(e)Health literacy brokering: bridging sociolinguistic gaps at the welfare office?

3.1 (e)Health literacy as social capital

In 2019, Norway’s Ministry of Health and Care Services put forth an action plan seeking the enhancement of the Norwegian population’s health literacy (Helse- og omsorgsdepartementet 2019), in line with similar public health actions elsewhere. Literacy can be narrowly defined as a person’s reading and writing abilities, but has been expanded by a range of sub-definitions, including digital literacy, which refers to people’s abilities to use media and technologies, as well as to read, write, and interact in online spaces (e.g., Perry 2009: 266). Meanwhile, health literacy concerns a bundle of competencies, broadly described as the ability to elicit, understand, and critically assess information, as well as interactive dimensions involving individuals’ ability to make meaning of/in interactions with healthcare providers, navigate the healthcare system, and manage health needs by applying information “to changing circumstances” in everyday life (Nutbeam 2008: 2075).

eHealth literacy, namely “the ability to seek, find, understand, and appraise health information from electronic sources and apply the knowledge gained to addressing or solving a health problem” (Norman and Skinner 2006, 2), is also increasingly relevant due to the digitization of Norway’s public sector. Limited eHealth literacy has been associated with social inequality: in fact, research in the Norwegian context has identified a correlation between low socioeconomic class, low levels of education, and poor eHealth literacy among the population in general (Brørs et al. 2020) and migrant populations in particular (Le et al. 2021: 9). In other words, individuals’ health literacies go beyond individual abilities, and are interfaced with institutional mechanisms and demands (e.g., Pelikan 2019). The ability to deal with text is mediated by the social context that it takes place in (Papen 2009: 23). Bourdieusian theory has contributed to critical literacy studies for decades (Carrington and Luke 1997: 97–99), bringing attention to the way that literacy intersects with social and linguistic capital (Bourdieu 1977). It has also gained traction in recent orientations toward the “social embeddedness” of health literacy (Bauer 2019), some of which have illuminated social capital as a central aspect of health literacy (Kim et al. 2015; Yang et al. 2013).

3.2 Literacy brokering

Literacy brokering has received significant attention in sociolinguistic research on multilingual contexts, including in the family (e.g., López 2020) and the workplace (e.g., Gonçalves and Schluter 2017). While an institutional interpreter in the Norwegian context needs to hold official qualifications (Buzungu 2022), the broker’s role is less formalized and more versatile, assumed by co-workers, friends, and family members. Tse (1995: 243) defines a language broker as someone who “facilitates communication between two linguistically and/or culturally different parties”. According to Perry (2009: 256, 262), a literacy broker is someone providing “informal assistance about some aspect of a given text or literacy practice” as well as different genres. Building on previous definitions, I propose a health literacy broker to be someone with sufficient knowledge of dominant languages and inherent language ideologies to mediate content, adjust epistemic asymmetries, establish common ground, and offer access to knowledge territories between two or more parties, as such enabling these parties to either manage their own health needs, or support others in doing so. An eHealth literacy broker holds additional technical abilities, enabling them to mediate between offline/online spaces.

The role of the broker is often dual, shedding light on underlying structures of power. Accordingly, Kraft (2020: 91) describes brokering as a “Janus-faced practice”: while brokering may bridge gaps between different parties within an interaction – parties that would not otherwise understand one another – brokering strategies support a particular logic and may result in increasing the responsibility of individuals. Kraft (2020: 81) writes that “despite brokers being a workplace need, the responsibility of becoming a broker is left with the individual worker”, a power dynamic that is transferable to other institutional contexts, including the welfare system. Several participants in my study relied on their husbands’ brokering, a finding that is also echoed in a study of how women from Thailand navigate Norway’s healthcare (Tschirhart et al. 2019). While NAV stipulates that “[f]amily, friends and children cannot interpret for you” (NAV 2022), Buzungu’s (2021: 102–103) research in Norway has indicates a few instances of family brokering, although only in cases in which social workers were not familiar with institutional stipulations.

In some cases, brokering may reproduce gender inequalities because migrant women, by relying on their husbands, may lose control over the content in their personal files, as well as other personal information. If the husband holds a dominating role in the relationship, brokering strategies may exacerbate the woman’s vulnerability. However, this was not the case for Suda and Maria, who were positive toward their husband’s brokering, and exercised control over when, how often, and how this strategy would take place. When communication misfired, this was seemingly not due to their underlying relationships with their husbands, but rather due to the bureaucratic tasks they were faced with, as I soon turn to illustrate. It is important to note that brokering is not unique to migrant women. Most people rely on support from their social network regardless of gender, socioeconomic class, and language background. In this analysis, I seek to contribute to the sociolinguistic literature on brokering by highlighting the under-examined experiences of people at the receiving end of brokering, rather than of brokers themselves. Additionally, this study sheds light on the complex relationship of brokering to power dynamics in interactions between migrants and the state.

3.3 Welfare state communication

Norwegian welfarism is based on egalitarian principles of universal tax-financed services, and individuals may access benefits in cases of illness and disability, among others. Norway has grown increasingly multilingual in recent decades, and to ensure the accessibility of public services for migrant individuals, the Norwegian Labour and Welfare Administration’s (NAV) language policy states that all clients are entitled to institutional interpreting by formally certified professionals in face-to-face appointments with their caseworkers (NAV 2022). This policy was previously based on the recommendations of The Directorate of Integration and Diversity (2020), but has since (after my data collection ended) been set out as part of Norwegian law (Tolkeloven 2021). Research has shown, however, that many employees are either not privy to the stipulated guidelines, or do not know how to implement them (Malmberg-Heimonen et al. 2019). Moreover, recent digitization efforts have led to a lower frequency of in-person meetings at NAV, which instead relies heavily on their website to communicate with the population. Due to increasingly textual communication online, clients with limited linguistic and digital literacy may be encouraged to mobilize alternative (and sometimes less reliable) strategies to understand information provided by the state – Google Translate, for instance (Bønnhoff 2019: 335), or, as I illustrate in this article, their social networks.

Indeed, recent digitization trends in the Norwegian public sector have rendered communication more textual than ever before (via messages, application forms, letters, and information published on websites), requiring individuals to be more digital – i.e., hold an e-mail address, an electronic ID, a smart phone, and so on (Bønnhoff 2020: 5). While increased digitization may ensure efficient management of tasks for some, Bønnhoff (2019: 335) has illuminated some of the mechanisms through which digitization increases the digital labor of individuals, a workload that may hamper accessibility especially for individuals who have learned Norwegian later in life and/or have limited digital literacy. In addition to Norwegian Bokmål, the Norwegian welfare office’s website is partially available in Norwegian Nynorsk, English, and Sámi. However, when I signed into my own profile on the website (all Norwegian residents have a digital profile that they can access by use of their personal electronic ID), I could not access all the relevant application forms in English – even after I changed the standard language on my profile into English. Lack of access to public sector websites (and translated information) goes against the state’s egalitarian ideals by leaving some clients more vulnerable. This calls for more research on “bottom-up views” on “multilingual state websites” and the “user experiences of particular online services” (Berezkina 2018: 72).

3.4 Structural vulnerabilities and constrained literacies

The vulnerability that some migrants experience in encounters with public institutions has sometimes been associated with low health literacy (Gele et al. 2016). By examining health literacy in a social context, I take as a premise that health literacies are constituted not by individuals’ abilities only, but by how individuals are enabled and constrained by the social context that they act within. I approach this via the prism of structural vulnerability, namely the “condition of being at risk for negative health outcomes through their interface with … multiple overlapping and mutually reinforcing power hierarchies (e.g., socioeconomic, racial, cultural) and institutional and policy-level statuses (e.g., immigration status, labor force participation)” which in turn “constrain their ability to access health care and pursue healthy lifestyles” (Bourgois et al. 2017: 300). Having one’s health literacy constrained by structural factors – e.g., being unable to use one’s most comfortable language in an institutional interaction or having to operate technological devices that one is not familiar with – does not translate to being unresourceful or consistently vulnerable.

Indeed, I seek to avoid victimizing terminology, highlighting instead the structural conditions that make individuals vulnerable under specific conditions and in specific power dynamics. People who are unemployed (Wanberg 2012) and/or on long-term sick leave are in the research literature considered to be at risk of experiencing diminished mental wellbeing – a situation that research indicates affects some migrant women more than the general population (Nortvedt et al. 2016). Moreover, people with limited digital literacy (Le et al. 2021: 9) and low Norwegian proficiency (Le et al. 2021: 89) may experience barriers in encounters with the welfare system. Multiple factors played into the structural vulnerability that Maria and Suda experienced. For instance, they are positioned as second-language speakers and have varying digital and bureaucratic abilities. Their financial stability depends on the welfare system and may be experienced as unpredictable, all the while dealing with limited energy and health challenges.

We lack research on how migrants such as Maria and Suda practice communication and health literacy in interaction with the welfare system, and the role brokering plays in these interactions. What makes this comparative case of eHealth literacy brokering important is that it illuminates some of the paradoxes and potential structural vulnerabilities that emerge when individuals are encouraged to receive brokering assistance from their own network in pursuit of filling sociolinguistic gaps.

6.1 Negotiating epistemic stance

I introduced this article with a rich point, illuminating how Anne and Maria relied on Maria’s husband and his brokering to accomplish the goal of their communication, namely, to make sure that Maria could access institutional services from home. A similar negotiation transpires in an interaction between Suda and Frida, which takes place in a meeting room at the local welfare office. Frida has not brought a computer with her to their meeting. Instead, she goes on to offer (what I observe as) a lengthy and bureaucratic description of the tasks that Suda needs to complete online to access benefits. At the beginning of the following excerpt, Frida draws on a communication strategy that she later points out to have learned at a course convened by her workplace. Frida says,

Extract 1

Lines one to five outline a salient stance negotiation between Suda and Frida. Suda’s brief reiteration (3–4) is soon self-interrupted (4). When stalling, she is quick to affirm that she can just ask her husband (4). We do not know if Suda did not understand the information provided by Frida prior to her question (1), if she has forgotten this information, or if she does not know how to repeat it to Frida. But the stance-act that Suda constructs by invoking her access to a broker, in fact someone who is presumably as stable a part of her network as possible, her husband, indexes social capital. Moreover, it holds a smoothing function, bringing her conversation with Frida forward to the next point on the list, rather than disrupting it by having to return to a previous point. By bringing up her husband, Suda avoids asking Frida for additional assistance, although this exchange was vital, concerning the digital (status) form that Suda is required to send Frida biweekly in order to receive social assistance.

As a recipient of welfare benefits, Suda describes feeling indebted, because “de som jobber, de betaler skatt” [those that are working, they pay taxes], adding that “også den skatten som må betales for meg som ikke jobber” [and these taxes must be paid on my behalf, who is not working]. Suda goes on: “Jeg føler dårlig. Jeg vil heller jobbe selv, og få penger selv, ikke få penger fra de som skal betale for meg” [I feel bad. I’d rather work myself, get money myself, not get money from those that pay for me]. Norway’s welfare model builds on a political model of workfare, premising that active workers are more likely to achieve social integration, thus greater wellbeing, while also being financially independent, and able to contribute to the collective through paid taxes (Ylvisaker and Rugkåsa 2021: 6). Indeed, the welfare administration’s strives to ensure “[m]ore people active and in work, fewer people on benefits” (NAV 2020). The stance that Suda builds, as exemplified above, stands in direct correlation to her experiences with this political ideal: Suda positions herself as someone who wishes to assume responsibility for herself, not be taken care of. Suda frequently constructs responsible stances in our interviews. To take another example, Suda once tells me how important it is for women to be active workers, not only for their own benefit, but because their husbands should not bear the whole economic responsibility alone. In these kinds of stance-acts, Suda positions herself against dominant gender discourses, claiming a stance of independence, but also responsibility and power to be an equal part in providing for her family.

Let us return to Extract 1. Frida’s aligns with Suda, reminding her of the welfare office’s website (5–6), which is integral to the agency’s administrative procedures. Frida goes on,

Extract 2

After Frida has pointed out the issue of digital literacy (9–10), Suda aligns with her (11–12), through which they form a moment of solidarity. This shared stance negotiation resembles what Goffman (1967: 27) has coined cooperative face-work – strategies that are invoked to manage sensitive conversation while also restoring harmony between stancetakers. Indeed, Frida tries to “save [Suda’s] face” (Hua et al. 2019: 496), possibly to avoid jeopardizing Suda’s epistemic stance and authority. In the end, though, this face-saving strategy conceals, rather than addressing, communicative dilemmas between Suda and her caseworker. This moment illuminates some of the “conflicting pressures” that Frida, like other social workers in Norway’s welfare state (Ylvisaker and Rugkåsa 2021: 3), finds herself confined by. Another caseworker that I interviewed, Ingrid, explained that she often avoided asking clients if they understood her when she spoke Norwegian, worrying that they would find it “støtende” [offensive]. Such stance-constructs indicate a tension between caseworkers’ responsibility to make sure that their clients feel comfortable, and their responsibility to ensure that their clients’ rights, duties, and access to information are maintained.

Frida and Suda have just negotiated a decrease in epistemic stance (11–12): They both struggle with screens. Suda now goes on to further downgrade her epistemic position by stating that she is “not so skilled with computers” (13) at all. Frida asks a rather leading question: “But you think it will be okay to submit the status form?” (14). After all, Frida needs to make sure that Suda has the digital abilities necessary to access social security forms. Having already received an indication that Suda may in fact not be able to do this, Frida may nonetheless “experience a sense of powerlessness”, feeling like there is little she can do to ensure that Suda’s rights are “sufficiently safeguarded” (Ylvisaker and Rugkåsa 2021: 8). Receiving these signals from Frida (in form of a leading question), Suda forms what Stivers (2004: 271) has dubbed a multiple stancetaking – repeating “yes” four times, thereby “display[ing] her stance that this trajectory of talk should properly halt and that further talk is unwarranted”. This is a stance-act that I have witnessed Suda invoke several times, especially when we get to sensitive topics. Suda and Frida’s conversation illustrates a struggle to maintain harmony due to conflicting interests between face-saving and ensuring Suda’s rights, between claiming knowledge and admitting to needing further assistance. Suda’s ability to use brokering is oriented to as a positive strategy in the collaborative face-work undertaken by Suda and Frida, similar to the negotiations that took place between Maria and Anne. I now turn to examine how the women’s language habitus and symbolic capital (see Bourdieu 1977, 1990) have affected their (e)Health literacies and their use of brokering as a strategy.

6.2 Symbolic capital

Maria and Suda share a similar habitus in many ways. They both migrated to Norway from countries in Asia to be with their Norwegian husbands and have attended two years of Norwegian language courses. Both women grew up with lower socioeconomic status and neither of them attended higher education. Instead, they both entered the unskilled, blue-collar workforce in their mid-teens and continued to work until they migrated to Norway. Suda experienced upward mobility in terms of her socioeconomic situation when migrating to Norway on a marriage visa. With few child-rearing responsibilities, few financial worries, and no work commute (because her family can afford housing close to her workplace), she is free to spend her spare time with her husband, his relatives, her Norwegian colleagues, and her friends, including many with whom she can practice Norwegian. While Suda grew up speaking a language with a different alphabet than the one used in Norwegian – which is based on the Latin alphabet – she may have already known the Latin alphabet because of her proficiency in English. In our conversations, Suda refers to a vast Norwegian network of friends and family. She tells me several stories about how her network has contributed to her accumulation of linguistic capital: when she first moved to Norway, her father-in-law, with whom she enjoys a close relationship, showed up at her house, surprising her with learning material that he had bought and offering to help her prepare for Norwegian class. Her friends at the language courses also helped her.

Intimate relationships equate to social capital for Suda, in turn facilitating her accumulation of linguistic capital in multiple ways. These strategies have been instrumental to Suda’s access to Norway’s job market in the past, and currently to her health literacy to access healthcare and the welfare office. While Suda used to bring her husband to institutional encounters, she now reports being comfortable communicating with health professionals and caseworkers on her own. She also explains that she reads and replies to most messages from these institutions (occurring in her message portals on the different public sector websites) on her own. I, too, observe this when I visit Suda in her home. If Suda is unsure of something or does not understand the information she has been provided with, she asks for help. She is often around people who can broker for her. Providing a concrete example, Suda explains that if she receives a text message from the welfare office or other institutions during workhours, she might ask one of her colleagues to translate the content for her. Similarly, if Suda is at home when she receives such a message, she asks her husband. He also helps her manage letters and other written bureaucracy. In other words, Suda’s (e)Health literacy is ensured by bridging gaps in linguistic capital and digital literacy with social capital. These forms of capital enable valuable transactions, ensuring Suda’s access to information and social benefits, and in turn enabling her to manage her health needs.

Maria, on the other hand, did not experience the same degree of upward mobility when arriving in Norway. She was literate in the Latin alphabet, but this does not seem to have translated into a significant advantage when learning Norwegian. She does not mention a Norwegian social network and underscores, anyway, that she does not have a lot of time to spend with friends. Maria works a blue-collar job, for the most part among other coworkers with migrant backgrounds. While they are expected to communicate in Norwegian only, Maria explains that they also communicate in English and other shared languages. Maria’s limited opportunities to practice Norwegian at work may limit her possibilities of accumulating linguistic capital. Maria and her husband cannot afford housing close to their respective workplaces, and thus have long work commutes. Her husband is one of the only people with whom she regularly speaks Norwegian, but due to this commute, they usually see each other on the weekends only. Thus, Maria is also the primary caregiver for her young children during the workweek. She mostly speaks with her children in her mother tongue. Maria takes on what Hochschild (2012) has called a second shift, that is, domestic responsibilities that come in addition to paid workhours. This second shift compromises the time left for Maria to practice Norwegian. In heterosexual couples, second shifts are often assigned to women, in which case they sustain gendered inequalities (Hochschild 2012).

While Maria experiences fewer opportunities than Suda to accumulate linguistic capital both at her place of work and at home, Suda only speaks Norwegian in her workplace. Maria does not feel confident in Norwegian and explains in one of our conversations that she often resorts to English instead – a strategy that I observe during our interviews, as well as in Maria’s meeting with Anne. Suda, by comparison, only uses English words on a few occasions during our interviews, and only when in search for a Norwegian word. Based on my own linguistic background as someone who grew up speaking Norwegian, I assessed Suda’s Norwegian literacy to be more proficient than Maria’s. As previously described, Maria was also less comfortable with the digital communication required by NAV and appeared to rely almost entirely on brokering in order to achieve required communication, although she had a limited social network of potential brokers around her. In conclusion: just like Maria’s socioeconomic status and limited social capital have impeded her accumulation of linguistic capital, Suda’s socioeconomic status and social capital have richly enabled hers. The social conditions that enable and constrict Maria and Suda’s accumulation of linguistic capital play into their linguistic strategies to craft epistemic stances, including their stances towards brokering, which I now turn to examine.

6.3 Bridging gaps

In an interview between Maria and I after her meeting with Anne, Maria explains: “that day we meet, of course I am afraid [of] what I am going to do and what I am going to tell … Because of how [I] speak Norwegian. And I hope I understand everything what they say”. Maria downgrades her epistemic stance as a Norwegian speaker, a positioning act that she naturalizes by invocation of the adverb “of course”, that is, this stance-act is conclusive and warrants no further questions. In addition to her concern that she will not understand, nor be well understood by Anne in Norwegian, Maria explains that she “do[es] not want to show off- also I am weak”. This resonates with what Goffman (1967: 15) has described as an “avoidance strategy” to protect the stancetaker from social threats against their face. Maria’s stance claim concerns her self-perceived inability to articulate herself in a profitable way “within the limits of the constraints of the censorship that is imposed on the occupant of that position” (Bourdieu 1977: 658), that is, to speak in a style that legitimizes her in the given linguistic marketplace (at the welfare office). Although no such sentiments were expressed by Anne – quite the opposite – studies have shown that limited linguistic literacy is “at times, given moral value” by social workers “in the sense of being a character flaw and a sign of resistance to making an effort” to integrate into Norwegian society (Buzungu 2021: 108), a judgement that Maria may fear, and which may now affect her stance.

Despite high Norwegian proficiency, Suda describes similar linguistic insecurities concerning her Norwegian literacy, and potential consequences of her (limited) literacy on her ability to understand Frida and make herself understood. During one of my visits to Suda’s home, she expresses the following about her meeting with Frida: “Jeg var redd for- forstår hun meg hvis jeg svarer henne?” [I was worried that, does she understand me if I reply to her?]. And on the other hand: “Hvor mye skal jeg forstår når hun gir informasjon til meg, ikke sant?” [How much will I understand when she gives me information, right?]. When I ask Suda if she was also “redd for hva [Frida] skulle tenke?” [worried about what [Frida] would think?], Suda replies: “Mm, hva skal hun tenke om meg? … ‘Hvorfor ikke hun trenger tolk?’, ikke sant” [What will she think of me? … ‘Why doesn’t she need an interpreter?’ right]. She expounds that it may be “fordi jeg er utlender (…). Kanskje hun tenker jeg er ikke bra norsk” [Because I am foreigner (…). Maybe she thinks that my Norwegian is not good]. For Suda, need for interpretation indexes a downgrade in epistemic stance and authority.

Indeed, Suda has never used institutional interpreters. When she first arrived in Norway, she would always bring her husband. Now, her Norwegian proficiency is stronger, and she is less reliant on her husband’s brokering during such meetings. Maria, on the other hand, has used interpreters several times, but explains that she prefers her husband’s brokering. In fact, the plan was that he would join Maria’s appointment with Anne. In the end, though, he needed to go to work and could not join their meeting. In my conversation with Maria, she explains that when we first met at NAV, she had assumed that I was there as an interpreter. She reiterates relief: “Så jeg- åh, heldig at jeg har tolk!” [So I just, ah, lucky that I have an interpreter!], a reaction that demonstrates desirability for institutional – or any kind of – interpreting. When she realized that I was not her interpreter, and that no interpreter was booked for the meeting, Maria still did not go on to request interpreting (perhaps because she did not know that she could).

To summarize, both Suda and Maria have at some point relied on their husbands’ brokering to bridge gaps in interactions with civil servants. At present, Maria, who has limited Norwegian literacy, depends on interactional health literacy brokering. Suda, on the other hand, rarely does. Both women rely on eHealth literacy brokering to bridge gaps in their digital literacy, although again, Maria to a larger extent than Suda. For Suda, access to her husband’s brokering primarily constitutes a reassurance. She explains that his brokering increases her emotional security and control, making it “trygt … eller behagelig for meg hvis han er med, også han kan forklare … da vet han hva som skjer … jeg skal spørre han dagen etter, da husker han” [… safe … or comfortable for me if he joins and he can explain, … then he knows what is going on … I can ask him the next day, then he remembers]. This is an important point: While institutional interpreters offer mediation during the meeting, an interpreter can neither remember information for Suda nor assist her at home, i.e., perform eHealth literacy brokering. This also demonstrates the interplay between social capital and the two parameters of linguistic and digital literacy, demonstrating again how Suda’s social capital enables her health literacy and her access to institutional services, while Maria’s comparatively weaker social network has led to less linguistic capital and a more dependent form of brokering. In the next part of this section, I turn to the perspectives of the caseworkers.

6.4 Competing interests

In our interview, Anne tells me, “Jeg opplevde at [Maria] fikk med seg hovedessensen, og at hun kunne òg gå hjem til sin mann og få hjelp av ham til å forstå gjennom referatet, som ligger ute på Ditt NAV” [My experience is that [Maria] understood the essence and that she can always go home to her husband and get help from him to understand via the report, which is on Ditt NAV]. Based on the insecurities that Maria expressed in our interviews, it is likely that Anne overestimates Maria’s level of comprehension, an evaluation that echoes the findings of a broad report on healthcare personnel in interactions with migrant patients with limited Norwegian proficiency (Le 2013: 11). Anne valorizes Maria’s linguistic capital (her ability to understand the “essence” of what Anne told her), as well as her social capital (Maria’s husband), then continues, “jeg [har] ikke mulighet til å ta stilling til det i møtet, for det tar for mye tid og ressurser å følge opp om det ble forstått, da. Da er det på en måte brukers eget ansvar å spørre” [I cannot take a position on that during the meeting because it takes too much time and resources to follow up on whether it was understood. Then it is kind of the client’s responsibility to ask]. In other words, Anne brings up Maria’s linguistic and social capital, as well as her own limited power, to justify reducing her own responsible stance, explicitly assigning responsibility to Maria.

This stance-act illuminates a structural issue: Buzungu (2022: 63), for instance, has argued that civil servants experience increased responsibility for implementing the goals of their institution, yet decreased power to make decisions based on the unique situations that they identify – a gap that here seems to encourage Anne to decrease her own responsible stance. Anne goes on,

Extract 3

Another caseworker participating in my study, Ingrid, constructed a similar stance, explaining that clients in need of interpreting “må egentlig si det selv” [kind of have to say so themselves]. It is likely that Ingrid and Anne, like many social workers at Norway’s welfare office (Malmberg-Heimonen et al. 2019), are unfamiliar with their institution’s stipulations and guidelines – that the office is responsible for assessing the need for and booking of interpreters – or that they lack the means (i.e., concrete guidelines) that are necessary (Buzungu 2022: 57) to implement the administrative goals of their employer. In any case, Anne’s stance demonstrates how her justification for assigning responsibility to Maria is underpinned by competing interests at the instituion, as well as her limited power as a caseworker.

Other solutions, Anne continues, would be for Maria to propose: “kan vi ta det en gang til med mannen min?” [‘May we go through it again with my husband?]. Anne concludes: “Altså, sånne ting, da. Det må- det må bare fungere sånn at de må tørre å si det. Vanskelig å følge opp” [Like, those kinds of things. It must- it just has to work that way that they have got to dare to let me know. Difficult to follow up]. “Street-level bureaucrats find themselves under significant pressure” (Holzinger 2020: 1804) to manage conflicting demands. While social workers hold “power to control certain resources and accept or reject clients’ own definitions of their needs” (Ylvisaker and Rugkåsa 2021: 9), Anne feels constrained in her attempt to manage frictions between limited time and resources, and her responsibility to make sure that Maria receives sufficient information. Interestingly, Anne’s stance links Maria’s ability to express her needs to a psychological disposition (courage), rather than to structural conditions – i.e., Maria’s limited power to adapt their communication to her preferences – as well as doxic linguistic ideals, influencing Maria’s positioning acts vis-à-vis Anne.

In my interview with Anne, she reflects that she does not necessarily know if messages received from Maria’s account are written by Maria herself or by her husband. Furthermore, she points out that responses from Maria’s account are usually quite minimal (“yes” or “no”), and thus rather ambiguous. All in all, Anne has limited insight into how much Maria understands when she – or her husband – replies to her messages online from home. By accepting Maria’s brokering strategy, however, Anne solves conflicts of interest at her own workplace – between limited time, administrative guidelines, and her responsibility to ensure that Maria has received information and will be able to submit status forms digitally from home.

Similarly, Frida reasons, “Jeg tenker at det for [Sudas] del så virker det som at hun har da grei forståelse, pluss at hun har en ektefelle som snakker norsk” [I think that, for [Suda], it seems that she does have an okay understanding and additionally, she has a spouse who speaks Norwegian]. Having evaluated Suda’s Norwegian literacy, rather ambiguously, as “okay”, Frida goes on to highlight that Suda can bridge gaps in digital and linguistic abilities with her social capital – her husband. Frida continues, “og det betyr jo- tror det betyr mye, for da er jo, det også en person som kanskje kan følge med hvis det er- hvis hun har misforstått noe så kan hun spørre, ikke sant?” [And that means- think it means a lot, because then there is also a person who can pay attention, if there is- if she has misunderstood anything then she can ask, right?], thus acknowledging the digital labor (Bønnhoff 2020) that Suda is expected to do from home. She also recognizes the importance of linguistic capital: “Så er nok vanskeligere for de som ikke har norske ektefeller, ja, ikke sant, og begge to er- ikke er opprinnelig norske” [So, it is probably more difficult for those who do not have Norwegian spouses, yeah, right, if both are not originally from Norway]. In concluding, both caseworkers bring up Maria and Suda’s social capital as part of their own stances to “follow the norms and procedures that set requirements for and influence social work” (Ylvisaker and Rugkåsa 2021: 8). The use of brokering overrides potential concerns for limited linguistic or digital literacies in both cases, and the caseworkers are unaware that Suda’s and Maria’s brokering practices in fact play out quite differently due to their different personal circumstances. The caseworkers’ stance claims for decreased responsibility sit uncomfortably with their epistemic status and rights to the knowledge territory of the institution but enable them to solve conflicts as employees.

6.5 Claiming responsibility

For the past year, Maria and Suda have primarily communicated with their caseworkers via the website. Like many other people struggling with long-term and chronic illness, the women need to communicate with numerous contact points within Norway’s healthcare and welfare system – medical specialists, hospitals, their general practitioner, the welfare office, and so on. Strategies draw on multiple forms of literacy, including bureaucratic, administrative, linguistic, digital, and interactional. Maria uses her smart phone – and different social media applications – many times a day to communicate with her family in Norway and abroad. Yet she is less comfortable using the welfare office’s website, something that she underscores several times throughout our conversations. She explains that “jeg er redd å for eksempel der er spørsmål, og jeg må svare, ikke sant, så jeg er redd å answer for dem, hvis jeg ikke, den veldig viktige spørsmål” [I am afraid, for instance that there are questions, and I have to answer them, right, if I don’t- that really important questions]. Instead, Maria prefers to ask her husband to read and reply to messages on her behalf. When we sit down together to look at her profile, Maria, noticing multiple unread messages in her inbox, gasps loudly. When I ask Maria about her familiarity with the website, she confirms my impression, explaining that her husband “ser for meg, også sier han ‘ikke så viktig, det er bare …’, han lese litt også sånn” [looks at it on my behalf, and then he says, ‘not so important, it is just …’, he reads a bit and stuff like that]. While Maria’s husband is instrumental to her navigation of the system, he is not always available. They have different work schedules, and he often spends the whole workweek in the office to save travel expenses.

Although Suda consistently decreases her epistemic stance concerning her digital literacy in our interviews, my observations indicate something else; she signs into her profile without my assistance and finds messages, reports, and letters from her caseworker on her own. Suda reports that she usually reads messages from and writes messages to NAV without help from her husband. Only after she has written a message, she explains, she might ask him for help. She relies more on her husband for organizational purposes, however: He keeps track of letters, dates, and deadlines on her behalf. Unlike Maria’s husband, Suda’s husband is often available. His office is geographically close to their home, and working a white-collar job, he often enjoys flexible workhours. In other words, while Suda requires less brokering assistance than Maria, her husband is more available to offer brokering than Maria’s husband.

But Suda also ran into obstacles when using the welfare office’s website. With time, Suda’s health made a turn for the worse, and when she was forced to extend her leave from work while receiving treatment, she was also required to renew her application for welfare benefits. This was an emotionally taxing period for Suda; her doctor had told her that there was a significant chance that she was facing chronic disability, and Suda eventually decided to resign from her job. While this was happening, the pandemic broke out, and Norway faced several periods of semi-lockdown. As many people were now furloughed, lost their jobs, fell ill, or could not leave their house because they were considered at particular risk, the welfare agency saw a stark rise in casework. Non-urgent cases were deprioritized – which may have affected Suda’s communication with the office. All contact was digital or phone mediated. For a total of four months, Suda’s social support ceased. When trying to figure out why, Suda struggled to get through to the welfare office, both on the phone and via the website. At present, her husband had other responsibilities, and was not as available to broker for her as he had previously been.

As Suda understood the situation, her previous rights to welfare benefits had discontinued because the administration could no longer find her residence permit on record, which rendered her unqualified for the services she had applied for. Because she did not get through to the office on the phone, Suda ended up applying for the wrong benefit – one that she was not eligible for. The sum of increased digitized bureaucracy and limited access to her husband’s brokering may have compromised Suda’s access to services, adding to her vulnerability in encounters with the welfare system. In the end, Suda spoke to Frida on the phone, and the issue was sorted out. Frida could not promise Suda that she would receive backpay for the previous four months that she had gone without receiving support. When Suda and I talk this story through, she claims responsibility for this misunderstanding, highlighting that “hvis jeg ikke får [tilbakebetalt] så jeg- det er sånn forståelse for at jeg gjorde feil. Det er min skyld at jeg gjorde feil” [if I don’t get [backpay], then I- I understand that I made a mistake, it is my fault that I made a mistake]. Suda crafts a stance that positions her as in control and responsible. This stance does not match her previous storyline – which involved significant confusion between several parties – but it enables Suda to manage conflicts between her limited power vis-à-vis NAV versus the authority and control she desires to convey to me.