Child Language Brokering in Healthcare: Exploring the Intersection of Power and Age in Mediation Practices

This paper aims to explore young people’s perspectives of a real-life scenario of child language brokering in a healthcare setting (the doctor’s office), when the topic of discussion is sensitive and potentially conflictual. Child Language brokers are migrant young people who translate and interpret for family members, peers and the local community. Often the spaces in which children broker (e.g., healthcare, banks), referred to here as a ‘contact zone’, are dominated by adults in positions of authority and unequal power differentials. The language broker and those for whom they are brokering may be in a less powerful position because of their migration status and/or age status. Existing research has focused mainly from the view of adults and young people’s perspectives on the practice are underexplored. We draw the existing literature to explore how brokers understand the wider societal context and the strategies they employ to manage conflict. Findings are presented from 29 individual qualitative vignette-based interviews with language brokers (aged 13–16) in the United Kingdom which were qualitatively analysed. Findings show how these children play a vital role in protecting those for whom they broker, often navigating sophisticated social interactions and tactics (such as delay and selective modification). Equally, they carry a weight of responsibility trying to manage complicated, perhaps morally questionable, situations. By asking brokers to reflect on a real-life healthcare scenario, we are advancing understanding of migrant youth brokers and the families they support in their day to day lives. Child language brokers find themselves brokering in health care settings with unequal power relations. Vignette methodology is a useful way of capturing complex sensitive and conflictual accounts of healthcare brokering in children. Brokering in the doctor’s office required children to have knowledge of tri-interactional actors, institutional factors and wider societal norms. Young people used different strategies for managing conflict (e.g., delay tactics and selective modification). Clearer guidelines are needed around child language brokering in medical settings and spaces with contentious power inequalities. Child language brokers find themselves brokering in health care settings with unequal power relations. Vignette methodology is a useful way of capturing complex sensitive and conflictual accounts of healthcare brokering in children. Brokering in the doctor’s office required children to have knowledge of tri-interactional actors, institutional factors and wider societal norms. Young people used different strategies for managing conflict (e.g., delay tactics and selective modification). Clearer guidelines are needed around child language brokering in medical settings and spaces with contentious power inequalities.

• Vignette methodology is a useful way of capturing complex sensitive and conflictual accounts of healthcare brokering in children.
• Brokering in the doctor's office required children to have knowledge of tri-interactional actors, institutional factors and wider societal norms.
• Young people used different strategies for managing conflict (e.g., delay tactics and selective modification).
• Clearer guidelines are needed around child language brokering in medical settings and spaces with contentious power inequalities.
This paper aims to explore young people's perspectives on child language brokering in a healthcare setting, namely a General Practitioner's office, when the topic of discussion is sensitive and potentially conflictual. In the UK, a general practitioner is a family doctor. In other parts of the world, the terms family physician, primary-care physicians or general internists are used. Child language brokers are children and young people who following migration, act as translators and mediators for their parents, peers and members of the local community (Tse, 1996;Orellana, 2009;Weisskirch, 2007). Empirical evidence from previous research suggests that in many countries it is considered unacceptable to use young people as translators in health care settings, a view endorsed in the small number of studies conducted with healthcare professionals (see Abreu & O'Dell, 2017;Cohen et al., 1999;Giordano, 2007). Yet the evidence also shows that both young people and healthcare professionals confirm that language brokering frequently occurs in doctors' offices, clinics and hospitals (Katz, 2014;García-Sánchez, 2018). Young people's personal perspectives on the practice are underexplored because its 'illegality' suggests it is an activity that remains invisible (Antonini, 2010).
Language brokering has been described as a triinteractional process of information exchange (Hall, 2004) because the three-way communication often involves the child language broker, the family member (or member of the local community), and the other adult, who is often a professional in a position of power. These young people may act as cultural mediators between the host culture and their home and community cultures. They develop skills to navigate and understand the cultural values and social norms that can help steer the interests of themselves and their families in such settings (Hall, 2004;Weisskirch, 2007). Importantly, this tri-interactional process (Hall, 2004) often does not occur on a level playing field in terms of power dynamics. Language brokers can experience a double vulnerability: as young people they are subject to adult power and control; as migrants, they can be on the receiving end of exclusion and discrimination (Massaroni, 2020;Reynolds & Orellana, 2009). Additionally, the practice of child language brokering has been flagged as a cause for concern because it gives the young person "too much" power within the family (see Crafter & Iqbal 2021). This is particularly the case when it comes to matters of health.
It has been suggested that during language brokering in healthcare settings there are three types of consultation: straightforward, complex and sensitive consultations (Cohen et al., 1999;Abreu & O'Dell, 2017). In straightforward consultations for routine illnesses, young people described translating as unproblematic (Free et al., 2003;Katz, 2014). Complex consultations, usually relate to the diagnosis of conditions and brings up challenges with vocabulary and cultural expression. Sensitive consultations relate to talk about intimate conditions of the body, such as reproductive health and personal emotional difficulties. This paper broadens the 'sensitive type' of consultation by exploring a morally ambiguous and potentially conflictual dilemma faced by a language broker during a general practitioner consultation. Using qualitative vignette interviews with twenty-nine young language brokers (aged 13-16 years old) in the United Kingdom (UK), this paper uses the theoretical framework of the contact zone to explore the language brokers responses to a sensitive and conflictual situation set in the doctor's office. The conflict situation involved three actors: a doctor (professional), an adult who is a neighbour to the young language broker and the young language broker. The vignette is based on a real incident of brokering.
Child Language Brokering in Healthcare: The Intersection of Power and Age A child language broker is defined as children or young people "who interpret and translate between culturally and linguistically different people and mediate interactions in a variety of situations" (Tse, 1996, p. 226). Children as young as 3 years old have been found to language broker or engage in a form of natural translation for their families (Harris, 1977). The concept of language brokering points to someone who both culturally and linguistically brokers between family, peers and members of the local community and others within the host or local culture (Kam & Lazarevic, 2014;Jones & Trickett, 2005). The term recognises that the language broker may take the initiative to negotiate or change the message to benefit their own or their family's situation.
With variable access to professional translating services in increasingly diverse language contexts in many countries, child language brokers are often the key linguistic and cultural mediators for migrating families between home and professional institutions (Crafter & Iqbal, 2020;Reynolds & Orellana, 2009). They facilitate the social and cultural integration of their family in the new host setting (Foner & Dreby, 2011) across a wide range of spaces including banks, retail, healthcare, law, social care, government offices and police; via translation of texts (e.g., emails, legal documents, school letters) and through interpreting conversations (Dorner et al., 2007;Tse & McQuillan, 1995;Valdés, 2003). The range of activities and contexts in which young people language broker means that their experiences can range from everyday or routine exchanges, to more complex or conflictual situations (Iqbal, 2019). Research on the social and psychological impacts of child language brokering present a complex and mixed picture ranging from associations with stress and anxiety to feelings of pride and accomplishment (Chao, 2006;Kam & Lazarevic, 2014). The physical and cultural setting in which language brokering occurs, the familial and relational influences and the type of task undertaken, and the goals of those involved, are all variables that can play a mediational role in the young person's experience (Kam & Lazarevic, 2014).
Language brokering in contexts like the General Practitioners office or other healthcare settings has understandably raised concerns from professionals working in frontline services (Gustafsson et al., 2019;Cohen et al., 1999). In situations where trained adult translators are unavailable, translating in medical settings can be a common occurrence for many young language brokers. This can require technical and systems knowledge, as well as interaction with a professional (Gustafsson et al., 2019;Martinez, 2019;Katz., 2014). The majority of research in healthcare settings is from the perspective of the adults, with limited research looking directly of children and young people's perspectives (Martinez, 2019). For example, Cohen et al. (1999) explored the views of doctors in a UK context of child language brokers when they translate for adult patients (usually family members) during consultation. They argue that while doctors accept children as interpreters due to the lack of adult community interpreters available, this acceptance is contingent on the nature of the consultation; whether it is straightforward, complicated or sensitive. The hesitancy and opposition expressed by doctors, was primarily associated with their traditional view of children as being innocent, non-agentic and their need to experience a care-free childhood (Cohen et al., 1999) but also of children's not having the knowledge to transmit medical concepts. This was reflected in interviews with healthcare providers in Los Angeles, USA, who worried about the transgression of age-appropriate boundaries, even when they could see some advantages to using child language brokers (Katz, 2014). They felt the quality of care was compromised and that interactions were made more difficult. Language competency and the crossing of ageappropriate boundaries were also stated as concerns. Furthermore, differences in opinions on treatment can exist between the patient (for whom the child is translating) and the medical professional, with the child language broker left to mediate and navigate such tensions (García-Sánchez, 2014). However, not all research has found such opposition. In one study in Italy, a manager of the health authority presented a contrasting opinion, suggesting the child language brokers should not be completely disregarded (Antonini, 2010).
Research that includes the perspectives of children and young people on language brokering in health care settings generally conclude that brokering here is more challenging than in settings like school because of uncertainty about unfamiliar procedures, complex language and the range of issues that might need addressing from emergency to nonemergency situations (Katz, 2014). However, in both Katz's work and our own (Crafter & Iqbal, 2021) we found that children did not always act alone and often strategized with parents about what would be discussed and how it would be discussed. García-Sánchez (2014) in her work on Moroccan immigrant children at the doctor's office in Spain describes how young people used strategies of 'selective modifications' (p. 232). García-Sánchez differentiates between instances when children had problems understanding information and thus withheld it unintentionally, and instances when children purposely modified information. She writes (p. 232): 'Moroccan immigrant children strategically deploy these modifications in their translations when they are confronted with conflicting moral universes, beliefs, and practices concerning child-rearing and health issues'. Some of the reasons for these modifications chime with Bauer's (2013) ideas around moral identities. These refer to children's management and creation of particular positive identities of those they broker on behalf of, through the careful transmission of their message. Others were related to protecting their family's interests when navigating different structures and systems, such that they represent their families in a particular light to the adult 'other'.
There is a complex interaction between the practice of child language brokering, the tangle of asymmetrical power relationships within the brokering context and understandable concerns about the age of the child language broker. Some have argued that conflicting demands of the adults in the tri-interactional process can leave the young person caught 'in the middle' (Hall, 2004). This means that despite children being in a position of power because they can affect or change the message, the terrain can be hard to navigate, resulting in their sometimes struggling with this power. Power relationships are fluid in this context because unlike the situation of professional translators the young person has close links with one of the adults leading to concerns that adults appear as 'de-skilled' (Hall, 2004, p. 12); unaware if their child is changing the content of the message during translation (Hall & Sham, 2007;Wagner, 1993) or whether young people are advocating too vigorously for them (Katz, 2014). Children have a different positioning compared to adult interpreters, because they live with and receive care from those for whom they perform these tasks, which mean that some professionals, even when viewing young people's roles positively, can frame parents as being passively protected by their children (Katz, 2014).
Previous scholars have used expressions such as 'parentification' and 'adultification' to describe a shifting of parent child dynamics because of the increased responsibilities taken-on by brokers within the family (Vasquez et al., 1994;Weisskirch, 2007). We argue in another paper (Crafter & Iqbal, 2021) that such terms suggest a very static power relationship between parents and children. Other scholars have argued language brokering can be framed as 'family interpreting' emphasising families as a performance team (Valdés, 2003) or engaged in acts of caring that enables families to function in sometimes challenging immigration environments (García-Sánchez, 2018). This has political ramifications in that these children facilitate families' access to health care, technology financial services and resources (which newly arrived immigrant families can be often be excluded from). It can also mean that migrant children who are language brokers may find themselves directly in the path of everyday racism and microaggressions (Nash, 2017) in public spaces in the host society such as the healthcare settings (Reynolds & Orellana, 2009).

Theoretical Framework: The Contact Zone
The focus of our paper is on young language brokers perspectives on a morally ambiguous and conflictual situation that occurs in the healthcare setting imbued with inequalities relating to power and age. Conceptually, this paper draws from Mary Pratt's (1991) work on the contact zone which is a 'term to refer to social spaces where cultures meet, clash and grapple with each other, often in contexts of highly asymmetrical relations of power, such as colonialism, slavery, or their aftermaths as they are lived out in many parts of the world today' (p. 33). The idea was initially used with reference to theories of literacy, yet it has subsequently been used across disciplines including psychology and education (Crafter & Iqbal, 2019;Hermans, 2001;Malsbary, 2014). It links particularly well with child language brokering, which as a practice is a manifestation of globalisation and migration that can place families and communities in a marginalised position due to lack of linguistic and cultural knowledge of the host society.
Language lies at the heart of the contact zone and can be seen as social code creating a shared sense of community and patrimony within a homogenous and unified social world. Yet, this assumption is an exclusionary one when speakers from different social classes and cultural backgrounds come into different contact zone settings such as classrooms, health and bureaucratic spaces. Dierkheimer and Helbrect (2015) argue when thinking about conflictual encounters between strangers, a struggle for prestige status and situational dominance can occur between social actors. However, when one of the actors within a social exchange is a young person, the status and dominance they are able to bring to an encounter can be different from that possessed by a professional adult (such as a doctor). Indeed, migrant brokers and their families may lack the social codes and sets of norms/rules needed for shared communication (Pratt, 1991). This can then result in one party (in this case the professional in a health care setting) exerting authority over the other.
The contact zone is a space which can be a dynamic one of confrontation and conflict-(an unsafe space which constitutes unequal power relations) but also one that can be comforting and conforming (what Pratt calls a safe house, which constitutes more equal power relations) (Atari, 2013). Importantly, the relation between safe/unsafe spaces is not fixed but fluid and shifting, at times ambiguous, illustrating that the power play within the contact zone is influenced by various social categories such as language, class, ethnicity, and religion. Pratt argues that social actors exercise power differently in different spaces: they might be dominant in one space but dominating in another. It is important to bear this in mind when thinking about encounters in the contact zone. We draw from these theoretical ideas in our consideration of how youth brokers deal with conflict in a doctor's office.

Research Design and Methods
The aim of the wider research project on which this paper is based, was to explore how young people who regularly act as child language brokers mediate cultural knowledge and identity across different spaces and contexts. Data was collected using a combination of traditional qualitative and quantitative social science methods (vignette interviews, observations, survey) and arts-based approaches (drama, podcast and art workshops). All data was collected following ethical approval for the study by the Research Ethics Committee at which the lead author is based, and all data was collected by the two authors of the paper. Both researchers were middle class women, one British South Asian and the other White British. In this paper we focus on the 29 vignette interviews (23 female, 6 male) with young people aged between 13-16 years old from three schools in Greater London in the United Kingdom. Separate consent was obtained from the young people and the carers of the young people. The consent form and information sheet were written in simple jargon free language, and we went through these forms with the young people. They took them home to their caregivers to explain and sign. The school also facilitated in this process and the parents had a chance to ask questions with the researchers and teachers before signing. We gave clear information about the project, ensured parents understood their children's participation was voluntary and that they had the right to withdraw. Full details of the method we used and further details of the sample can be found here: https://la nguagebrokeringidentities.files.wordpress.com/2019/03/a hrc-final-report-clb-as-mediators.pdf.

Research Setting and Participants
In Greater London, almost 40% of the population are born outside the UK and over 300 languages spoken throughout the city (Rogers, 2013). We were invited to collect our data in three secondary schools with young people aged 13-16 years old, namely-Murry Green (N = 9, all female), Rutland (N = 9, 6 female, 3 male) and Drake (N = 11, 9 female, 2 male) (The names of all the schools have been changed).
Each were carefully chosen based on a number of points, including: their localities (they were all in areas with high levels of recent migrant arrivals); high levels of students with English as an Additional Language (EAL) which is an important signifier for the practice of child language brokering; school links with a local Young Interpreters organisation; and a past research associations in one of the schools. The young people in the schools completed an online survey which helped us establish which young people were active brokers. Those who brokered more than once a week for family members were invited to participate in the interviews and workshops.
Our participants have varied migration backgrounds. Some young people and their families made direct journeys straight to England from their home countries including Poland, Bulgaria, Colombia and Ecuador. Others made transnational migratory journeys, in which they or their parents were born in countries like Bangladesh, Sri Lanka or Nigeria and came via another country such as Italy or Spain before setting in England. Nineteen of our sample arrived in the UK between the ages of 11-16 years, 5 described arriving between the ages of 6-9 years, 1 told us they arrived between 2-3 years and 4 of our sample did not give us this information. We did not tease out in our analysis how age of arrival for respondents shaped how they interpreted the moral ambiguity of the vignette scenario (presented in the next section). However, we did carefully select our sample. Before we conducted the interviews and administered our survey, we spent time getting to know the young people and conducting a series of workshops around spaces, places and situations in which they brokered. All the young people selected were active brokers for their families and had experience and knowledge of brokering in a range of different systems and settings in the UK (including the doctor's office). All of the young people interviewed were able to be interviewed in English. Our excerpts have retained any grammatical errors and local London accents.

Vignette Interviewing
We collected data using qualitative vignette interviews. A vignette provides a story stimulus that enables the respondent to reflect on the dilemmas of the vignette characters. Previously framed as an ethically sensitive approach (Barter & Renold, 2000), vignette methodology allows researchers to discuss issues from a perspective that is non-personal and thus less threatening . As our participants did not have English as their first language, it was also felt that the vignette would help them communicate about their own perspectives and experiences during the interview.
All of our vignettes were developed from examples in the language brokering literature. We created each vignette such that the respondent had a difficult or conflictual situation to negotiate as well as potential power struggles. In other words, we created a contact zone (Pratt, 1991). We share findings from one of four vignettes that were presented to young people during the interview (for details of the other three see Crafter & Iqbal, 2020). The vignette story discussed in this paper is based on a visit to the doctors and was adapted from a real-life healthcare brokering situation described by García-Sánchez (2014) in her book about Moroccan immigrants in Spain. The scenario involves a young child language broker-Sorraya (the character), interpreting for her neighbour at the doctor's office. The research team read out the following vignette to their respondents: Sorraya is translating for her neighbour at the doctors. The woman's little son is not eating and this is the second visit in four months. After a little discussion the doctor says that the mother has to stop giving him so many sweets. The mother cheerfully asks Sorraya to tell the doctor that when he starts screaming, she gives him a smack. Sorraya knows that the doctor will not approve of hitting her child.
Each respondent was then asked the following questions: (1) What do you think is happening in this situation?
(2) How would you describe the cultural misunderstanding that has occurred? (3) How do you think the child in the story would deal with it? (4) Role play exercise: Imagine you are in this situation. I am the doctor and you are Sorraya. You have just explained to your neighbour that she should stop giving her son sweets and the neighbour has responded with the comment about smacking. Interviewer says in a professional tone-'Can you explain to me what your neighbour has said?'. What would you do/say?
The conflict in the story arises when the young person (Sorraya) hears that the neighbour (for whom she is translating) presents her with unexpected information-that she smacks her child. Thus, the vignette character and our interviewees are presented with moral conflict that navigates cultural norms and values. The first three questions ask our interviewees to comment on what they think the character would do in the situation she finds herself in, while question 4 asks the young people to imagine themselves as the character. This question aimed to facilitate our interviewees in bringing in their own personal accounts of similar experiences in medical settings. Of the 29 in depth interviews with the young people, 26 responded to this particular vignette in detail. This was the final vignette in the interview, and time constraints had led to three interviews ending abruptly. This is one of the limitations of working in schools.

The Analytic Process
The analysis of vignette interviews was conducted through a combination of hand coding and qualitative data analysis software (NVivo). We read through transcribed interview transcripts and developed initial thematic categories based on existing theory, our research questions and literature on child language brokering. This initial code book was developed from team discussions focusing on an initial sample of transcripts. These were then refined and challenged through further engagement and analysis of the data and grouped into themes (Flick, 2014). The analysis team comprised of three people (the two authors and a colleague, external to the project) who independently analysed transcripts and then came together to compare themes. Where there was disagreement, themes were abandoned or reworked until a consensus was reached. When distinctive codes were introduced by team members, these were compared with existing codes for similarities and either integrated, added to the coding list, or let go of. Once finalised, the framework was used to analyse remaining transcripts with team members continuing to confer where queries arose. During this analysis process, we focused on the dyadic relationship between the character in the vignette and the young person's own positions during the interview. When we discuss the vignette findings, we are clear that while these can often be projections of the children's decision making on the character, here we use them as a means to explore multiple positions and power dynamics . In this case, the vignette linked to a scenario relating to the character (Sorraya) encountering a sensitive consultation involving a moral dilemma in the doctor's office while brokering for her neighbour. We were interested not only in how our participants reported Sorraya would handle the situation, but also in the accounts of their own similar experiences in medical settings that participants brought forward. Our coding strategy incorporated this distinction. In the paper we will clarify when respondents have drawn from their own experiences and when they are reflecting on the vignette character.

Findings
All of the brokers had had real-life experience of brokering in a medical context, such as a doctor's office, a hospital or a pharmacy. None had faced a similar moral dilemma or situation to Sorraya (the vignette character) but our respondents were able to reflect on the complexities raised by the vignette. Within the scenario, there are three key power-relationships associated with brokering in a medical context: (1) the adult neighbour and their relationship with the child broker, (2) the doctor (who acts as a knowledgeable authority figure) and their relationship with the broker and lastly (3) the broker's understanding of the wider possible consequences for the neighbour of revealing that they smack their child. These power relations map on to different elements in the contact zone, such as the tri-interactional relationship between the neighbour, the doctor and Sorraya and wider national policy and cultural norms around child welfare and protection.

Understanding of the Problem and Institutional Power Dynamics in the Contact Zone
All of our young respondents were conscious that language brokering took place within complex social situations that required knowledge of both the tri-interactional actors, institutional factors and wider societal norms. Participants reflected on Sorraya's (the vignette character) dilemma, and of the repercussions of her translating to the doctor what her neighbour told her. Tereza, age 16, discusses this: I think Sorraya, … knows that if she tells the doctor that the neighbour hits the little kid this will get to social workers and then they will just try to take the child from her… the social workers will come to your house and they will try to investigate what's happening. … so, I think she's a bit concerned like, what words shall I use, should I tell him that, or should I just explain it in the different way.
Many of the other participants reflected on the potential problems faced by the neighbour in the vignette, particularly as they understood the doctor's office to be a point of linkage between wider social services and the family. They recognised that the state had the right to take the child into protective care if the child's welfare was assessed to be at risk. Although not all children were familiar with the terminology, all understood worst case implications, such as Anca, age 13, 'you know how it is there [in the UK], if you beat your child the people come to take your child away'.
When the young people reflected on the wider sociocultural context and a contentious parenting practice, some of our respondents compared societal norms between premigration countries and their new life in England. Samir, age 14 from Afghanistan, reflected on the vignette story and how his own personal experience of being disciplined by a parent: Cause in different countries you have different thoughts about stuff and I think if you do the same thing in my country the doctor will approve it.
In previous research, it has been argued that the practice of language brokering enhanced the young person's understanding of both the home culture and the host culture (Birman & Trickett, 2001). Samir's quote suggests that he understands smacking to be an acceptable form of discipline by the neighbour in one context, but also that Sorraya's indepth knowledge of the host culture's cultural norms place her in a more knowledgeable position than the neighbour. Tereza, age 16, raises a similar point, positioning the neighbour as being in a less powerful position because of her lack of understanding about social and welfare norms: The mum must be from another country. For example, if you're African, if you smack your child, you're just educating them, its ok. But like, in British [sic], I think they take it too much, to a point that you cannot put your hands on your kid, end… I think the mother does not understand that if she says that to a British doctor you get it another way and the mother will get in trouble Outside of her perspective on parenting practices in a British context, Tereza recognises what is needed to navigate the health system. Within this contact zone of the medical setting, Tereza demonstrates she is able to draw from her acquired social codes to facilitate an 'orderly, coherent exchange' (Pratt, 1991, p. 38) and thus avoid a situation that would cause trouble for the mother.
For our language brokering respondents, the moral dilemma presented in the vignette also presented emotional challenges for Sorraya-the vignette character. Ellora, aged 14, talks through the range of different complex dimensions that Sorraya might consider: Cause I know that if a child gets abused and slapped and um, beaten out. There is like legal consequences. I think Sorraya knows it, so she doesn't want the police [to] separate the mother from the child so maybe Sorraya is thinking about the child. She doesn't want the child to live without a mother, cause a mother is something important … So she's thinking about this and she's like, I need to find a solution to make them get away. So it's something really difficult. If I was in Sorraya's place I would be, freaking out.
The first issue of note is Ellora's escalation of the word 'smack' to something more severe (beating and abuse). This escalation explains her link to police intervention, that in turn leads to a reflection on the implications the importance of the mother-child relationship. The language broker character carries the moral burden in this scenario of needing to think about the loss of a mother for a young child. The idea of a moral dilemma running as a thread through some of our participant's responses to the vignette ties into Bauer's (2013) work on the creation and management of morality in the process of language brokering. The responsibility of making consequential and moral decisions positioned the vignette character as being in a position of being both powerful and powerless at the same time. On the one hand, Sorraya (the character) has a leading position in directing the conversation. Equally, Ellora describes how if she were Sorraya, she would be 'freaking out'. Some of our other respondents described Sorraya as 'feeling stuck' to feeling 'sad' and 'uncomfortable' should she have to lie to the doctor. Anca, aged 13 whose first language is Romanian, described Sorraya as being a 'martyr' but used the word 'vinovat', which in Romanian captures emotion also linked to guilt and culpability. In doing so, Anca was able to describe the weight of responsibility she perceived Sorraya to be carrying.

Strategies Employed by Young People for Managing Moral and Cultural Dilemmas
Many participants discussed the use of delaying tactics within their own real-life translating situations, particularly where they were translating a lot of information, or where the information was complex and there was a moral dilemma. In the same way, they suggested the vignette character might use such tactics. This they believed, would give Sorraya time to think about the best way to respond to what the neighbour said. Rabiatou, aged 16, for example suggests, 'If I was Sorraya, I would tell the doctor we were having a conversation about (the issue) and I didn't understand what she meant (when) she was explaining it to me.' In this instance, Rabiatou states she would claim a lack of understanding of the problem described by the neighbour as a means of stalling (as the doctor did not understand her home language), thus buying her time. Pratt (1991) discusses the role of language and power in the contact zone, and it is clear that in this case, English (as the host society language) occupies a dominant position being used by the authority figure (doctor). However, through such delay tactics using the subordinate home language, the young people are able to subvert power dynamics within the contact zone. Indeed, Daria, aged 14, suggests Sorraya should have a conversation with the neighbour in their home language to inform them of the possible repercussions of translating directly what they stated.
Yeah, but if they are talking in another language, Sorraya, she can tell the doctor 'Oh excuse me I need to talk about, to explain her something'. She can lie to him just to explain to her that it's not a good thing to tell the doctor that you will hit your child if he will scream at you or shout or something like that.
On the one hand, Sorraya's power position within the contact zone of the doctors' office is lower because of her status as a child. On the other hand, Daria describes how the character could use her language brokering as a strategy to give her time to explain the conflictual problem to the neighbour. In this sense, she exercises some degree of power as a language broker to form a closer allegiance with the mother (neighbour) rather than the figure of authority, in order to impress upon the neighbour, the seriousness of the situation. In doing so, she would be protecting the mother from the more powerful figure in the interaction-in this case, the doctor. Recognising the gravity of the brokering situation, some children, such as Rabiatou, aged 16, said they would be very direct with the neighbour as a lot was at stake: I would have told the neighbour that you're not allowed to hit children in the UK. That's what I would have said. Because they can take your child away, in a heartbeat.
Although we do not claim that the young people's responses would actually reflect what they would do in real life, their responses are interesting none-the-less. Of the 26 detailed responses to this scenario, 22 young people said they would withhold or selectively modify the information about the neighbour smacking her child. Three young people were unclear about what decision they would take in real life and only one child claimed that they would inform the doctor that the mother smacked her child. Reflecting similar findings regarding 'selective modification' reported by García-Sánchez (2010, there were different reasons for doing this including protecting the interests of the family, recognising a moral dilemma, wanting to avoid conflict and wanting to present their home culture in a positive light. In our study, Tereza, age 16, discusses how she would change the content of the message. I don't think she would tell the doctor that the neighbour just smacks the child. I think it would say it in another [sic] words. She would be like 'I don't like it when he cries a lot and I don't want to smack him. However, in this case, Tereza differentiates herself from the character in the story. I don't want to do this. Because even though you tell the doctor, I think, from my point of view, if you tell the doctor 'oh no, this is just a tradition' (they would say) 'you're in this country, you're in Britain, you don't do that'. Because then if the kid appears bruised somewhere it's gonna be the doctor who's gonna see it There are two important features to Tereza distancing her own position from that of the vignette character. Firstly, she understands the cultural knowledge of 'tradition' does not have the same weight as dominant understandings of childhood and child protection reflected in the stance of the doctor authority figure. Secondly, previous research has shown that language brokers are very conscious of adults being suspicious of whether they tell the truth, and consequently talk about the importance of being truthful to adults (Crafter et al., 2017).
The power of linguistic terms was brought up by participants in relation to cultural context. For example, multiple children brought up the word 'smack' and said they would change this to soften the message to the doctor for example 'spank'. Ania, age 13, uses the word 'tap' or 'lightly hit' alongside an explanation about parenting differences: I would just say that when he starts screaming … like she gives him a tap or you know lightly hits him because that's what they do in that culture and like she doesn't know that you are not supposed to do that so I would kind of like just explain to the doctor that that's what they do in that culture The way in which our participants described the child's explanation of the situation to the neighbour varied. Ania, aged 13, for example says Sorraya would explain to the mother the possibility of punishment from external sources and that smacking was not correct: To the mother I would say that, you shouldn't really be doing that because like you just don't do that here and you can get punished for hitting a child here and it's not really appropriate Selective modification was also used as a means of safeguarding or shielding the individual with less power in the brokering context; in this instance the neighbour. Dimitar, age 14, discusses how altering messages is inevitable to protect the interests of the party they are translating for: I mean she [the character -Sorraya, as a child language broker] should probably,… lie to the doctor what they said to each other because she wants her neighbour to keep her child because that would devastate her neighbour. She should probably say we were having some difficulties saying what the issue was exactly which are her eating habits.
Interviewer: So she should protect her. That's quite a big thing for a child to do, do you think?
Yeah it is, but sometimes you gotta do what you gotta do.
Dimitar's sympathies here lie with the neighbour, who is positioned as the character in most need of protection. This perspective bears resemblance to the caring responsibilities that language brokers are said to do on behalf of family members, which have a moral-affective component that help children maintain relationships (García-Sánchez, 2018) and protects those who cannot communicate in the local language from perceived injustices (Nash, 2017). Dimitar's follow up comments to the interviewer's question here 'you gotta do what you gotta do' are also particularly interesting. In other parts of his interview, Dimitar reflected on how mindful he is of his status as a child and that he recognises this status enables him to challenge expectations and confront conflict where needed. This isn't necessarily the case for all young brokers, but it was for Dimitar who understood that traditional ideas around childhood meant professional others might take him for granted and not expect him to behave in certain ways. He often used this to his advantage during the brokering process (Crafter et al., 2009).
Others, in a very different way to Dimitar, reflected on their position as a less powerful child to an adult figure, and felt they didn't have the confidence and expertise to deal with such a complex situation. As a result, they came up with alternative approaches to how the character would deal with this situation. Mihal, aged 13 states the character would ask her own parents to intervene and 'sort things out' because he believed the neighbour would not take kindly to a child giving her advice: Because it will be good if she, she will tell, like, a member of her family, because if a member of family, family is friendly with the neighbour, maybe she can talk with the neighbour and sort the things out. Maybe.
By bringing in an external family member, Mihal reflects that this would alleviate the pressure Sorraya was under and in a way minimise the tension in the contact zone. In some ways this is an example of how age can be used in another advantageous way. As a child brokers can legitimately draw on the resources of other adults to help with their dilemma.
Ellora, age 14 also thinks about how resources and knowledge can be obtained from adults. She decides that Sorraya would be an active broker (Crafter & Iqbal, 2020) and take the initiative to gather information about good parenting techniques for the neighbour, 'So maybe ask advice of the doctor of how to behave in that situation'. In this way, she manages to create a safe space within a contact zone but while also selectively modifying the information.

Conclusion, Implications and Future Research
Conclusion This paper set out to explore how child language brokers talk about a real-life brokering scenario within a healthcare setting (the doctor's office), wherein unequal power relations exist between actors. Each of our brokers had experience of brokering in a medical context but none had faced the kind of moral and conflictual dilemma that we presented through our vignette character Sorraya. Our participants responses are informative because we were able to get a sense of how young people reflect on the responsibility of holding power in institutional spaces and of acting as a voice for an adult.
Two main findings emerged around young people's understanding of the problem. The first was, related to the moral dilemma and power inequalities present in the brokering scenario. The young people in our study, reflected on the array of institutional factors, societal norms and triinteractional actors involved in the healthcare scenario. They recognised that the responsibility of language brokering (in relation to the character), was one that could leave the young person feeling simultaneously powerful (with the brokering duty they had been bestowed with) and powerless (when the outcome of their brokering could have serious implications for others). The second finding was associated with the strategies employed by young people for managing complex brokering situations. These strategies included delay tactics, especially when there was a lot of information and it was complex or controversial. Also, withholding information or selective modification of the message. Through the use of these strategies, the language brokers were able to subvert power dynamics in the contact zone, and assert a degree of self-agency. Another important consideration for the paper was how age played a role in a complex brokering situation.
The participants in our study seemed to recognise that that child brokers possess in-depth knowledge about other cultural values and traditions. Yet, such practices and traditions do not always bear the same weight in all contexts. Their knowledge thus extends to an understanding of both home culture and host culture practices and wider consequences surrounding these when linked to institutional forces. For example, on the surface it might seem that Sorraya (our vignette character) is in a powerful position because she can 'selectively modify' the talk. However, in a situation where there is a moral-affective and conflictual situation, Sorraya is in a delicate position. Her words carry serious repercussions and our respondents positioned doctor as the authoritative other who represents social care practice and norms of societal and cultural values in the host setting.
Our use of Mary Pratt's (1991) work on the contact zone provides a useful means for thinking about the experiences and power dimensions encountered by a child broker in the process of mediation on behalf of a family/community member. It builds on the notion of migration, marginalisation, and the superordinate/subordinate positions held by certain languages, individuals and spaces. As Atari (2013) highlights, contact zones are connected, and actions within one (in this case the doctor's office) can have wider ramifications in another (the state welfare system). The role of language within social settings is particularly important. While Pratt discusses how different languages carry different representations of power, the sociolinguist Hymes (1971) thinks about how children across contexts may not have access to equal sociolinguistic resources and language acquisition. This is particularly true for the case of child language brokers who carry different migration histories and take on roles which are considered to be adult ones. In the contact zone setting of the doctor's office (explored in this paper) for example, young brokers can be faced with linguistically complex medical terminology which can be challenging for them to interpret.
The paper also highlights that children can be forced to make big decisions which will have an impact on those close to them. Along these lines, children in this study talked about a number of strategies used to buy time, shield and protect others through changing language and understanding more about wider context. Language brokering is arguably a practice that aligns with cultural expectations of familial responsibilities (Spyrou et al., 2018) and some scholars have framed the process as a burden (Tomasi & Narchal, 2019) due to the responsibly it places on children and youth. Yet we have argued elsewhere that children themselves can view participation in such work in a positive light in many instances. It can act as a means in which they gain additional skills and development relative to their chronological age (Crafter & Iqbal, 2021).
We recognise one limitation of our study is that we purposefully developed vignette stories that contained conflictual and challenging situations, which brought up moral dilemmas. Nevertheless, each of these stories was based on a real-life incident of brokering, and this indirect method of capturing children's experiences was ethically more sensitive and less threatening than having researchers in-situ during a tense moment of translation.

Implications and Future Research
With austerity cuts made to budgets for public services (such as professional interpreting), across many parts of the world, less money is being spent on providing easy access of translators and interpreters to migrant families and individuals. It is likely that children and young people will continue to do this activity. Moreover, in the current political climate in which anti-immigrant sentiment is high in many parts of the Global North accompanied by often hostile immigration policy, issues around trust mean that parents and adults may prefer family members to support them, even if these are children.
In the UK there are few clear formalised guidelines for the use of child language brokers in public settings (for e.g., legal, medical, financial and governmental). While some guidance does exist for education settings: see https://www.nuffieldfoundation.org/project/child-langua ge-brokering-at-school), it does not adequately address the challenges which are unique to others such as healthcare (discussed in this paper). Yet, we know this practice goes on. Given this, there is a need for improved and tailored practical guidance for healthcare professionals and a recognition that there are some conversations which children should not be involved with. There should be an understanding of boundary setting and a recognition of the power differential at play between themselves as adult professionals and children who are brokering for their families. For example, there may be times during a consultation when a discussion can start out in an acceptable manner, but then turn into something the child shouldn't be involved with, in which case the healthcare provider would need to step in. Additionally, an understanding is needed of wider contextual factors that may be at playimmigration status, lack of understanding of systems.
At a policy level, more discussion and understanding are needed about the complexity of language brokering with an inclusion of young and adult language brokers in the conversation. Moreover, there is a need for more research in the area of brokering and health care. We conclude with a call for research which specifically seeks to understand and categorise the range of brokering activities in which young people engage in medical settings and other spaces with contentious power inequities. One of these areas of urgent need is in health emergency situations where children are often the only available interpreter.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.