'We don't worry that much about language': street-level bureaucracy in the context of linguistic diversity.

The way we deal with diversity is crucial for social equity in the context of migration-related super-diversityand represents a challenge for all actors involved. The present article aims to contribute to the understanding of linguistic discrimination by contrasting the perceptions of institutional actors and mobile European citizens concerning language-related barriers when accessing labour market mediation services and benefits. The article draws exemplarily on empirical data (mainly qualitative interviews) relating to the provision of labour market-related services by the Austrian Employment Service and Hungarian migrants' experiences with this institution. The juxtaposition of these two complementary perspectives reveals the challenges that managing linguistic diversity poses for institutions and the actors involved alike. Likewise, it permits investigations into how emerging language-related problems may translate into experiences of inequity.

Introduction

In a minor media scandal, the semi-public Austrian Employment Service (AMS) was recently accused of ‘not being able to cope with migration-related challenges’ (Die Presse 2018). While this incident of an Austrian institution allegedly overwhelmed by migration-related challenges could be regarded as a purely national matter, it is symptomatic for the discontent that diversity may cause for street-level bureaucracies in many Western European countries. These organisations operate in a context characterised by the inconsistency between minor political attention given to diversity management and the social reality of ethnically, culturally and linguistically diverse societies. Only occasionally, as in this particular instance, does the institutional management of diversity receive public and media attention. The way diversity is addressed is an inherently political issue and critical for social equity in the context of what Vertovec (2007) terms ‘super-diversity’. The steep rise of migration in recent decades as well as the development of supra-national institutions such as the European Union (EU) together result in linguistically, ethnically and culturally more diverse settings. Specific attention must be given to the political, economic, cultural and social processes through which difference is transformed into inequality (Brubaker 2014, 3). Managing diversity represents a challenge for all actors involved and is of particular importance regarding street-level bureaucracies that are characterised by their frontline position, mediating between government policy and the public (Lipsky 1980). Due to the ‘pervasiveness of language’, and ‘the increasingly and inescapably “languaged” nature’ of modern social life (Brubaker 2014, 3), linguistic diversity has become particularly relevant in this respect (Arnaut et al. 2015; Vertovec and Wessendorf 2006). This was also apparent in the aforementioned media coverage in which ‘serious problems with unemployed people with a mother tongue other than German’ (Jankowski 2018) were positioned at the centre of the debate. However, linguistic competencies were thereby often used as a pars pro toto for assumed cultural and religious ‘otherness’ and associated with a refusal of integration and even propensity to violence (Jankowski 2018; Vettermann 2018). While this predicament is common in public debates, the focus in the following will be on language-related barriers experienced by street-level bureaucrats and migrants. At the same time, as illustrated by the telling statement of a diversity appointee at the AMS quoted in the title of this article (‘We don’t worry that much about language’), linguistic discrimination is often neglected, talked down and, consequently, insufficiently addressed by institutional strategies in the handling of migration-related diversity. By contrasting the perspectives of institutional actors and their clients, the present article aims to show how intra-institutional regulations of language practices affect all actors involved and may lead to linguistic discrimination. It illustrates this by drawing on empirical data linked to the AMS’ provision of labour-market related services within the context of linguistic diversity and Hungarian migrants’ experiences with this institution.

The case study of Austria is illustrative of many Western European countries in several regards, chiefly due to the discrepancy between political denial and the social reality of an ethnically and culturally diverse country with a broad spectrum of immigrants and officially recognised minorities (Krzyżanowski and Wodak 2008). Furthermore, Austria constitutes a noteworthy example for its historical experience in coping politically with the different aspects of multilingual discourse: The Habsburg Empire represented a model of linguistic pluralism that stood out in nineteenth-century Europe marked by the emergence of linguistically unified nation states (Rindler-Schjerve and Vetter 2007). As labour market access is decisive for participation in society, the semi-public AMS constitutes a central institution regarding social equity in the context of migration-related diversity. Mobile Hungarians are an interesting migrant group in this research framework: Apart from being the largest non-German native speaker group in the Austrian labour market with a longstanding migration history to Austria, they, as EU citizens, are under specific circumstances accorded the same legal status as their Austrian peers in terms of accessing unemployment benefits. Hungarians’ migration to Austria must be understood within the significant socio-economic asymmetries between a ‘new’, Eastern EU member state marked by widespread insecurity in their domestic labour market (Szalai 2013; Szikra 2014) and their well-off Western neighbour, an ‘old’ EU member state with a comparatively low unemployment rate. However, Hungarians in Austria are less associated with cultural or religious ‘otherness’ than other (mostly Southern European or Muslim) migrant groups: Notwithstanding decisive socio-economic differences and a decade-long division by the Iron Curtain, the neighbouring countries Austria and Hungary share a century-old common history (beyond the well-known Dual Monarchy) as well as cultural and religious (predominantly Roman Catholic) traditions. One interview participant at the AMS explicitly denied the existence of racism against Hungarians, while another ascribed profound knowledge of the Austrian bureaucratic system to them (‘they know the ropes perfectly’). In contrast to the interviewee’s perspective, it was possible to identify instances of xenophobia and discrimination that the interviewed Hungarians experienced. While other dimensions of discrimination are present in Austrian bureaucracy, focusing on Hungarian migrants promotes concentrating on linguistic barriers that migrants and street-level bureaucrats (Lipsky 1980) may encounter. Thus, the juxtaposition of these two complementary perspectives (institutional actors and mobile European citizens) reveals the challenges that managing linguistic diversity poses to institutions and the actors involved. It also supports investigations into how emerging language-related problems may translate into experiences of inequity. Drawing theoretically on Bourdieu’s work (1982, 1986) about the social function of language and his differentiation of forms of capital, the present article emphasises the political and historical situatedness of linguistic competence (Blommaert, Collins, and Slembrouck 2005; Heller 2007). Its objective is to address the under-researched (re)production of social asymmetries through language and linguistic practices in the context of migration-related diversity. It is often difficult to separate discrimination on the basis of language from other types of discrimination, as ‘unequal access to power and resources is often multi-causal, and differences based on language often overlap with other socially constructed differences’ (Skutnabb-Kangas 2015, 2–3). Although neglected by the relevant literature and often subsumed under ethnic or cultural diversity, linguistic ‘differences are consequential for the production and reproduction of inequality’ and attention must therefore be paid to the ways in which they are ‘phenomenologically experienced, interactionally negotiated, culturally elaborated, socially organised, legally regulated and politically contested’ (Brubaker 2014, 20–21).

As Bourdieu (1982) notes, language is not merely a neutral means of communication but also has a social function, ascribing status positions and thus reproducing social inequalities on a symbolic level. The officially authorised mode of language (including accent, variety and repertoires) – the ‘legitimate language’ – is the one spoken by the powerful and constitutes the general norm (especially in official contexts), engendering symbolic domination and hierarchisation among speakers who are more or less skilled in the use of that language – and thus more or less legitimate speakers. While Bourdieu explored foremost within-language inequalities, the social function of language is of special relevance in regard to questions of multilingualism. In accordance with Blommaert, Collins, and Slembrouck (2005, 213), this article views multilingualism not as ‘what individuals have or lack, but [as] what the environment, as structured determination and interactional emergence, enables and disables them to deploy’. Communication problems are to be considered as socially constructed rather than a priori existing, as ‘the result of how individuals and their communicative “baggage” are inserted into regimes of language’ valid in a particular space (Blommaert, Collins, and Slembrouck 2005, 198), or in other words, into ‘linguistic environments’ that are the objects of language policy (Grin 2003, 28). Questions regarding domination, multilingualism and linguistic barriers are of particular importance in the context of the EU where an official ideal of cultural and linguistic diversity contrasts with the territorial organisation of linguistic difference and where manifestly multilingual societies are paradoxically perceived as essentially monolingual at the national level (see, e.g. Busch [2009], Gal [2006] or Stevenson [2006]).

Another tension of particular relevance in the context of the EU concerns the multi-level governance of social rights: While European integration has gradually restricted the authority of EU member states to delimit granting social rights to other EU citizens, the organisation of social security still relies on national institutions to design and provide social benefits (Seeleib-Kaiser and Pennings 2018). The organisation of social security, traditionally located within the sovereignty of the nation state, generally faces new challenges in an increasingly mobile world (Sabates-Wheeler, Koettl, and Avato 2011). At the normative-political level, welfare and migration scholars have debated whether and how increasing immigration may weaken support for the welfare state that builds on a presumably shared national identity and the ensuing consequences for contemporary societies (Ehata and Seeleib-Kaiser 2017; Favell 2016; Kymlicka 2015). When analysing social security rights from a sociological perspective, not only must normative as well as a political-legal registers be considered, but also practical elements that encompass the mechanisms of institutional implementation and the actual experiences of access (Carmel 2013; Marshall 1977; Soysal 2012). While the EU has indeed created a unique transnational portability system to secure the social security rights of mobile Europeans, substantial access barriers still persist in practice (Scheibelhofer, Holzinger, and Regös forthcoming; Seeleib-Kaiser and Pennings 2018). A recent study of EU migrants’ access to subsistence support in Germany has shown that the main informal barrier in this respect is the enforcement of German as the language of administration and communication (Ratzmann 2018). Likewise, in the Austrian context, language competence arguably exercises a fundamental gatekeeping function, especially in terms of increasingly ‘Germanised’ immigration policies that restrict the rights of those less fluent in the official language (de Cillia and Wodak 2006; Krzyżanowski and Wodak 2008; Mourão Permoser 2012; Wodak and Boukala 2015).

Research design and methodological-ethical implications

The empirical base for this article’s analysis encompasses qualitative interviews and participant observations conducted within the international and interdisciplinary TRANSWEL project (2015–2018) as well as in the course of an accompanying study (see Holzinger forthcoming). Focusing within this research framework on intra-European migrants’ experiences when accessing social security in four pairs of countries (Hungary–Austria, Bulgaria–Germany, Estonia–Sweden and Poland–United Kingdom), multiple obstacles and barriers that interviewees themselves linked to language and communication problems became apparent. In the Hungary–Austria country pair, this was especially relevant for labour market mediation and unemployment benefits (for methodology and main findings, see Regös, Holzinger, and Scheibelhofer forthcoming; Scheibelhofer, Holzinger, and Regös forthcoming). In total, 25 problem-centred interviews (Scheibelhofer 2008; Witzel and Reiter 2012) with Hungarian migrants and their significant others (family members and partners) were conducted by the Austria-Hungary country team (eight among them reported experiences with the AMS) as well as twelve (six per country) additional expert interviews with high-level officials, policy makers and legal specialists (for the present article, of special relevance were two interviews with senior management officials at the AMS and the national coordination office of the European Employment Services (EURES)in Austria). Additionally, seven problem-centred interviews with AMS employees working either in management positions or at the street level were conducted in a side study, which focused on institutional actors’ perceptions of linguistic diversity study (see Holzinger forthcoming). The interviews were conducted in Hungarian (with migrants) and German (with institutional actors) by members of the research team speaking the respective language. Additionally, participant observations (Spradley 2009) were carried out at regional offices while accompanying three of the interviewees to appointments at the AMS in order to provide assistance with administrative formalities. The entire research process was based on the methodological principles of constructivist grounded theory (Charmaz 2006). The spatial focus of the study was Vienna, the Austrian capital, because the federal structure of the AMS (and hence differences between the regional organisations) made a territorial delimitation meaningful. Vienna was of particular interest in this context as it has the highest share of migrants nationwide (Statistik Austria 2017) and migration-related linguistic diversity is thus especially relevant in this area.

Ethical considerations were integral to the research process, in terms of informed consent, confidentiality and the communication of findings. An emphasis was placed on the (linguistic) accessibility of the research results, which was accompanied by a multilingual dissemination strategy (English, Hungarian and German; see, e.g. Scheibelhofer et al. [2016a, 2016b]). A high level of sensitivity in regard to how findings were presented was necessary to avoid their improper use for political objectives (as was observed in the above-mentioned media scandal). In this respect, another important methodological point was the tension between analysing and constituting difference when researching diversity (Mecheril and Melter 2012). Grounded theory constitutes an ideal approach in this connection, as it emphasises that each category must earn its way into the data and thus enhances the constant scrutiny of external concepts applied. Furthermore, the problem-centred interview technique (Scheibelhofer 2008; Witzel and Reiter 2012) offered the advantage that it enables accounting for interviewees’ subjective interpretations without imposing a preconceived view of the problem onto their narrations. An additional specific challenge in the project was that several interviewees solicited the researchers to help them access official channels and claim benefits. The researchers attempted to provide assistance (e.g. accompanying them to public offices, translating documents and interpreting in administrative procedures) whenever requested and possible, as the matters were of critical importance to the respective migrants’ individual life chances. Once again, the systematic methodological reflection of the researcher’s involvement and positioning, as promoted by a constructivist grounded theory approach, was regarded as necessary to meet scientific criteria (Charmaz 2006).

‘Overstress on both sides’: the enhanced relevance of linguistic problems in the absence of a consistent institutional language strategy

The juxtaposition of the migrants’ and institutional perspectives suggests that the current manner of addressing linguistic diversity within the AMS is predominantly characterised by a shift of responsibilities to the individual, mainly the concerned claimant. This shift is accompanied by problems for both migrants and street-level bureaucrats (Lipsky 1980), which manifest themselves inter alia in language barriers, potential exclusion from benefits and services as well as in pressure exerted on employees. Because the importance of language in the process of labour market integration (which turned out to be especially relevant in regard to deskilling from the perspectives of institutional actors and migrants, see Holzinger forthcoming) is beyond the scope of this paper, as its focus is limited to accessing AMS services. However, it should be noted that the AMS is the main provider of German language courses for unemployed migrants, as well as in basic qualifications (including literacy training). The services considered in this article comprise mainly information, job search assistance and training opportunities as well as access to financial support. Advisors do not require an educational background in social work but must complete internal basic training at the AMS.

As stated earlier, this article departs from the assumption that multilingualism is not an individual but a relational competence (Blommaert, Collins, and Slembrouck 2005) and that, consequently, individuals’ linguistic capacities to act are instrumentally conditioned by language regimes, institutional constraints and facilitations. Language barriers are thus not to be considered as a priori given but as socially constructed and contingent. The ways in which institutions deal with linguistic diversity fundamentally influence the individual deployment of resources. Regarding the AMS, the analyses in this paper suggest that the described language barriers are exacerbated in the specific institutional work contexts that are characterised by migration-related super-diversity (having increased after the EU enlargements in 2004 and 2007), rising unemployment rates and bureaucratic reform processes under the heading of new public management (for a recent study on the work conditions of AMS employees, see Penz et al. [2017]). All interviewed street-level officials perceived high pressure in terms of temporal, financial and personnel resources, in addition to considerable legal complexity and bureaucratisation. Furthermore, the AMS’ internal control system contributed to the high level of stress experienced by the interviewees. Language-related barriers were often presented as an additional issue that further increases the workload to be handled by the street-level bureaucrats. Considering chronic underfunding and understaffing resulting from the combination of rising unemployment rates and cutbacks, communication issues were described as a type of ‘luxury problem’. However, the analyses show that linguistic barriers are in part created, or at least decisively conditioned, by the general institutional contexts and the high level of pressure exerted on street-level bureaucrats. The rigid timelines (standardised short intervals per client) and the scarcity of written information, in particular, render linguistic diversity problematic. The absence of a common institutional strategy in terms of language(s) and linguistic practices, which is closely related to the neglect (and to a certain degree, denial) of communication problems, is critical in this regard. Public life is inescapably ‘languaged’ and public institutions necessarily operate in and through language, thereby privileging certain groups with certain language repertoires over others (Brubaker 2014). Thus, the state and its institutions cannot take a neutral stance towards language(s) but, as language regimes decisively impact individual life chances, language regulations and linguistic practices in public life are inevitably politicised (Ibid.). When researching the AMS, there was an identifiable absence of a common institutional strategy in terms of linguistic practices, which ensued from an apparent refusal on the institutional side to address issues of unequal access to language-mediated resources (encompassing services as well as benefits). Issues of communication problems in the context of linguistic diversity were not articulated clearly by the institution. The consequential lack of clear institutional guidelines entails a shift of responsibilities to individual actors because communication problems become foremost imminent at the individual level, leading (in the words of an AMS advisor in Vienna) to experiences of ‘overstress […] on both sides’, i.e. for both the advisors and the migrants.

The following analysis demonstrates how migrants with certain language repertoires are disadvantaged when accessing services and benefits of the AMS. It will thus distinguish between ‘manifest’ and ‘symbolic’ language-related barriers. It next explores both institutional actors’ and migrants’ strategies and illustrates how language-related tasks fall within the responsibility of individuals and consequently lead to experiences of social inequity.

‘First, I don’t speak German well’: manifest and symbolic language-related barriers when accessing services of the AMS

As mentioned above, given the ‘pervasiveness of language’ (Brubaker 2014, 3) in public life, AMS claimants with certain language repertoires are disadvantaged in their access to (necessarily) language-mediated services and benefits. This became apparent in the empirical data, which identified a range of barriers that both the officials and migrants described. Although these barriers usually occurred simultaneously (and also intertwined with other discrimination mechanisms), an analytical differentiation between what can be described as ‘manifest’ and ‘symbolic’ language barriers is indicated. The first category applies to experiences of apparently impeded oral and written communication between the institution and migrants. Both migrants and street-level bureaucrats related that they did not understand the information provided by their interlocutors or could not make themselves understood. In this regard, street-level officials often stated that they would require more support than they currently have at their disposal (such as translators or multilingual information material) to communicate with migrants and provide them with relevant information (‘The question is always whether you have understood each other. I am often not sure whether my clients understand me correctly’). Furthermore, the migrants pointed out that they found it difficult to inform themselves about and consequently insist on their rights. In particular, written information was described as scarce, while very high competence in German (termed ‘legal literacy’, as it comprises technical language expertise) was mostly needed to understand the complex legal regulations. Several instances were observed in which AMS decisions detrimental to the migrants (rejection of applicationsfor unemployment benefits) were based on language-related misunderstandings, i.e. in which migrants claimed to have misunderstood the question and/or to have expressed themselves incorrectly. These instances have far-reaching consequences, as the following example shows. To appeal the rejection of his application for unemployment benefit, Ferenc, a Hungarian plumber who had been working in Austria for the previous five years, emphasised the following in a letter to the AMS: Finally, after over a year of investing a considerable amount of resources (time, money) and by calling upon the help of his wife, other social contacts and even legal assistance, Ferenc received unemployment benefits in arrears for a short period of seasonal unemployment. Language (especially the complex legal formulations employed in the written communication with the AMS) thereby played a decisive role, as Ferenc had serious problems understanding the reasons for the rejection of his application – which in turn made it more difficult to contest the decision. In this study’s sample, a further migrant’s application for unemployment was dismissed in the first instance for the same reasons as in Ferenc’s case: The AMS categorised both concerned migrants as cross-border workers from Hungary and consequently declined the competence for the benefit payments. A questionnaire, which was handed out to the applicants during their interviews with the street-level bureaucrats, was critical in this context. The questionnaire’s aim (which, however, was not explicitly explained to the interviewees at the time of its distribution) was to determine the competence of the AMS for the migrants’ claims. The statement made by Ferenc’s wife – the questionnaire ‘was easy to misunderstand’ – was confirmed by one of our researchers whose first language is German and who accompanied the interviewee to the AMS. Misleading formulations and ambiguous wording made the questionnaire prone to misunderstandings, with potentially severe consequences for the individual migrant who risks losing their entitlement to Austrian unemployment benefits. In such instances, it thus became clearly visible how ‘manifest’ language-related barriers may translate into concrete experiences of social inequity.

First, I don’t speak German well and my partner does understand German, but it is not the mother tongue of Ms. Veronika Nagyne, so she couldn’t understand everything either, the German officialese, and Mr. Huber [the responsible AMS advisor] doesn’t help. (Ferenc, 46 years old, plumber)

In addition to observing these ‘manifest’ instances of linguistic disadvantages in the empirical data, latent (ideological) structures which condition the access to benefits and services were also present. Instances of ‘symbolic barriers’ were identified, described in the interviews with both officials and migrants alike, in which competence in German implicitly functioned (without a legal basis) as a symbolic exclusion criterion. Some officials perceived applicants with limited German skills as undeserving, remarking that migrants ‘should learn German first’ before making their (legally rightful) claims. This attitude is exemplified in the following statement made by the head of a regional AMS office: Although our interviewee justified his support for a German-only policy with allegedly ‘pragmatic’ reasons, his quote reproduces arguments of a national closure of services offered by public institutions and thus illustrates succinctly how language may function as a symbolic instrument of exclusion. As has been argued widely in the literature (Bourdieu 1982; Brubaker 2014; Busch 2009; Gal 2006; Stevenson 2006), linguistic homogenisation has been at the core of nation state-building processes in the modern era, making language ‘a central focus of personal and collective identity, and a key terrain of political struggle’ (Brubaker 2014, 6), which also justifies the exclusion of those who are not part of the (linguistic) national community. On a symbolic level, this has entailed the ‘stigmatisation of non-standard varieties and non-prevailing languages’ (16), i.e. a devaluation of the ‘illegitimate’ language (Bourdieu 1982), which is also interiorised by its speakers through the deployment of symbolic violence (Bourdieu 1982).

But of course, there are attitudes in the organisation that say: Well, if you want something from us, then you have to speak German. Yes. I can understand this as well, yes. […] But this is not the expression of a political stance, I wish to clarify, but … well, I am also rather in favour of saying: There is no sense in talking to people in their mother tongue if we have to place them in sectors where they can’t use their mother tongue but simply have to learn German. (Director of a regional AMS office)

The data accordingly revealed instances in which migrants, due to their (allegedly insufficient) proficiency in German, felt less self-confident in interacting with officials or even less entitled to benefits – and consequently tended to waive their rights and claims. They were considered (by themselves and others) as ‘illegitimate’ speakers due to their insufficient competences of the ‘legitimate language’ (Bourdieu 1982), i.e. the language spoken by the dominant group in a given context. In this regard, the aforementioned concept of symbolic violence can elucidate how status positions are ascribed through linguistic practices and asymmetrical language competences – and how, consequently, certain speakers fall silent in official contexts or are even silenced, as shown by the extreme case of Lajos, a construction worker who had been working in Austria for several years. When requesting to register himself as unemployed, he was at first denied the right to file his claim by the clerk at the reception, who sent him away with harsh words and referred him to the Hungarian Employment Service: Although his sister, whom Lajos called for help and whose German was better than his, was finally able to solve the problem, this case vividly shows how some applicants may be barred from accessing their social rights because they are not acknowledged as legitimate speakers (see also Scheibelhofer and Holzinger [2018]). Lajos’ example also illustrates paradigmatically the intersection of different dimensions and categories of discrimination. In his narration, instances of discrimination on the base of nationality, language competence and possibly socio-economic status are almost inextricably intertwined. While Lajos is thus an extreme case – as he was explicitly denied his right to speak – symbolic barriers can also take effect on a more implicit level, leading migrants to fall silent and consequently not stand up for their rights merely on the base of anticipated symbolic sanctions. Speakers who perceive themselves as insufficiently competent in the legitimate language may feel embarrassed when communicating, as they have interiorised the symbolic domination. This is illustrated aptly in the following interview passage in which an interviewed AMS advisor ascribed Hungarian applicants a certain feeling of unease when speaking German: Knowledge of German thus proves to be of vital importance in order to insist on one’s social rights. As an AMS advisor at a Viennese regional office put it: ‘If you speak the language, you can defend yourself. Well … you also do not have to let them fool you’. This applies to both manifest and symbolic barriers, as demonstrated by the case of Dorina, a 27-year-old white-collar worker who had completed her bachelor studies at the University of Vienna. When applying for unemployment benefits after having worked in a prestigious company for several years in Austria, the responsible AMS advisor confronted Dorina with what she perceived as irritating questions regarding her ‘centre of vital interest’. In Dorina’s words, the official ‘wanted to convince [her] somehow’ that she was in Hungary quite often, thus indirectly suggesting that (to her own disadvantage) Dorina could be considered a commuter whose case would fall into the competency of the respective Hungarian institution, resulting in considerably lower unemployment benefits. Dorina, whose proficiency in German is very high, eloquently defended herself from this categorisation and the official’s attempt to exclude her from the Austrian welfare community by suggesting that she in fact belonged to another one, namely the Hungarian welfare state (‘Indeed, I think I said it three or four times: Yes, I live in Vienna, I do not want to go to Hungary’). Dorina’s narration thus shows how proficiency in the official language can be a tool to defend oneself from exclusion. However, her case illustrates, not only on a manifest level, how linguistic competence is critical to understand and adequately respond to questions made by street-level bureaucrats. It also reveals how such competence plays a central role on a symbolic level by differentiating between insiders and outsiders, legitimate and illegitimate speakers – thus decisively influencing their disposition to advocate for their social rights.

She told me […] I should leave. She told me to go home, to Hungary. And I was like, why should I? I said: I have been here for five years and that’s it. [She] was like ‘pfft, Hungarians’, and threw my papers down. (Lajos, 49 years old, construction worker)

I have the feeling that they feel a bit embarrassed then for not being able to express themselves in German as they would like to. I think for them, it is easier if someone would sit in front of them who speaks Hungarian. (AMS advisor at a regional office in Lower Austria)

‘You can’t speak German to people who do not speak German’: the individualisation of institutional communication responsibilities

Although the existence of language-related barriers was rarely admitted explicitly by the interviewed institutional actors, they narrated how challenges ensuing from linguistic diversity are to be met by the employees directly dealing with clients. Summarised by the aforementioned director of a regional branch of the AMS: ‘All of the consulting services are in German, but naturally we know that you can’t speak German to people who do not speak German. That means we have several possibilities’. Concerning these ‘possibilities’ of handling language-related problems, the interviewees mentioned a (frequently disjointed) variety of strategies, ranging from communicating with gestures and hand signals, the use of a lingua franca (mostly English) to the deployment of technologies (e.g. pilot testing online interpreting via video). However, in the absence of a consistent institutional strategy, what turned out to be of central importance was the shuffling off of responsibilities onto individuals. Three aspects are of special interest to this research focus and will thus be elaborated in the following: First, a certain ambiguity was discerned in regard to employees with a migration background. Second, the externalisation of language-related tasks and competencies became evident. Third, the ensuing importance of lay interpreters enhanced the relevance of social capital.

Regarding the first point, observed communication problems were in part internally shifted onto the individual employees who were in direct contact with clients. Implicitly, a special status was thereby accorded to employees with a migration background, especially to those who spoke a common ‘migrant language’, such as Turkish or Serbian. However, a certain level of ambiguity in this regard became discernible: While non-German consultationwas presented as not desirable (although officials in Austria are not prohibited to do so by law), reliance on multilingual officials was reported. For example, a senior executive responsible for a diversity agenda stated: ‘Of course, advisors with a migration background are deliberately chosen […] BUT not to offer advice in the mother tongue’. The refusal to offer advice in any language other than German was mainly justified by referring to German as the sole official language (e.g. ‘Our executive committee always says: the official language is German’, employee at the AMS national executive office) and its importance for labour market integration. Another recurrent reasoning was the avoidance of ‘fraternisation’ among co-ethnics and the protection of multilingual employees from clients’ attempts to ensure favourable treatment based on an assumedly shared cultural identity. Still, occasional reliance on multilingual employees was also mentioned recurrently and vindicated mainly on the grounds of pragmatic reasons. The fact that multilingual employees indeed provided unofficially valuable (and necessary) services with respect to communication problems was even confirmed by interviewees who generally defended a ‘German-only’ standpoint, such as the aforementioned head of a regional office in Vienna. According to this director, one strategy to solve language-related problems is to rely on ‘colleagues whom we somehow can exploit a bit if we need information’.

Second, an institutional strategy of externalising/outsourcing language-related tasks and competencies was identified. This study’s empirical data provides evidence that the AMS often refutes competence in regard to communication problems. Ensuring understanding was implicitly seen as an obligation to be fulfilled by migrants. This strategy is clearly represented by the only multilingual document that several interviewees described, namely a document which states in several languages that counselling is only offered in German, urging those not sufficiently proficient to bring along an interpreter. As one interviewee openly admitted, AMS employees often had to ask their clients to ‘bring along someone who can interpret’, because they themselves ‘are unable to manage it’. In the case of EU migrants, a recurrent reasoning that served to justify the dismissal of institutional competence for ensuring that communication and understanding were guaranteed was discerned: By opposing them to refugees (who were described as vulnerable and indigent), EU migrants were positioned as independent, self-sufficient and consequently not in need of institutional help. The creation of special service points for refugees was emphasised, in which first-language consultation was offered. In contrast, EU migrants ‘who are willing to move’ were repeatedly described as being capable of autonomously finding their way through the labyrinths of social security (Scheibelhofer, Holzinger, and Regös forthcoming). This constructed image of self-sufficient EU citizens thus serves to justify the shift of language-related responsibilities onto the individual (‘People from the EU area are rather used to this, they know this very well. […] Hungary is the best example, […] they know the ropes perfectly’) – thus consequently relieving state institutions of this task. As a senior executive at the EURES network, who emphasised the importance of individual efforts when migrating, put it: ‘You can’t expect all states to be structured like a supermarket’.

Third, due to the observed externalisation of language-related tasks, a major role is attached to lay interpreters who are largely to be arranged for by the unemployed migrants themselves (as indicated in the quote above). While NGOs were reported to provide lay interpreters on behalf of refugees, EU migrants were described as being left on their own. In all analysed cases, the respective migrants were observed as resorting to relatives, friends, colleagues and/or even to us researchers to help in translating, interpreting and writing letters or to accompany them. Such reliance on lay interpreters is problematic for at least two reasons. First, mostly reported by institutional actors, is that quality control cannot be guaranteed by non-professional interpreters/translators. The following quote, in which Ferenc’s wife Veronika described her unorthodox strategy of composing a complaint letter for her husband, exemplifies the fact that lay interpreters themselves are often overwhelmed by their tasks: This highly creative strategy applied by the interviewee indicates extensive cross-language legal literacy (i.e. cultural capital) but also demonstrates that lay interpreters cannot offer professional services. The second reason that reliance on lay interpreters is problematic has even further-reaching consequences for individual life chances: The fact that in general, lay interpreters must be organised by the migrants themselves, implying that social capital (Bourdieu 1986), is highly significant when accessing services of the AMS. Migrants thus must not only activate cultural capital resources (language competence, education, ‘legal literacy’) or financial means (e.g. for the official translation of documents or even legal assistance). Social capital also proves to be important in this aspect (for a general discussion of the importance of capital forms in our research, see Regös, Holzinger, and Scheibelhofer forthcoming). As social capital (similar to other forms of capital) is distributed unevenly among claimants, reliance on lay interpreters thus ultimately leads to experiences of inequity. Co-ethnic help through websites and associations (especially in terms of an initial orientation in the welfare system) was reported to compensate to a certain extent the absence of a social net, but reliance on family, friends or colleagues able to help proved to be of utmost importance (see Scheibelhofer, Holzinger, and Regös forthcoming; Regös, Holzinger, and Scheibelhofer forthcoming). Teréz, who received unemployment benefits only after appealing a prior rejection of her claim, highlighted the vital support her Austrian partner provided: While forms of capital are unevenly distributed among members of society and thus may also result in socially stratified experiences among Austrian-born unemployed individuals, Teréz’ quote reinforced our findings, namely that EU migrants face additional specific linguistic barriers when accessing AMS services.

And I had no other choice than to go as far as to observe the […] EU Charter of Fundamental Rights, what is said regarding mobility and work. […] you can check it on the website of the EU in Hungarian, in German. Thus, I read it in Hungarian, I understand the law quite well, and afterwards I copied the German version of this part of the Charter and forwarded it to the court. (Veronika, Ferenc’s wife)

The problem is that the Hungarians, if you don’t have a help like I have [reference to her Austrian partner], if they don’t have it, then the Austrians would know what to do. They [the Hungarians] wouldn’t know where to look for help. (Teréz, 44 years old, former schoolteacher in Hungary/secretary in Austria)

Conclusion: linguistic barriers impeding participation

The present article aimed to make an innovative contribution to the understanding of migrants’ language-related inequity experiences by exploring how both migrants and institutional actors perceive the handling of linguistic diversity, which problems and coping strategies they identify in this respect and their implications for experiences of social (in)equity. We identified that although public and political discourse mainly focuses on cultural and religious otherness in regard to the institutional handling of diversity, linguistic barriers can essentially be defined (whether on the manifest or the symbolic level). Our empirical data shows how language skills (both of the official language German but also cross-language skills such as ‘legal literacy’) must be capitalised upon to access services and benefits of the AMS – thus contributing to the social stratification of access. Furthermore, the social function of language, i.e. the symbolic domination and hierarchisation among more or less ‘legitimate speakers’ (Bourdieu 1982), constitutes an important dimension of linguistic discrimination.

As stated initially, it is often difficult to separate discrimination on the basis of language from other types of discrimination. Our analyses were thus geared to the disentanglement of language-related barriers from other socially constructed barriers in order to disclose mechanisms of discrimination based largely on language practices. This was demonstrated within the case study of Hungarian migrants in Vienna who represent a suitable research group in this regard. As set out above, Hungarians are less perceived as culturally different than other migrant groups,are ascribed system knowledge and portrayed as not in need of institutional help. Yet, it became evident that, even for such an apparently privileged group, barriers exist that can largely be traced back to institutional language practices. Drawing on Skutnabb-Kangas (2015, 2), we refer to these practices ‘where people get unequal access to power and both material and immaterial resources, based on their language/s’ as instances of linguistic discrimination, i.e. linguicism.It must be noted that such experiences of linguistic discrimination are not limited to Hungarians in Austria or migrants in general. Rather, critical inequalities also exist among ‘native speakers’ who are more or less skilled in the use of the ‘legitimate language’ (Bourdieu 1982). This becomes particularly evident with respect to legal literacy, which is a requirement for effective participation in modern societybut whose mastery correlates strongly with educational background and social status. However, as argued above, the social function of language is of special relevance in regard to questions of multilingualism. This was also perceived as such by the interviewed officials who presented language issues almost exclusively in relation with migrants. Hungarians were thereby not described as more different than migrants from other non-German speaking countries.

Our empirical material provided evidence that communication problems are to be considered as socially constructed, as ‘the result of how individuals and their communicative “baggage” are inserted into regimes of language’ valid in a particular space (Blommaert, Collins, and Slembrouck 2005, 198). In the researched space of the AMS, the lack of consideration of structural problems concerning the institutional handling of linguistic diversity and the consequent absence of a common institutional strategy leads – especially under the conditions of limited financial resources and neoliberal restructuring in the public sector – to responsibility being delegated to individuals. Street-level bureaucrats find themselves under significant pressure and migrants in situations of social inequity. If we consider our particular findings regarding the AMS in a wider, European context, we see them as embedded in what welfare scholars termed a new European social project that strongly endorses the active participation of citizens, while burdening the individual rather than the state. As Soysal (2012, 14) argued, ‘the factors and conditions that entrench differential capacities and the very obstacles to parity of individuality’ are not equally addressed in this new social project. The ‘emerging fault lines’ (15) thus also cut through Europe and exclude the ‘lesser Europeans’ (15), i.e. those who face ethnic, religious and – we must add – linguistic discrimination. We thus advocate the importance of further research to investigate language-related barriers disentangled from ideological debates and to declare their identity as instances of linguistic discrimination that hinder social equity and participation in de facto multilingual societies.