Whose language counts?

1 Introduction

Sociolinguists critically explore the ways in which language reflects, reinforces, or contests social inequalities. They investigate not only how people use language in interaction, but also how they think they (and others) are using language—what we call language ideologies. One problem, however, is that too often, only so-called ‘native speakers’ who learned a language in their childhood or ‘monolinguals’ are being asked what they think about ‘their’ language. This means that when linguists ask people what they think of e. g. gender-inclusive language in, say, German, they neither ask people who learned German as adults only, nor German users who do not live in a German-speaking country. As most studies focus on native speakers, monolinguals and/or non-mobile people, which are an increasingly small minority in contemporary times characterized by “increased connectedness” (Bock, Busch & Truan 2023: 193), what are we actually describing?

In this position paper, I intend to show that our discourses on language are affected by the multilingual experience. While it has been recognized that the concepts of ‘mother tongue’ or ‘native speaker’ are problematic in times of superdiversity (Vertovec 2007; Blommaert & Backus 2013), I argue that linguists still (implicitly) rely on native speakerism and a monolingual bias. But, then, as the title of this paper says: Whose language counts? In other words, who is considered the most representative for a language or a community?

Let us take an example. The (overly simplified) stance “I find gender-inclusive language important”, either overtly expressed, or reconstructed through language practices, serves as an instance of language ideology. Throughout this position paper, I will use the German-speaking world as a case study, both as it is my field of specialization and an interesting case of a pluricentric language—a language dominant in many centers such as Berlin and Cologne (Germany), Graz and Vienna (Austria), Basel and Zurich (Switzerland), but also Windhoek (Namibia), or Bulverde (Texas, USA). So, from what kind of people do we know whether they align or not with the stance “I find gender-inclusive language important”?

The population who is traditionally being surveyed consists of Germans who use German as a first language and live in a German-speaking country. Conversely, if an individual, despite being proficient in German and residing in Berlin, is French and learnt German later in life, their insights are systematically excluded from such linguistic examinations. And if you are German, speak German, but happen to live in The Hague, New York, or Buenos Aires, then suddenly you also disappear from studies on language ideologies. Recentering these overlooked speakers is the focus of this paper.

The paper is structured as follows. I first briefly define language ideologies, then show more precisely what previous research at the intersection between language ideologies and multilingualism has looked into, and why this still contributes to othering multilinguals (Wiese et al. 2022). Section 2 is a presentation of two well-identified language ideologies: native speakerism and monolingual bias, which, as I will argue, are prime examples of erasure in language ideological research (Irvine & Gal 2000). In Section 3, I show how research on gender-inclusive language is still imbued with linguists’ ideologies of native speakerism and monolingual bias, also entangled in ideological constructions of ‘one language’ belonging to ‘one state’. Drawing on empirical research on L2 speakers of German, Section 4 shows why integrating a higher variety of language users is necessary—not only on ethical grounds, but also on theoretical ones. Ultimately, in sketching hopes for a more inclusive sociolinguistic future, I plead for considering multilingualism a relevant factor in analyzing language ideologies—nothing else, nothing more.

1 What we know until now: Language ideologies and multilingualism

Broadly speaking, language ideologies are language users’ conceptualizations about linguistic practices (Silverstein 1979; Schieffelin, Woolard & Kroskrity 1998; Blommaert 1999)^[1], and how we, in turn, value these language users. For instance, what do we think about gender-inclusive language and those who use it or do not? Although these questions may seem neutral, they are in fact closely connected to issues of power and social justice (Cavanaugh 2020: 55). Just as language varies, language ideologies also are exposed to variation (Kroskrity 2004).

The term ideology encapsulates a “totalizing vision” (Gal & Irvine 2019: 49) that foregrounds power relations within linguistic contexts, while also “dislodging the notion of ‘language’” (Blommaert 2006: 512). When we focus on ideologies indeed, we become able to identify language as a “complex of metapragmatic qualifications projected onto situated language usage” (Blommaert 2006: 512), thus acknowledging that not only language practices, but also judgements about language, shape language variation and change. By adopting the term language ideologies, there is thus an intentional emphasis on the comprehensive examination of how societal and political forces shape and are reflected in linguistic expressions (Cavanaugh 2020).

Since the multilingual turn (Conteh & Meier 2014), studies on multilingualism and multilinguals have grown tremendously. How has it translated in language ideological research? Until now, language ideological research has dealt with three major trends:

1. language ideologies in monolingual settings, especially how speakers position themselves towards nonstandard varieties (early works cited above);

2. language ideologies and language acquisition (Hall 2018), with a growing body of literature on the role of teachers in language ideologies (Ruck 2020; Takeuchi 2021; Milojičić 2022; Cushing 2023). Interestingly enough, the terminology is often different, as many publications investigate ‘teachers’ beliefs’ (Razfar 2012; Haukås 2016; Dobbs & Leider 2021) rather than ‘ideologies’;

3. language ideologies in multilingual settings (Busch 2015), i. e. what it means for multilinguals to experience multilingualism in their everyday lives, at work, etc., with many studies targeted at students, probably due to easy access for researchers (Vogl 2018; Bodis 2021; Cushing & Helks 2021).

In parallel, a body of quantitative studies integrating multilinguals has established that multilinguals have a different perception of appropriateness and standard in their different languages. For instance, the perceived emotional force of swear and taboo words is highest in the L1 and gradually lower in languages learned subsequently (Dewaele 2004). Surprisingly, higher degrees of multilingualism are not automatically linked to more positive attitudes towards code-switching (Dewaele & Wei 2014). Finally, participants knowing more languages scored significantly higher than those knowing fewer languages on tolerance of ambiguity (Dewaele & Wei 2013).

Altogether, the language ideologies of multilinguals and language ideologies around multilingualism have already been widely explored. However, although language ideology research is increasingly concerned with multilinguals and multilingualism, it does not combine original questions about the positioning of particular language phenomena with a multilingual perspective. In other words: While language ideology research deals with language ideologies about multilingualism, it does not ask what multilinguals understand by gender-inclusive language, anglicisms, non-standard varieties, accents, etc. and how they position themselves in their different languages.

Quite paradoxically, research on language ideologies still focuses on people who are considered ‘native speakers’ (or may not be asked which languages they use), people who (supposedly) use a single language (‘monolinguals’), and people who live somewhere where the language they use is dominant and/or (officially) recognized. Specifically, people who have acquired new language(s) in adulthood and people who do not live somewhere where their first language (L1) is the dominant language are often not considered when it comes to current language ideological debates.

This means that while multilinguals are included in language ideological research, it is mostly as multilinguals—and not on the same grounds as other language users. Focusing research on multilinguals on their multilingual identity tends to give them a status of exception. This othering of multilinguals occurs surprisingly often in linguistic research, and often implicitly relies on the assumptions according to which multilinguals are migrants and/or have a foreign origin (Wiese et al. 2022). In the next section, I will review the two interrelated biases—native speakerism and monolingual bias—at play in this process of othering.

2 Ideologies of native speakerism and monolingual bias

In some ways, multilinguals are ‘special’ on two grounds: First, they invite us to rethink the notion of ‘native speaker’ as the sole authority on a language defined as a bound entity; second, they show by their very mere existence that the world is not as monolingual as many (nonlinguists) may think. While both ideologies are closely intertwined—as the ‘perfect’ native speaker is the one who speaks ‘their’ language without ‘contamination’ from other languages—I will present both ideologies separately for the sake of exposition.

3 Invisibilization or erasure in language ideological research

Language ideological research is concerned with how our representations or assumptions about language varieties and their users may lead to processes of “differentiation” (Irvine & Gal 2000). Importantly, linguists are also imbued with these “ideological representations of linguistic differences” (Irvine & Gal 2000: 37). The bias towards native speakers, monolinguals, and nonmobile people in language ideological research thus also proceeds to invisibilization or erasure:

Erasure is the process in which ideology, in simplifying the sociolinguistic field, renders some persons or activities (or sociolinguistic phenomena) invisible. Facts that are inconsistent with the ideological scheme either go unnoticed or get explained away. So, for example, a social group or a language may be imagined as homogeneous, its internal variation disregarded. Because a linguistic ideology is a totalizing vision, elements that do not fit its interpretive structure—that cannot be seen to fit—must be either ignored or transformed. Erasure in ideological representation does not, however, necessarily mean actual eradication of the awkward element, whose very existence may be unobserved or unattended to. (Irvine & Gal 2000: 38)

The idea is that by focusing on the seemingly default norm of ‘native speakers’ who would be L1 monolinguals living in an environment where their (only) language is dominant, sociolinguists implicitly contribute to the homogenization described by Irvine & Gal (2000). Importantly, Irvine & Gal (2000) explain that erasure may proceed to the fact that some “awkward” elements are “unobserved or unattended to” rather than purposely and actively disregarded. When they focus on, say, the language ideologies of Germans using German and living in Germany, sociolinguists do not necessarily imply that you would have to be German to use German, or that living in Germany is a requirement for holding language ideologies on German. Still, the persistent collective emphasis on e. g. German as a dominant/majority language (or, when German as a minority language, the focus on German as officially recognized) disproportionately advantages nonmobile L1 users at the expense of the diaspora, heritage speakers, and L2 users. It is in this sense that the implicit focus on languages as national entities contributes to the invisibilization or erasure of other users of German, such as those who learnt German as a minority language, or in adulthood.

The implicit and misleading assumption that ‘Germans using German’ are the norm relies on “some sociological contrast [that] seems to require display”:

By focusing on linguistic differences, we intend to draw attention to some semiotic properties of those processes of identity formation that depend on defining the self as against some imagined ‘Other.’ [...] Such representations may serve to interpret linguistic differences that have arisen through drift or long-term separation. But they may also serve to influence or even generate linguistic differences in those cases where some sociological contrast (in presumed essential attributes of persons or activities) seems to require display. (Irvine & Gal 2000: 39)

The advantage is that ‘native speakers’ and ‘monolinguals’ get to be asked first when it comes to language ideologies about ‘their’ language, which further emphasizes “linguistic differences” that may not be empirically sound and rather be the consequence of “sociological contrast”. To understand whether these differences actually exist, and whether multilingualism plays a role, we need to integrate a wide variety of language users when considering the language ideologies around (what is perceived and valued as) ‘one’ language.

4 Research on gender-inclusive language as an example of linguists’ ideologies

Despite the focus on “transnational identities” (De Fina 2016) and the development of migration linguistics (Stehl 2011; Canagarajah 2017), the native speaker and monolingual bias is visible in many studies’ design. One way linguists still contribute to native speakerism and monolingual bias is by considering that ‘native speakers’ and ‘monolinguals’ may (should?) be the first ones to be asked about current language ideological debates, based on the assumption that their language proficiency and emotional attachment to a language make them the best informants on the ideological debates revolving around that language.

Let us take the example of gender-inclusive language, a highly topical and controversial subject of interest in many communities. A sneak peek at the program of the international conference “Attitudes towards gender-inclusive language. A multinational perspective”^[3] organized by Falco Pfalzgraf in September 2022 gives us insights into the default angle. While nine out of thirty-three contributions are interested in possible transfers or comparisons between languages, even these contributions are not protected from problematic assumptions. One for instance directly relies on the distinction between “participants who were both native and non-native speakers”, while another focuses on “translating gender-neutral language from English into their native language”^[4]. Many other contributions define nationality as an excluding criterion, as this one which relies on an “online survey with [language X]-speaking [nationality X’] as participants”. Here we see that the association between a language and the citizens of a European nation-state in which the language is dominant and officially recognized is taken as default, and that speakers of a language X are expected to also have the nationality (in the singular) ‘matching’ that language.

Linguists also tend to focus on the debate in a specific country where the language is either the majority language or officially recognized. Corpus studies, for instance, usually tackle media discourse in German in German-speaking countries such as Germany (Waldendorf 2023) or Austria (Drüeke, Pascher & Peil 2018), thus letting aside the fact that media moves beyond national borders. Even if we assumed that newspapers articles would reflect a national position, which may already be a bold claim, the readership of German media does not consist only of Germans or L1 German users. Thus we can expect articles from German newspapers to be read beyond Germany, or online comments of Austrian newspapers to be written by non-Austrians chiming in while being influenced by the German-speaking discourse on e. g. social media. The debate on gender-inclusive language is currently often polemical and present in the media—both in Germany and in neighbouring countries. Parallel discussions in other language communities may thus influence the (possibly nation-based) public and political discourse. It can be assumed, for example, that participants in the debate on gender-inclusive language in Germany are inspired by the ideas and approaches of other language communities, or that cultural comparison leads to increased language sensitivity.

Comparisons between language varieties of pluricentric languages beyond English are also rare, thus limiting our understanding of how language contact may impact language ideologies. We do not know yet, for instance, whether French users in Switzerland or Québec may exhibit different ideologies around gender-inclusive language depending on the influence of the German-speaking and Italian-speaking discourses in Switzerland vs. English-speaking discourse in Canada (see Jack-Monroe 2023 for a first MA thesis on the topic). As Rivera Alfaro & Cuba (2022) put it, these study designs reveal “ideological restrictions underlying nation-state construction, which were naturalized in several sociolinguistic studies” (also see Schneider 2019 for a critique of methodological nationalism).

As a consequence, we learn more every day on the language practices and ideologies of, say, German users in Germany, Switzerland, and Austria, while overlooking L1 German users who do not live (and may have never lived) in a German-speaking country, have migrated to another country, and/or people for whom German is their LX—even if they live in a dominantly German-speaking country. This is particularly visible in psycholinguistic studies, in which having “German as their native language” and “residence in Germany” (Zacharski & Ferstl 2023: 300) are often named as excluding criteria. For language ideological research however, it remains to be shown why being a highly competent user of German learnt after the age of 3 (having German as an L2, L3, etc.) and/or having German as an L1 but having moved to another (possibly German-speaking) country would make you an unsuitable candidate to assess gender-inclusive practices in a language you may be using daily.

As language ideologies are still imbued with a native speaker, nonmobile, and monolingual lens, a significant group of speakers who may exert an influence on the dynamics of language ideologies is not adequately represented. In a nutshell, we currently know more and more about what L1 speakers of a given language think about e. g. gender-inclusive language, but not what multilinguals, L2 users, and mobile people think about it. This means that people who learned a new language in adulthood and those who do not live anymore in a country in which their L1 is the majority language are not interviewed nor surveyed, no matter how proficient they are in the language investigated. Speakers of, say, German, for whom German is not their first language or who speak other languages next to German, are currently de facto excluded from language ideological research.

5 Why integrating a wide variety of language users of a language is worth it

As I have argued until now, investigating multilinguals is necessary on ethical grounds: Recognizing their legitimacy comes to social justice in an increasingly mobile, globalized world (Blommaert 2010) in which ‘native speakers’ and ‘monolinguals’ represent only part of the population (although a privileged one). Going further, it is also on theoretical grounds that we need to know (a) whether the language ideologies of multilinguals differ from those of so-called (monolingual) ‘native speakers’; (b) why this is so; (c) how this may impact language (and thus social) change. Importantly, question (a) relies on the assumption that comparing the language ideologies of multilinguals to the language ideologies of ‘monolinguals’ may be “comparing apples with pears” (Dewaele 2018: 239). In all likelihood, even ‘monolingual native speakers’ have been exposed to other varieties or languages, and may be considered multilinguals. An option is then to see multilingualism as a “continuum” (Leivada et al. 2023: 2) where degree of exposure, proficiency, and emotional attachment to the languages all are involved in how language ideologies emerge and are further reproduced.

Research has shown that language ideologies, as part of the “total linguistic fact” (Silverstein 1985: 220, 257), play a central role in shaping language change—which Silverstein has precisely shown with the example of gendered forms, “a particular formal indexical distinction” (Silverstein 1985: 251) that gets assigned value based on usage linked to a specific socio-economic status, prestige, etc. Notably, a favorable disposition toward gender-inclusive language contributes to its increased usage (Burnett & Bonami 2019), and when adopted collectively, induces gradual yet perceptible changes in language dynamics.

Based on this, we may hypothesize that multilinguals are also the drivers of language ideologies. Specifically, I am asking the question: Can we speak of a transfer of language ideologies between the many languages we use? Since we know from research on language contact that all domains from linguistics are affected by contact-induced change (Grant 2020), could it also be true for language ideologies? Can we then speak of a transfer of language ideologies between the first language(s) and the second language(s), and if so, in which direction? If multilinguals bring their ideologies into the L1 and LX, this transfer may lead to language change, so that usage finds its way through multilingual speakers. These questions can be articulated as such:

1. Are language ideologies expressed similarly in the L1 and LX?

2. What role do language ideologies in the L1 play on the LX and vice versa?

3. Can we speak of a transfer of language ideologies (and in which direction?)?

It has been suggested that multilinguals may “apply different notions of the standard language ideology to their different languages” (McLelland 2020: 119). As defined by Lippi-Green (1994: 166), standard language ideology is “a bias toward an abstract, idealized homogeneous language, which is imposed and maintained by dominant institutions”. Applied to gender-inclusive language: Is the positioning of L2 speakers of German towards gender-inclusive language different than the one they have in their L1? Conversely, are the language ideologies of L1 speakers of German who have various LX different than those of L1 German speakers who experience their everyday life through a (more) monolingual lens?

In a paper based on interviews with Francophones who moved to Berlin and learnt German in adulthood (Truan accepted), I attempted to answer these three questions. First, are language ideologies expressed similarly in the L1 and LX? The data analysis showed that speakers of French and German have a different understanding of what counts as standard when it comes to gender-inclusive language. While almost all respondents are equally positive about gender-inclusive language in all languages, they only consider it to be unmarked their L2, German, while in the L1, French, being in favor of it is a political stance.

Gender-inclusive language itself is not always clearly recognizable for many interviewees in the L2, which can possibly be explained by its frequency and thus inconspicuousness. In Berlin and especially in academic, cultural, and left-wing political milieus, gender-inclusive language is considered the (implicit) norm. Against this background, it is assumed that variation is less visible in the L2: Local usage norms become the norm. Beyond multilingualism, the influence of social class in language ideologies about gender-inclusive language is thus clear.

Second, what role do language ideologies in the L1 play on the LX and vice versa? In the L1, French, language ideologies play a more important role and are more deeply rooted. The easier acceptance of gender-inclusive language in the L2 can be explained by several factors. Firstly, controversies regarding gender-inclusive language are perceived less intensely in the L2, possibly due to a lower emotional charge. Critical debates about gender-inclusive language, especially those that could be considered conservative or purist, receive little attention in the L2. In this respect, language ideologies in the L2 are less influenced by public-media discourses. Secondly, in the L2, language-immanent argumentation (“gender-inclusive language is easier in German”) is more frequent.

Third, can we speak of a transfer of language ideologies (and in which direction?)? Compared to the L1, multilinguals are more open to language change and language variation in their L2. The interviews indicate that there is a transfer of language ideologies from the L2 to the L1: Progressive language ideologies in the L2 have an influence on language ideologies in the L1. After multilinguals have experienced what gender-inclusive language can look like in their less emotional language, similar language practices are adopted in the L1. The effects of multilingualism thus prove to be decisive for active participation in inclusive language practices. Going further, we may hypothesize that L2 users also contribute to language change as a slow, gradual process informed by internal, social, and cognitive and cultural factors (Labov 2006 a; Labov 2006 b; Labov 2010).

6 Conclusion and hopes for a more inclusive sociolinguistic future

The question I tackled in this position paper was: Whose language counts? In other words, who is considered the most representative for a language or a community? An apparent disconnect pervades our field: While language ideological research precisely examines representations and assumptions about language and its users, we as linguists still implicitly perpetuate biases. In language ideological research, the default or prototypical person asked what they think about particular language phenomena remains the ‘native speaker’ or ‘monolingual’ language user. Specifically, the first persons deemed relevant for the study of, say, ideologies around gender-inclusive language, are the ones living in a context in which their L1 is dominant, officially recognized, and unmarked. While it has long been acknowledged that a language is not only used by those for whom it would be their ‘native language’ or ‘only language’, it looks like these critical insights have not entirely made it into language ideological research.

Relevant factors for assessing language ideologies nevertheless include multilingualism. Importantly, the central question ‘what counts as standard’ can be rephrased as ‘where do multilinguals perceive a standard?’. When tracing back what is perceived as the norm in a given community, multilingualism is a factor deserving attention next to age, gender, occupation, social class, etc. The impact of the learning context and socialization in the LX language may be more prevalent than compliance to a standardized norm (as in textbooks).

I have shown that language ideology research needs to include multilinguals much more, and in particular L2 users and mobile L1 users. Language ideologies circulate in a multilingual, globalized society in which not only L1 speakers shape language use and change. How, then, can we contribute to a more inclusive sociolinguistic future? The first step is to stop considering ‘native speakers’ and ‘monolinguals’ as the default or the norm. While some may argue that linguists’ efforts may have found their way in education and applied linguistics already (Flores & Rosa 2015; Ramjattan 2019; Oldani & Truan 2022; Cushing 2023), a very strong resistance against change, even in these fields, remains perceptible (Wiese et al. 2022), and we crucially need to “decolonize linguistics” (Criser & Malakaj 2020; Canagarajah 2022; Deumert & Makoni 2023; Hudley et al. 2024). The first move would be, when asking language ideological questions, not to start with L1 users living in a country where their L1 is the majority language, nor to restrict investigations to these people. We can indeed safely assume that people for whom e. g. German is their L2 but have been living for many years in a German-speaking country are shaping language ideological debates on German. Of course, factors such as proficiency level, reception of public discourse, participation in activities in the given language, emotional attachment to their many languages, etc. may all play a role in how L2 users express language ideologies in their L2. Similarly, there is no reason to consider a priori that L1 users who are living (temporarily or not) in a place where their L1 is not dominant have stopped taking part in language ideological debates in their L1. Here again, the (perceived) closeness or distance to the L1, reasons for and duration of the migration experience, familiarity with the local language(s), etc. may also shape the language ideologies of mobile people.

My hope is then that next to (still important!) investigations on their multilingual identities, we also ask multilinguals the same questions we would ask ‘native speakers’, ‘monolinguals’ (if they exist), and (at the moment of the investigation) nonmobile people. As I have sketched above, it is important that we keep accounting for all the factors (gender, race, age, class, etc.) that may play a role in language ideologies without suddenly giving multilingualism a status of exception. Language ideologies circulate among L1, LX, mobile, multilingual users. It is high time we recognize what the multilingual experience does to language ideologies in the many languages we inhabit.