Whose English? Whose Diversity? Towards a More Holistic Understanding of Global English

In this paper, I call for an increased dialogue between different branches of Applied Linguistics. After discussing conceptualizations of the English language in the context of Applied Linguistics research and the ways in which English is perceived to form part of linguistic diversity, I argue that further dialogue is needed in order to gain a more holistic understanding of English and its multiple facets. The use of English is diverse and can also form part of the social justice agenda. At the same time, as our enquiry into the global spread of English moves beyond essentialized abstractions and metaphors, it is also necessary to critically question some new orthodoxies, e.g. the direct connection between translanguaging and social justice. Overall, I agree with Hultgren that widening our analytical lens is important. In particular, the suggested focus on the socio-material aspects is needed for drawing our attention to less known contexts of language use and to under-represented study participants. Before we tone down language in our scholarly inquiry, we still need to hear more voices.


Introduction
In her opening statement to this Special Issue, Hultgren (2020) challenges the assumption that the global spread of English is a threat to social justice and questions a wide-spread belief that the discipline of Applied Linguistics plays a central role in finding a solution to this problem. She argues that the spread of English is only a symptom of larger societal issues and urges us to take a more holistic approach and to look beyond language to identify causes of injustice and inequality. In particular, Hultgren underscores the importance of socio-material factors in shaping language uses. To conclude, she calls for increased researcher reflexivity among applied and sociolinguists, and for broadening the analytical lens and scope of our disciplinary inquiry.
Reading Hultgren's statement has made me reflect on my own positioning in the field, on the current ways of conceptualizing English, and on possible ways of overcoming the stigma that has developed around this object of inquiry. This article sums up these reflections. Considering my disciplinary and institutional affiliation, it is little surprising that my response to Hultgren's position statement is supportive. Having done research in the sub-fields of Applied Linguistics mentioned in Hultgren's article (e.g. ERPP, EMI and LPP), I truly welcome new perspectives on, and fresh insights into, how to overcome what Hultgren (2020) describes as "the disciplinary impasse". After discussing conceptualizations of the English language in the context of Applied Linguistics research and the ways in which English is perceived to form part of linguistic diversity, I call for an increased dialogue between different branches of Applied Linguistics. This continued dialogue is needed in order to gain a more holistic understanding of English and its multiple facets. Above all, taking into consideration the socio-material aspects can contribute to drawing researchers' attention to less known contexts of language use and to under-represented study participants.

English in Applied Linguistics
In her opening position statement, Hultgren mentions multiple facets and conceptualizations of English, such as a lingua franca (e.g. Jenkins 2017), a local practice (e.g. Pennycook 2010), or multiple distinct varieties of English around the globe (e.g. Kachru, 1990). At the same time, she points out that the critics of the global spread of English tend to perceive this language in largely homogenous terms. There is something common to the long list of metaphors describing the spread of Global English: in all its monstrous personifications, English-presumably standard English-is largely perceived as an abstract entity, an artifact linked to various geopolitical trends and actors and detached from its everyday users, particularly those outside the anglophone world to whom it poses a threat. To a large extent, it is this conceptualization of English-or indeed any other language-that makes it possible to use it as a "red herring" or, I would add, a "scape goat" to distract our attention from other issues involved in promoting social justice.
Nearly three decades after the publication of thought-provoking and influential works such as Linguistic Imperialism (Phillipson 1992) and The Cultural Politics of English as an International Language (Pennycook 1994), portraying English as a threat to other languages has become a bit of a trope that dominates academic and public discourses. At the same time, as Pennycook (1994: 262) showed some twenty-five years ago, English functions not only as a language of "imperialism" but also a language of opposition and opportunity, e.g. in the aftermath of the fall of the Berlin Wall and disintegration of the USSR. I happen to have experienced this opposition on the other side of the Iron Curtain in my hometown of St Petersburg, the epicenter of underground culture and music in the former USSR. The same applies to some postcolonial contexts. For example, in 2010, a temple to the English languagedeified as the goddess Angrezi Devi-was consecrated in the Indian village of Banka. Despite the legacy of British colonialism, the English language is seen by the repressed Dalit groups as 'a route out of social, intellectual and economic depression' (Krishnan 2017).
In the meantime, many of the questions concerning language and inequality raised by Pennycook in 1994 remain to be addressed, as evidenced by its re-release in 2016. Applied linguists involved in English language teaching should maintain their critical stance and reflect on their own practices and the role of educational institutions in promoting political agendas, e.g. neoliberalism. As my anonymous reviewer pointed out, neoliberalism is particularly subtle in its construal of English as a form of diversity, flexibility, and cosmopolitalism, and researchers still need to find ways of identifying how power manifests itself in such positive claims. For us, i.e. educators and researchers based in the Nordic countries, this is particularly important today as student populations in classrooms located in major metropolitan areas have diversified as a result of both privileged and needs-driven mobility. Students, and often their teachers, may come from different socioeconomic backgrounds and may not share the same first language. This poses new challenges to educational institutions. English-in its multiple facets-is implicated in these challenges: as an academic subject, as a functional lingua franca, and as an abstract entity in language policies.
Despite ongoing debates about its status and the introduction of protectionist language policies in the Nordic countries, English has been gaining further ground in different spheres. As Hultgren mentions in her article, today English has many recognizable faces in high-stakes domains: it is the language of scientists, politicians, artists, and business people. It is also widely used in communication supported by digital media, although, contrary to some earlier predictions, it does not dominate the internet (Kuteeva & Mauranen 2018). At the same time, English has faces that we may recognize as lay people but know much less about as researchers, such as that of a self-taught Afghani schoolboy who crossed Europe alone to seek asylum in Sweden (Resare Jansson 2018), or that of construction and domestic workers. Unlike the Englishes spoken in high-stakes domains, the latter kinds do not hold the same prestige, and we know less about their role in promoting or impeding social justice. Before we come up with any new metaphorical descriptions of English, we need to get out of our comfort zone and to consider all elements of the "English Language Complex" (Mesthrie & Bhatt 2008: 1-10), continuing our inquiry into how it interacts with other languages. Hultgren (2020) suggests that applied linguists should turn their attention away from language towards other, material and socioeconomic factors. Her challenge comes at a time when many branches of Applied Linguistics and Sociolinguistics have been questioning what constitutes the object of the discipline and challenging established conceptualisations of language (e.g. Canagarajah 2013, Otsuji 2015, Li Wei 2018). While I generally agree that Applied Linguistics would benefit from embracing a wider, interdisciplinary approach to "the theoretical and empirical investigation of real-world problems in which language is a central issue" (Brumfit 1995: 27), as we are venturing into other, less known realms, we still need to give language a chance and to further explore the complex interactions and contradictions between the different ways in which languages are used and perceived by their L1 and LX 1 users (Dewaele 2018).

English and diversity
Hultgren revokes Piller's argument that "linguistic diversity intersects with social justice" (2016: 5). English-in its multiple forms and 1 I adopt Dewaele's term LX as a shorthand for any additional language regardless of the order and manner of acquisition or total number of languages spoken by an individual. I am aware that this term also presupposes that languages are distinct and countable entities but, for now, I find it the most encompassing and easiest way to refer to different kinds of speakers of more than one language. manifestations-contributes to this diversity. Jenkins (2017) argues that English does not pose a threat to multilingualism since it is often used in combination with other languages. Canagarajah (2013: 68) conceptualizes English as "a translingual practice" in connection to less conventional uses which involve blending standard English with other varieties, codes, and semiotic resources. In other words, it is not only the use of standard varieties of English that is growing but also its hybrid uses, both in high-stakes domains (e.g. Mauranen 2012) and on the grassroots level, often in interaction with other semiotic resources (e.g. Pennycook and Otsuji 2015). If we view the spread of English through a Bakhtinian (1981) lens, it is not surprising that it has been accompanied both by an increasing use of its standard varieties in high-stakes domains (e.g. publishing, translation, the English Language Teaching industry) and by growing stratification on different levels, i.e. professional, generational, or local, leading to heteroglossic variation in the ways it is used by both L1 and LX speakers. This stratification leads to what Bakhtin (1981: 67) terms heteroglossia (the English rendering of Russian raznorechiye, literally "diversity of speech") referring to variation within a given language. Some language purists are concerned about a possible deterioration of the English language in the anglophone world. For example, the veteran broadcaster John Humphrys, a recently retired leader of the BBC Radio 4 Today programme, talks about "language obesity" resulting from an overconsumption of junk words (cited in Shariatmadari, 2019). Recent research into English as a lingua franca has shown how non-standard usages shape the development of English and contribute to language change (Mauranen and Vetchinnikova, forthcoming). In other words, English is varied and is changing rapidly, not least due to its global spread.
As mentioned above, the conceptualization of English as an artifact or an abstract, monolithic entity has made it possible to give it various metaphorical descriptions. Over the last decade, a great deal of research in applied and sociolinguistics has attempted to reconceptualize our understanding of language use by challenging the boundaries between languages, varieties, and codes, e.g. Pennycook's (2010) concept of language as a local practice or Li Wei's (2018) theory of translanguaging. This line of research has often laid blame for our understanding of languages as separate entities bound to specific locations and nation-states on movements and ideologies such as German Romanticism. At the same time, it must be pointed out that each discipline needs to have its own set of basic concepts, and the concept of a language has worked very well for general linguistics, e.g. from its early days of comparative philology to today's typological research. To this day, the more theoretical branches of linguistics have had no problem with a language as an abstract concept. The more applied branches of linguistics dealing with "real-world problems in which language is a central issue" (Brumfit 1995: 27) have been finding the focus on language alone less satisfactory because studying "real-world problems" requires a wider analytical lens, as Hultgren argues in her position statement.
In Applied Linguistics, equating languages with nation-states has been misguiding our understanding of the complexity of language uses by individual speakers and communities whose linguistic resources and practices may include different languages, varieties, and codes. The questioning of linguistic boundaries has also occurred in the study of world literatures, leading to different conceptualisations of "translingualism" (e.g. Helgesson & Kullberg 2018;Kellman 2000). As a pedagogical practice in language teaching, translanguaging is traced to the Welsh trawsieithu, to describe the use of two languages, Welsh and English, to promote the acquisition of Welsh (Jaspers 2018). Many foreign language teachers who share L1 with their students would recognize this common practice but translanguaging as a theoretical concept took off in Applied Linguistics research about a decade ago (e.g. García 2009;García & Li Wei 2014). Research on translanguaging and its pedagogical applications became part of the social justice agenda in the US (e.g. García 2009) and was also developed by Li Wei, e.g. to examine the language practices of Chinese students studying in the UK. This line of research has played an important role in showing how translingual practices can help the non-anglophone speakers, often from minority backgrounds, by allowing them to make use of their L1s, local varieties, or other linguistic or semiotic resources (cf. Canagarajah 2013). At the same time, some critics argue that, while liberating and revealing in many respects, this line of research is still routed in the "old" concept of languages as codes and makes use of the same methodological tools as previous research (e.g. Pavlenko 2018 for an overview). I would add that, paradoxically, translanguaging appears to be a largely anglocentric concept since English occupies a key position in the studies which were conducted in predominantly monolingual educational contexts. The popularity of translanguaging research is particularly pronounced in the anglophone scholarship, as evidenced by the number of published papers and presentations at conferences such as AAAL.
Seen from a non-anglophone perspective, the contribution of translanguaging to the social justice agenda is not so straightforward. For example, research has shown that translingual practices breaking the monolingual English-speaking status quo can be liberating and empowering (e.g. Canagarajah 2013). However, the findings of the research conducted in the US are not necessarily transferable to other contexts, and it cannot be taken for granted that all translingual practices encouraging the use of languages other than English will promote social justice. When used outside the anglophone world, translanguaging between English and the local language can function as a mechanism of exclusion by speakers of the local language who may also value and preserve standard English norms. A recent study conducted in an international EMI programme at a Swedish university, where about half of the students do not know the local language, found that while clearly helpful in some contexts, translingual practices may also result in excluding all those who do not share the required linguistic resources and whose English is considered to be below perceived standard (Kuteeva 2020a).
The current fascination with fluidity may have led to what Duchêne (2019) describes as contemporary "polyglottophilia", the critique of the monolingual nation-state and the nostalgia of a multilingual past, which may in turn lead to romanticizing linguistic diversity and nonessentialism and at the same time ignoring inequality and maintaining the status quo. May (2019) puts forward a similar critique towards the concept of "superdiversity" in critical sociolinguistics (e.g. Blommaert 2015), arguing that its focus on language uses in large metropolitan areas neglects the more traditional and less mobile communities. Pavlenko (2018) shows how the seemingly happy celebration of "superdiversity" in contemporary Western societies is appealing as an academic "slogan" and cites Makoni (2012: 193) who argues that it creates "an illusion of equality in a highly asymmetrical world" (cf. May 2019).
There is no doubt that, compared to other languages, the use of English today is highly asymmetrical and it occupies a "hypercentral" position (de Swaan 2001) in many linguistic constellations. Some critics of the global spread of English mentioned by Hultgren have linked it to endangering linguistic diversity and to fostering monolingualism. Gramling's (2016) book The Invention of Monolingualism traces the historical development of languages and states, showing how the scientific discovery of monolingualism in the seventeenth century was responsible for forging "technologies of governance" and "language states" (Gramling 2016: 10). This discovery of monolingualism resulted in the linguacene in which all of us currently operate and conduct research. Gramling also shows how, since the 1990s, the idea of monolingualism has been accelerated by the rise of technology-supported translation industries. It would be tempting to link this development with the global spread of English but a great deal of discussion and examples in Gramling's book focus on the establishment of standard German and various technologies that have enabled it, from Bach's Well-Tempered Clavier in the 1720s to today's EU legislation and Google Translate. This supports Hultgren's argument that material and socioeconomic factors contribute to shaping our language uses, and that we should broaden the scope of our discussion beyond any particular language.

Towards an Increased Dialogue Within Applied Linguistics
While many disciplines, including Applied Linguistics, can benefit from interdisciplinary collaboration, e.g. with Political Science, I would argue that before we turn away our attention from language to other factors, there is still room for a more holistic inquiry within our field which involves an increased dialogue between and across different subbranches of Applied Linguistics. In her position statement, Hultgren (2020) cites May (2019) to liken our field to an isolated "singular", Bernstein's (2000) term to describe an established academic discipline with a strong self-regulating community and strong boundary maintenance. A note of caution is necessary here since Bernstein's (1999Bernstein's ( , 2000 discussion of disciplines generally refers to the older, established "pure" and more theoretical disciplines (Becher and Trowler 2001), with longer histories than Applied Linguistics, which only developed in the second half of the twentieth century. Also, May's (2019) discussion of "singulars" was referring to Second Language Acquisition (SLA), which is one of the largest-if not the oldest, largest and most influential-branches of Applied Linguistics. Nevertheless, SLA does not represent Applied Linguistics as a whole.
Taken as a whole, the field of Applied Linguistics appears to be quite fragmented and looks more like an assemblage of "warring triangles" (Wignell 2007, drawing on Bernstein 1999) than a "singular" (Bernstein 2000). Bernstein (1999) accounts for disciplinary differences based on epistemological factors and knowledge structures. At one end of his disciplinary continuum lie hierarchical knowledge structures, which seek "to create very general propositions and theories, which integrate knowledge at lower levels" (Bernstein 1999: 162). These are typical of the natural sciences, in which knowledge rests upon the same foundations and is accumulated through empirical enquiry. These horizontal knowledge structures can be visualized as pyramids or triangles with broad and shared foundations. At the other end of the continuum lie horizontal knowledge structures, such as the humanities, in which knowledge is built through multiple interpretations of the same phenomena and artefacts. Wignell's (2007) discussion of the social sciences shows these disciplines often model themselves on the natural sciences but cannot agree on their theoretical foundations. While some theories may become dominant and marginalize others, a single theory does not take over the whole discipline. Such competing theories can be visualized as "warring triangles" (e.g. Kuteeva & Airey 2014). Applied Linguistics can be viewed as one such discipline. The above-mentioned debate concerning language as an object of inquiry is just one example of several conceptual and theoretical disagreements in Applied Linguistics.
To illustrate my argument with an example from Hultgren's paper, the field of English for Research Publication Purposes (ERPP) and EAP in general would gain a deeper understanding of the complexity of academic writing by engaging with the literature on bilingualism and multilingualism. For example, some research focusing on the "non-native speaker" writing for publication has looked at the threshold of nonstandard forms in published articles (Rozycki & Johnson 2013) and in manuscripts submitted for publication (Flowerdew and Wang 2016). It is certainly useful for ERPP practitioners to know how many non-standard forms and of what kind are acceptable when writing for publication and to find ways of addressing possible shortcomings among their students. As mentioned above, the English language is undergoing rapid change, not least due to the sheer number of active LX users, whose less conventional Englishes are more acceptable today (e.g. Rozycki & Johnson 2013 on engineering journals). Whether these LX users of English may feel disadvantaged by the dominance of English and experience additional challenges in their writing seems to be another question, as illustrated by a heated debate which followed the publication of Hyland's (2016) article on the "myth of linguistic injustice" in the Journal of Second Language Writing. Both linguistic and socioeconomic factors have been brought to the fore in this debate, whose main conclusion was that we need further research on this topic.
As Hultgren demonstrates with the support of O'Neil's (2018) study, six non-anglophone European countries, including Sweden, occupy top ten positions in the production of (largely English-medium) research output. Some LX users of English, such as Nordic scientists in Hultgren's (2018) study may perceive English as "just a tool for communication" and may not feel personally disadvantaged by its dominance in research communication (Bolton & Kuteeva, 2012). At the same time, this does not deny the fact that LX users, even those with advanced proficiency, often have complex and emotionally charged attitudes towards their own and others' language uses. They often have different perceptions of their L1 and LX selves (e.g. Pavlenko 2006), which in turn may translate into anxiety in professional contexts, for example, when writing for publication or presenting at international conferences. The debate about "linguistic injustice" as being caused by native-English-speaking publication gatekeepers is further complicated by the finding that LX users of English often attach high value to linguistic correctness when acting as peer reviewers or editors (Hynninen & Kuteeva 2017). Other research conducted in non-anglophone contexts, e.g. on English-medium instruction, has also shown that LX users of English can be harsh judges of their peers' language use (e.g. Kuteeva 2020a). Thus, while there is more to the problem of social injustice than language alone and other socioeconomic factors need to be considered, the interplay between language perceptions and practices of both L1 and LX users also calls for further research.

Identifying and Overcoming Fallacies
Writing about the complexities of multilingualism in connection with social justice, Ortega (2019: 35) identifies a number of fallacies that have muddled SLA research since its early days. Among the main suspects, she lists five: "the monolingual bias, nativespeakerism, essentialist ontologies of languages, raciolinguistics, and social selectivity". We can say that the same shortcomings have affected our understanding of the spread of English. The first and second fallacy rest on the assumption that monolingualism is the default for human communication and that nativeness is a superior form of language competence. The assumption that the ideal user of English is a monolingual educated native speaker, preferably from the US or the UK, is at the heart of many criticisms directed towards its global spread. This assumption also underpins a great deal of English language teaching and associated research worldwide, although these practices have been challenged by research into English as a lingua franca (e.g. Jenkins 2017, Mauranen 2012. Thirdly, the assumption that the target English to be learnt is "pure" and free from any traces of its local, non-anglophone context is supported by the essentialist ontology of language, as discussed above. Likewise, "linguistic imperialism" (Phillipson 1992) largely equates language with nation-states and the US in particular. Fourthly, we need to pay close attention not only to the ways in which language is used but by whom it is used. Raciolinguistics (e.g. Flores and Rosa 2015) is a relatively recent concept which arose in connection with the study of multilingual speakers in the US. These speakers may be marginalized and minoritized not only for their ways of speaking but also for who they are in terms of race and socioeconomic status. The approach adopted by raciolinguistics can help us gain an insight into how multilingualism can be viewed as an asset to be praised or a problem to be remedied or how it can be made invisible altogether (e.g. García and Tupas 2019). Research into the use of English outside the anglophone world has a lot to gain from adopting the raciolinguistics lens, as standard language ideologies and language hierarchies usually affect minority groups regardless of their geographical location. Finally, social selectivity refers to the overrepresentation of educated study participants based in the well-resourced countries where most applied linguistic research is conducted. Ortega (2019) adopts the acronym WEIRD coined by Henrich, Heine, and Norenzayan (2010) to refer to study participants based in Western, Educated, Industrialized, Rich, and Democratic societies. With few exceptions, research into the uses of English suffers from the same handicap, which adds insult to injury given the fact that English is already the most researched language.
In other words, it looks like there is still a lot to be done with language in Applied Linguistics. Below, I demonstrate the relevance of the five fallacies identified by Ortega (2019) in connection with the three assumptions analysed and critiqued by Hultgren in her opening statement.
Assumption 1: Non-native speakers are disadvantaged by the spread of English Non-native speakers are often disadvantaged by nativespeakerism and standard language ideologies, more than by the use of any particular language. I have discussed this commonplace attitude in connection to the use of English and Swedish in higher education, showing that the ability to use an academic language was often equated with "nativeness" by many university students and staff (Kuteeva 2014). A great deal of research in SLA and in other sub-fields of applied linguistics has rested on the premise that LX language uses are somehow "deficient" in comparison to "native" uses. This paradigm has dominated many subfields of Applied Linguistics (e.g. EMI, EAP/ERPP, LPP), although it has been recently challenged by research analyzing multilinguals' creative uses of linguistic and other semiotic resources (e.g. Canagarajah 2018). Nevertheless, the dominant discourses surrounding language use still rest on the premise that native speakers are better language users than non-native speakers. Moreover, nativespeakerism and standard language ideologies contribute to creating hierarchies of LX users of English, where the L1 speakers of language A are perceived as being superior to the L1 speakers of language B (e.g. L1 speakers of Nordic languages are better at English than L1 speakers of Romance languages). This perception of speakers, in turn, creates hierarchies of languages that may stigmatize some groups of LX users by others and actually deprive the stigmatized groups from the opportunities offered by the use of English. One example is a belief that English-medium instruction at Swedish universities is not suitable for students who do not have Swedish as their L1. As one university lecturer put it, Swedish students of immigrant and minority backgrounds may be particularly disadvantaged if English is adopted as medium of instruction, as they may already have put great effort into learning Swedish. (cited in Kuteeva, 2020b) In a university-wide survey (Bolton & Kuteeva 2012), such comments were not made by students with immigrant or minority backgrounds. Rather, these seem to be assumptions by L1 speakers of Swedish and English which reveal their language ideologies concerning the "'nonmainstream speaker'-the transnational migrant, the indigenous minority, or the socioeconomically disadvantaged" (Stroud 2010: 195). The monolingual bias, nativespeakerism, and standard language ideologies also seem to be at work here. Further research adopting a raciolinguistics lens can shed more light on the multilingualism of these students and the role that English plays in promoting or impeding social justice.

Assumption 2: English threatens other languages
Language uses among individuals are evolving and shifting, adapting to the individual trajectories of the speakers who use them. They are learnt, used, forgotten, recovered, and forgotten again. In transnational flows, accents and language combinations change, depending on the contexts of their usage. As someone who speaks an X number of named languages, I can certainly observe these changes in myself, my children, friends, and family members. But even if we detach languages from their users and perceive them as abstract entities, it is hard to imagine how any language per se would threaten another. Rather, it is the regulatory mechanisms behind the language, such as the educational establishment, state interventions, market economy, family language policies or technological affordances, that tend to threaten non-dominant languages and varieties. This is true for English or any other language (e.g. Gramling 2016 on German, Bourdieu 1991 on French), although the mechanisms behind English operate on a larger scale today compared to other dominant languages.
Standard language ideologies and language hierarchies can contribute to the process of silencing multilingual individuals. This is often the case in high-stakes domains such as business, education, and science which tend to regulate their language uses to a greater extent. For example, Holmes (2020) shows how the dichotomy between Swedish and English in connection to the internationalization of higher education leads to essentializing languages and their speakers and reproducing language hierarchies at a science department of a Swedish university.
The over-simplified and essentialized perceptions of English as "international" and Swedish as "national" leads to rendering the multilingual practices and identities of international scientists invisible and ignores the complexity of the interplay between different linguistic resources which these scientists employ in their everyday work and in the production of knowledge, even when the final output is in standard English or Swedish. In this case, far from being simply "a tool for communication", both English and Swedish appear as ideologized monolingual and monocultural categories which are being reproduced in a hegemonic fashion. Thus, contrary to the argument that English alone poses a threat to linguistic diversity, Holmes' study provides some evidence to the fact that regulatory mechanisms surrounding both English and Swedish (e.g. official and informal language policies) undermine the diversity of language uses which do not enjoy the same status and institutional support. Further research on how the essentialist ontologies of language impact language practices is needed.

Assumption 3: Language policy will curb the spread of English
Hultgren provides a number of examples, mainly from the Nordic countries, where "parallel language" policies promoting the use of national languages are often at odds with other policies, e.g. university internationalization or bibliometric evaluation regimes, which indirectly promote an increasing use of English. The latter are certainly more powerful in determining which language(s) are used because they form a basis for rankings, funding, and reward systems. Language policies, such as those in Nordic higher education, tend to reflect the beliefs and dominant discourses of their authors, in many cases university-based applied or sociolinguists on the quest for social justice and protection of language(s) which form part of their country's or region's national heritage. These are important documents that have the potential to shape language practices but their implementation is not always clearly understood or followed up, as shown, for example, by a recent study of supplementary language summaries in doctoral theses (Salö 2018).
Since English is so widely used in the Nordic countries, it has also been made responsible for a lack of interest in studying other modern languages which have been traditionally taught at schools, e.g. French, German, or Spanish (e.g. Josephson 2015). It is certainly often the case that young people learn a great deal of English outside educational settings (e.g. Sundqvist & Sylvén 2016). Their first experience of learning a new foreign language is very different from English and poses so-far unknown challenges. The same applies to higher education where English is widely used. In order to foster this "academic" multilingualism, the latest Nordic university language policy document titled More Parallel, Please! (Gregersen et al. 2018) lays emphasis on the development of language support and resources for major European academic languages other than English. It also encourages universities to support grassroots multilingualism by valuing the resources of the "international classroom" (Gregersen et al. 2018: 21). The document states that in today's society bilingualism is not enough (meaning the national language and English), and doctors and many other high-stakes professionals need to know more than two languages. This call comes in the aftermath of recent research underscoring the potential of multilingual resources in educational settings and re-assessing what counts as acceptable language norms.
For now, it seems that promoting academic multilingualism and supporting grassroot multilingualism may have more impact than protectionist policies juxtaposing English and the national language alone could do but it remains to be seen how this new policy would work out in practice. As discussed above, if language hierarchies are not questioned, there is still a danger of romanticizing multilingualism of the educated, elite kind (cf. Kuteeva 2020b) and ignoring the grassroots multilingualism of the less privileged population who may be working hard to convert themselves into "perfect monolinguals" in order to be socially accepted. Further studies on the role of English in social integration, e.g. of recent migrants, will address a glowing gap in the literature on language policy and the global spread of English.
To conclude, I agree with Hultgren that widening our analytical lens is important but I also believe that increasing dialogue between different branches of Applied Linguistics is just as urgent as interdisciplinary research. In particular, the suggested focus on the socio-material aspects is needed for drawing our attention to less known contexts of language use and to under-represented study participants. Before we tone down language in our scholarly inquiry, we still need to hear more voices.