, and Royal Institute of Technology, emphasis on Swedish competence is especially likely to restrict the participation of or even exclude international employees who do not speak Swedish from decision-making processes (as Salö et al. 2022 observe), thus negatively impacting the climate for inclusion, and damaging the climate for diversity by hampering representation and opportunities for advancement.

In the context of the policy documents, it is clear that the value of protecting Swedish in accordance with the Language Act is provided as a justified ground for lessening the degree of inclusion, especially through accepting the exclusion of non-Swedish-speaking employees from managerial and administrative assignments. University West, like most of the HEIs, begins their language policy by declaring that “The official communication language at the university is Swedish. The Language Act (2009: 600) emphasizes the authorities’ responsibility for use and development of the Swedish language” (University West, 1). However, as we discussed in 1.3, governmental commentary accompanying the Language Act suggests that the degree to which the Language Act is to be complied with may be adapted as needed in different situations. Moreover, the Language Act may provide grounds for language-based discrimination (Boyd 2011) by privileging the first language speakers of the dominant language (Shohamy 2006).

3.2 How do the policies likely facilitate or compromise inclusion? How do HEIs motivate these policies?

As seen in the discussion of the first research question, inclusion did not appear to be one of the explicitly prioritized values in Swedish HEIs’ language policies. However, a close reading of the policy documents shows how HEIs try to use language choice, requirements, and support so as to enable all staff and students to participate in work and study. We identified six approaches to try to facilitate inclusion:

1. Providing immediate language support

2. Encouraging bottom-up, flexible language choice

3. Emphasizing individuals’ rights to choose what language they use for speaking and writing

4. Taking a top-down, monolinguistic approach to the parallel use of Swedish and English for different functions and situations

5. Requiring staff to be highly proficient in both Swedish and English

6. Offering unspecified language support

Interestingly, as we discuss below, although seemingly intended to increase participation and inclusion, options 4, 5, and 6 may actually compromise such goals. The third option – emphasizing individuals’ rights to choose the language they use – can be helpful (Karlsson and Karlsson 2020), but is not unproblematic, as discussed below.