Cultivating capabilities and coping: accepting and analysing moments of communicative opacity in multilingual encounters

1 Introduction

While current preoccupations with (super)diverse populations rest on the presupposition that such demographics are something new (Pavlenko 2016; Vertovec 2007), there is no doubt that researching in linguistically, culturally, and ethnically heterogeneous communities may present challenges for the most experienced of researchers, even when highly prepared with a linguistically representative team, attuned to situational sensitivities (Andrews et al. 2019). This is especially true in the context of ethnographic work, where the very nature of the methodology can make it difficult to anticipate the languages spoken by participants encountered in dynamic and unpredictable settings.

Building on experience gathered during linguistic ethnographic research exploring multilingual practices in antenatal health consultations, and undertaken for my doctorate, this paper reflects upon how a naïve understanding of ‘multiple cultural competence’ (Vertovec 2009) gave way to a gradual, uncomfortable recognition of my own ‘linguistic incompetence’ (Phipps 2013), when interaction extended beyond the scope of my distinctly limited repertoire. Unfamiliar lects, varieties and languages not only served to highlight my linguistic shortcomings but often restricted the immediacy of inference. In turn, discomfort with my ‘incompetence’ was frequently exacerbated by ongoing processes of analysis, when the circularity of transcription, translation, clarification and (re)transcription, complexified what could be understood as ‘interpretation’ and contributed additional layers to possible (mis)representation. On the other hand, as my ethnographic journey continued, I began to explore some affordances of linguistic non-understanding (van Hest and Jacobs 2022). With a growing appreciation of the consulting room as a liminal, ‘third’ space (Bhabha 1990) of potential unknowability, a capabilities approach (Nussbaum 2011; Phipps 2013) offered a chance to focus more keenly on what I was able to do. My attention was drawn to the linguistically strategic, corporeal, and non-verbal aspects of relationality embodied by healthcare professionals, which appeared core to enhancing women’s experience. Simultaneously, and perhaps for the first time, I began to really understand the notion of transcription as an integral dimension of data interpretation, a joint endeavour shaped with collaborators, and fraught with personal subjectivities (Bucholtz 2000, 2007]).

This paper discusses some of the emergent vulnerabilities which came to light during ethnographic fieldwork and data analysis and seeks to illustrate how moments of communicative opacity can be navigated. I begin first with an overview of the literature on researching multilingually, drawing on work which explores some of the challenges posed by linguistically unpredictable contexts, the limits of personal repertoire and the unanticipated affordances prompted by liminal spaces.

Methodological reflection on translation and transcription follows, before an introduction to the research context and participants. In sharing data extracts from the field, I will exemplify some of the complexities presented by language incongruence and multiple transcriptions, exploring the impact of unsolicited commentary and potential strategies for coping with (un)expected tensions. In documenting practical and ethical considerations, I explore my own personal limitations in the hope that these will contribute to open and ongoing conversations within the research community. I end this paper with recommendations for institutions supporting sole doctoral researchers working with multilingual participants.

2 Challenges to researching in multilingual contexts

The number and variety of people migrating from different parts of the world has increased exponentially in the 21st century, dispelling previous ideas of diverse populations as living within bounded, diasporic communities, linked by either space or knowledge and sharing cultural and communicative norms. Such conceptualisations have been replaced with understanding of contemporary urban populations as having ‘meshed’ realities (Canagarajah 2011), where a heterogeneity of ethnicity, nationality, language, education, age and gender profiles, immigration and work status has become the new norm. In response to the demographic changes that may present challenges for traditional ethnographic fieldwork, and in the anticipation that researchers may encounter non-hegemonic and/or unfamiliar repertoires, Andrews and colleagues (2019) propose a framework for researching multilingually. One recommendation is that studies should be approached with a degree of clear intentionality, with researchers giving thorough consideration as to how they will respond to the linguistically diverse populations with whom they may engage. This preparatory stage should not be underestimated, for there may be a number of issues which need prior deliberation. In their recent paper proposing a ‘methodological multilingual turn’, Costley and Reilly (2021) reflect upon the degree of planning and preparation that may be needed before embarking on fieldwork in linguistically diverse contexts and offer a prescient reminder that we cannot expect to engage with multilingualism from a monolingual perspective.

This implies a need for flexibility, resourcefulness, and heightened awareness of the local context. From the outset, one may be well advised to pay attention to the kinds of practices and language that may or may not be generated, or even tolerated, in a specific research space (Andrews et al. 2019). At the same time, the multivalence of language(s) may mean that while speaking a language loaded with historical legacies may inhibit the potential for collaborative relationships with some participants (Sepielak et al. 2023), it may nonetheless hold different associations for others. Of course, it is clearly advantageous to have a degree of proficiency in (a) shared language(s) when working with specific communities. Not only can direct communication enhance comprehension and strengthen relationships but there are likely to be additional benefits in terms of increased ethical sensitivities, authenticity, and the reduction of asymmetries (Costley and Reilly 2021; Ganassin and Holmes 2013, 2020]; Holmes et al. 2013). However, conducting research in a superdiverse environment nevertheless poses challenges in terms of practicalities, and to suggest otherwise would be disingenuous (van Hest and Jacobs 2022). While some multilingual researchers may have the resources, and occasion, to draw on different parts of their repertoire (Ganassin and Holmes 2013; Polo-Perez and Holmes 2023), others may need to reflect on personal limitations. Indeed, research by Sepielak et al. (2023) finds that linguistic concordance can frequently be a scholarly aspiration rather than an everyday reality. Communicative ambitions can remain unfulfilled if a language is not spoken fluently by a researcher and if a variety is considered rare or difficult, there may be limited inclination, time, or budget to learn it.

Pre-empting such scenarios, Costley and Reilly (2021) follow calls to problematise the notion of a sole ethnographer (Creese et al. 2015). They underline the importance of creating flexible alliances with ‘knowledgeable collaborators’ (Lorette 2023), where researchers can work hand-in-hand with participants, guides and interpreters in an as non-hierarchical manner as possible (Ganassin and Holmes 2020). By adopting an ‘ontological imperative’ to prioritise language (Costley and Reilly 2021: 1041) in such settings, researchers also have the advantage of making audible a “plurality of multiple authorities” (ibid 1042) to strengthen insight, nuance reflexivity and enhance the potential for real-world impact. It is claimed that this approach can be especially powerful for projects that seek to advance representation of linguistically diverse and minoritized populations.

While extensive recommendations are likely to prove helpful in shaping future research plans for collaborative endeavours, doctoral students comprise one group of researchers for whom the journey is often uncomfortably solitary. Although some may solicit informal advice and practical support from fellow students, participants, and individuals they meet in the field, the inherent unpredictability of ethnographic fieldwork in a diverse context can make it difficult to anticipate the languages one is likely to meet and what one should do if, or when, faced with stretches of incomprehensible talk (Andrews et al. 2019; Lorette 2023). This is not to suggest that doctoral students are any more likely to be confronted by ‘spaces of linguistic non-understanding’ (henceforth, LNU) than more experienced researchers: communicative opacity can of course be alienating and destabilising for both those in the early stages of their career (van Hest and Jacobs 2022), and beyond. In Phipps’ (2013) highly personal and moving account, reflecting on perceptions of her own ‘linguistic incompetence’, she lays bare the complexities of holding oneself to account on a lack of language(s), and lects lost. Highlighting the ethical obligation to acknowledge the limits of one’s repertoire, Phipps encourages researchers to avoid dwelling in a state of lanxiety, side-stepping unpredictable situations or unfamiliar repertoires: this may be to ignore the superdiverse realities of everyday communication and research demands, as well as running the risk of observational and analytical paralysis. Instead, I suggest that we may be better placed to begin with an acceptance that stretches of LNU are likely to be both commonplace and rhizomic in nature (c.f. Deleuze and Guattari 1988), with a network of invisible, convoluted, interrelated and symbiotic constituent factors complexifying interaction, and our understanding of it.

In this vein, it is important to note that linguistically and culturally diverse environments are regularly recognised as being particularly well placed to foster a sense of liminality and hybridity (Bhabha 1990; Li 2018). Following a post-colonial logic, Bhabha argues that cultures can be seen as social constructions, “constituted in relation to that otherness internal to their own symbol-forming activity” (1990]: 210). As mutually constitutive concepts, they are therefore “subject to intrinsic forms of translation” (ibid), with iterative processes of othering and reinscribing identit(ies) creating liminal, or ‘third’ space(s). Bhabha maintains that such spaces can enable new cultural articulations and translation of difference. Closely linked to this is the importance of relationality, that is how one uses language(s) to position oneself and build relationships with participants and colleagues (Andrews et al. 2019), while still remaining cognisant to the delicate line between forming connections and remaining sensitised to difference (Rampton 1995). Canagarajah encourages researchers to develop “cooperative dispositions and performative competence for cosmopolitan relationships” (2013]: 202), whereas Andrews and Fay (2020) emphasise the need to nurture a ‘translingual mindset’ that can enable individuals to be “prepared for the unexpected, dynamic, or even playful uses of language in their research contexts” (Polo-Perez and Holmes 2023: 741; see also, Rampton 1995). Leading by example, Phipps (2013) encourages us then to remain mindful of our own linguistic shortcomings, while simultaneously urging us to focus on personal capability, that is, not just the skills and assets that we do possess, but also those that we have freedom or the opportunity to be able to develop in a specific context, or at specific times (Nussbaum 2011). By enhancing our capacity for attentiveness and attuning to the paralinguistic and affective aspects of communication, as well as sensitivities to research context, our ability to build relationships with collaborators, may be enhanced and our analyses improved (Rolland et al. 2023).

3 Methodology

Underpinning the tenets of qualitative research is an understanding that every aspect of data collection is saturated with subjectivities, from the moment of project inception. The questions as to where research is conducted and who is recruited to take part, are ostensibly guided by a (pre)determined orientation to phenomena of (presumed) relevance to a (research) community. In the context of the doctoral project introduced today, my overriding concern was in exploring the role of multilingualism in intercultural health encounters. My interest in the topic had been piqued by migrant accounts of antenatal care in the UK and guided by the potential role of language (in its broadest form) as a social determinant of health (Federici 2022). This paper shares extracts from fieldwork, a corpus which comprises 12 clinical observations, focus group and individual interviews, and extensive fieldnotes: these were collated over a period of 6 months in an antenatal department, in an NHS London hospital. In line with previous sociolinguistic research in diverse, as well as healthcare, settings, I approached my research site from a linguistic ethnographic perspective (e.g. Blommaert 2013; Cox 2017; Roberts et al. 2004; Simpson 2016). An understanding that “language and social life are mutually shaping” (Rampton et al. 2004), enabled me to explore “the processes … that shape urban encounters [and] everyday negotiations with difference, and practices of accommodation” (Wise and Noble 2016: 427). Similarly, a close focus on language allowed further examination of (tensions within the) wider sociocultural context that interactions may index (Gumperz 1982).

During the research journey, in the linguistically and culturally diverse setting where my study was situated, it became increasingly apparent that languages spoken in addition to English were difficult to anticipate. This inevitably raised anxieties about being able to gain informed consent from patient participants. Given that I also knew that I would have very little time to adequately approximate patients’ understanding of spoken or written English prior to their appointment, I designed a series of differentiated consent forms. To try and make the forms both syntactically and epistemologically comprehensible, and as ethically robust as possible, I followed principles underpinning the creation of differentiated resources for language learners, and those proposed by NHS Health Education, to address readers with low literacy skills (Learning and Work Institute n.d.; NHS Scotland n.d.). Consequently, forms were designed for readers with English competency equivalence to CEFR levels of B2/A2/A1. Moving away from presumptions of literacy skills in English, a pre-literacy form was designed with a series of picture prompts to enable me to talk through my research with potential participants: ideally these were to be used with support from interpreters.

Although key members of staff helped me to understand the relevance of specific appointments and the institutional processes of engaging interpreters, spaces of LNU emerged during observations, and frequently extended to the introduction of unfamiliar medical terminology and specialist acronyms. In the material conditions of a consulting room, I had no recourse to linguistic or insider support and was precluded from spontaneous participation in dialogue, because of the normative constraints of medical appointments. In this context, it was not just my limited linguistic repertoire which prevented inference and interaction but also my lack of epistemic and institutional authority. Lacking the immediate affordances of language congruence and professional knowledge, I was often only able to focus on what was possible in the moment and impelled to embrace a capabilities approach (Phipps 2013). I therefore sought to capture interactants’ intentionality through ‘thick description’ (Geertz 1973) and by enriching fieldnotes with detailing on context cues such as expression, gesture, prosody, touch, mime and/or drawings.

On returning to my desk, I adopted a reflexive stance, thinking back on the implications of observations and shaping an intentionality of gaze for future forays into the field. Central to these reflections were ethical questions prompted by the consultations I had attended. Although my proposed research had been approved by the hospital and university ethics boards, preparatory work had not fully equipped me for the highly personal and sensitive information that I was to encounter on an everyday basis. Ensuing discussions with supervisory and experienced colleagues gave me space to think about whether specific data were relevant to research questions or principled to share. They also allowed me the time to reflect on Holliday and MacDonald’s argument that “interpretation begins to some degree even at the data generation stage” (2020]: 634). With this in mind, I recognise that my presence as an observer, and the notes that I made, create a complicity in interaction as well as in interpretation and documentation of what I perceived was going on (Bucholtz 2000). Accordingly, it was my input, what I later decided to foreground and transcribe, and who I chose to undertake translation, that set in motion an assemblage of co-constructed realities that can only claim to be a partial representation of an authentic health encounter (Holliday and MacDonald 2020; Károly 2022; Ochs 1979; Vigouroux 2009).

Despite recommendations that advocate the use of (a) research assistant(s) who may be able to support the fieldwork linguistically, this is rarely possible in doctoral research (Sepielak et al. 2023). Instead, I was reliant on help after data had been collected, and this took the form of various kinds of formal and informal translation: while I was able to employ a professional translator on several occasions, at other times it was less easy to both find and/or finance additional professional services. On these occasions, a pragmatic decision was to recruit fellow doctoral students, and/or colleagues, who were proficient speakers of the sought-after varieties. I did so with the understanding that they were likely to have experienced (doctoral) training on the ethical, sensitive, and confidential dimensions of research, and to consider the responsibility carefully. It is well recognised that translators inevitably bring their own perspectives to a script, consciously or subconsciously, and add a layer of subjectivity to the interpretation (Károly 2022; van Hest and Jacobs 2022). Once dialogic practices of back translation, checking and confirmation are introduced the process becomes complex, especially if misunderstandings and variability emerge (Hennink 2008; King 2023; Thompson and Dooley 2019). Thus, far from being neutral texts that capture an objective reality, transcripts can be seen as ‘creative and politicized’ textualized documents, shaped by researcher(s) and collaborator(s) (Bucholtz 2000: 1440). We must recognise that material is selected to best support our argumentation, epistemologies, and ontologies: but even without guile, it is an opaque process, interpolated by ideologies and demanding a vigilant self-awareness and reflexivity.

In the following section, I introduce extracts from three case studies which document episodes and different aspects of LNU, as they occur in authentic medical consultations. Data is transcribed using conventions associated with applied conversational analysis in order to shed light on communicative detail (see appendix 1; Jefferson 2004; ten Have 1990) and to help explore interaction in diverse healthcare encounters. I begin with an account of a brief, unrecorded, consultation which is mediated in Hindi, a language I do not speak. It leads me, for the first time, to think explicitly about what researchers can gain from paying attention to body language, prosody, gaze, and physical contact. Following this is an extract from another appointment conducted entirely in Hindi, this time with a consultant whose repertoire mirrors that of her patient: faced once more with opaque dialogue, I focus on corporeal and relational aspects of the exchange. The final three extracts come from different stages of a multi-authored consultation, featuring a Portuguese patient, her companion, a midwife and, for a time, a professional interpreter. These illustrate how episodes of interpretation can enable clarification but also hold the potential to compound or obfuscate (epistemic) (mis)understanding. As this appointment continues, the apparent arbitrariness of content prompted me to engage another translator (translator B), to double-check that the first (translator R) was not mishearing utterances. I seek to demonstrate that as iterative processes of translation, transcription, questioning and clarification bring multiple opinions to the fore, they blur what can be understood as interpretation in its broadest sense.

4 Data

5 Discussion and conclusion

In this final section, I reflect on the complexities of conducting research in contexts where our comprehension and interpretation may be fragile or contingent: I conclude by making some tentative recommendations to support researchers at the beginning of their careers.

Responding to calls for open conversations about the complexities presented by researching in multilingual contexts, this paper has sought to illustrate some of the challenges that may be presented when a researcher does not share the linguistic repertoires of their participants, where stretches of talk are witnessed but not fully comprehended, and assumptions are made about (false) fluency (Costley and Reilly 2021; Sepielak et al. 2023; van Hest and Jacobs 2022). Far from being an unusual situation, as populations grow and diversify, it is increasingly likely that researchers, especially ethnographers who immerse themselves in a particular context, may encounter language incongruence. This can sometimes be predicted, prepared for and in some way mitigated by employing co-investigators, participatory methods, interpreters, translators and/or cultural mediators. Indeed, there exists a plethora of helpful recommendations by experienced colleagues that pave the way for ethically responsible, linguistically representative and rigorous research, and which encourage engagement with multilingual methodologies (see for example, Andrews et al. 2019; Blackledge and Creese 2010; Byrd-Clark and Roy 2022; Creese 2015).

However, there is less guidance for the sole researcher working in diverse settings and who may be obliged to explore other ways of knowing and being, in order to make some sense of the interaction that they observe. Phipps reminds us that although approaching spaces of LNU may invoke frustration and a sense of personal ineptitude, to accept the unknown and the unknowable, will increase our capacity to both infer and “engender qualities of empathy” (2013]: 340). Indeed, on personal reflection, I can see that the way in which I felt able to respond to the experiences of extended linguistic non-understanding (as shown in extracts 1 and 2) developed considerably over the period of fieldwork as I endeavoured to focus on modes of communication, other than spoken language(s). My gaze was drawn to the non-verbal and embodied relational work of nursing staff, a capacity for which they are frequently recognised (Bredmar and Linell 1996) and that seems to have the potential to enhance patient comprehension (see for example, Brooks 2022).

Nevertheless, there is undoubtedly a need for further methodological guidance for doctoral students working in multilingual environments and with diverse populations. It is challenging to navigate complex linguistic landscapes unaided, often with limited recourse to institutional funding and remote support from supervisors and/or colleagues. Although rapport can be built with participants and institutions, good relationships may not always create the conditions for collaboration or assistance with interpreting and translation. During my doctoral journey, for example, I had the support of wonderful hospital staff who facilitated access to consultations during data collection. Ironically, an unforeseen consequence of my ethnographic fieldwork was that the more I began to be seen as part of the team, the more I was included in technical conversations. The ebb and flow of jargon, containing references to acronyms and specialist terms, often proved epistemically challenging. Yet it was not incumbent on the professional to help me with unfamiliar medical terminology or to translate on occasions that they shared the same repertoire as a patient. Their responsibility was to help the patient, not the observer. These experiences point to the fact that, despite linguistic concordance, spaces of LNU can emerge when one enters a new community of practice, and it may be helpful to prepare early career researchers for this.

Costley and Reilly (2021) note that while some challenges are just seen as part and parcel of the research process, little attention is given to pre-fieldwork guidance, and I echo their call for a ‘methodological multilingual turn’ to prioritise a rigour that is “reflected in our training, preparation, professional practices, and disciplinary expectations” (2021]: 1039). While acknowledging the unpredictability of diverse populations, and the tensions that the costs and feasibility of preparatory language lessons surely provoke, there is much that could be done to prepare students for situations where they may not speak the language(s) spoken in their research setting. Doctoral training centres are often uniquely well placed to facilitate relationships between different departments, staff, and students. Working more closely with others, to offer reciprocal translation or reflect on challenges encountered in fieldwork for example, can mitigate the sense of isolation often felt by the novice researcher: indeed, the benefits of cooperation are clearly valuable in practical, financial, developmental, interdisciplinary, and ethical terms. However, it is also through encouraging researchers to accept that they will encounter moments of communicative opacity, that we can begin to sow the seeds necessary to approach alternative dimensions of relationality with humility, curiosity, and purpose.