Illustration: Diwas Bisht

What has COVID-19 taught us about how to share information and build trust among minority ethnic communities who use oral languages?

An active study of UK Bangladeshi experiences of the pandemic suggests that good evidence communication requires trusted local mediators and a narrative framing that is driven by the data, but relevant to the situations being faced by local communities

By Mark Cabling and Chris Tang

In their recent IPPO blog, Joanna Chataway and Molly Morgan Jones presented a strong evidence base for how trust is vital for the enactment of COVID-19 policy and the uptake of public health measures. They highlighted that trust inequalities among some local UK communities are linked to ethnicity and how public institutions are perceived.

It is easy to see why people from minority ethnic communities feel disappointed when the pandemic has thrown a spotlight on and exacerbated the social inequalities they face. Importantly, trust is won by word as well as deed, and government messages and information about COVID-19 have not been framed in ways that are linguistically accessible, culturally relevant or resonant for specific groups. The call for a focus on local interventions that improve and widen information accessibility is a welcome one.

Successful trust-building on a local level requires insights into the dynamics of trust for different communities. In this blog, we consider some communities who share an ‘oral language’ – a language that is uncodified, with no widely used written form. This has important consequences for how, for example, public health information is shared, interpreted and socially validated.

Oral languages in the UK

The notion of an oral language is nothing new. It is, in fact, the linguistic norm: in 1971, Munro Edmonson noted how, among the tens of thousands of documented languages, only 106 had evolved literature. It is a common situation for minority ethnic groups in the UK to rely primarily on a regional variety or dialect that is uncodified and oral for intergenerational communication, rather than the written codified version that is officially associated with that group. For instance, the majority of UK Bangladeshis speak Sylheti – an oral language from the Sylhet region of Bangladesh.

A team of community researchers, activists, and local GPs working alongside a group of linguists at King’s College London have an active UKRI study on UK Bangladeshi experiences of the pandemic. Our work includes promoting compliance and vaccine uptake amongst ethnic groups that speak Pahari, Sylheti and Pashto – languages spoken by an estimated 1.5 million people in the UK.

We have found that evolving shared understandings of COVD-19 and its mitigation has been dependent on the intervention of local doctors and politicians, community workers and leaders. All have acted as mediators between the written conceptualisations inherent to English messaging and information, and the spoken ways of knowing and understanding that are inherent to their own communities.

Informational vulnerabilities during COVID-19

The higher rates of vaccine hesitancy in people of Black and South Asian ethnicity reported in this study (cited by Chataway and Jones) were sadly unsurprising considering the informational vulnerabilities faced by communities that speak an oral language. Such vulnerabilities are linked to an informational ecosystem that is not geared to the way oral language communities communicate and spread information.

During the pandemic, those dominant in an oral language and with low English language proficiency – more often belonging to older generations – have been unable to access information about COVID-19 independently, and have therefore been reliant on family members, carers, friends and acquaintances.

Based on our interviews with UK Bangladeshi community elders and information-sharers living in East London, the task of information sharing from a written, codified language into an oral, uncodified one is rather daunting. Words such as ‘clinically vulnerable’ or ‘droplet’ are the product of written language – they refer to abstracted concepts that do not translate into an oral language, where expression is typically rooted in the physical and cultural world of the speaker.

For instance, older participants in our study expressed highly visceral conceptualisations of COVID risk. Their notion of comorbidities and poor immunity as ‘physical weakness’, for example, led to perceptions that personal behaviours such as healthy dieting and exercising, as well as folk remedies including special teas, could boost immunity and physical strength, thereby providing protection from COVID-19.

There is often a risk that linguistic and cultural reframing by information-sharers can lead to simplifications or alternative interpretations. For example, the interpretation of social distancing based on the conceptualisation of a ‘household’ as a safe container for family members both in the same residence and living nearby may lead to people socialising with a wider group of family members. For the elderly who have to shield, this may also lead to them having close physical contact with members of their household despite shielding measures recommending otherwise.

Social networks and misinformation

‘Narratives’ also play a crucial role in the spread of information within oral communities. By this we mean not just oral communication, but the exchange of information through anecdotal evidence and stories – where personal accounts, or the retelling of another’s personal experience, is used to share information by way of characters, plot and climax.

For some oral language speakers in our work, conversations with family members, friends and neighbours in their social circle have been the main source of information about the pandemic. But in different risk scenarios, social networks can just as easily exacerbate vulnerability as they can reduce it.

Our data suggest a reliance on immediate (verbal) social networks makes oral language speakers particularly vulnerable to misinformation, as information is harder to verify when reliable sources are inaccessible. It is, therefore, important to distinguish between the information-sharing that happens as a fundamental part of social learning in oral language communities, and an activity we refer to as ‘mediation’: the strategic cross-linguistic and cross-cultural reframing of information to promote messages as more intelligible, personally relevant, credible and actionable as part of formal and informal risk communication systems.

Note that mediators could be any trusted individual with the tech savviness, health literacy, and linguistic and cultural knowledge to access, interpret, assess and select from the information available and reframe it in ways that are meaningful to others in their community. There are strong suggestions from our data that the massive jump in vaccine uptake by Bangladeshis and Pakistani groups from February to April 2021 (which was triumphantly announced earlier this year) was partly driven by a localised response by doctors, councillors, religious and community leaders, as well as community outreach workers acting as mediators.

Crucially, as we will now discuss, these mediators accessed oral communities because they were already a part of them, and functioned within the dynamics of these communities’ pre-existing information systems.

Dynamics of trust for oral language speakers

A consistency we have noted in our work with Sylheti, Pahari and Pashto speakers is the high value placed on medical and scientific knowledge, which translates into trusting COVID-related information provided by experts and providers. This trust is beneficial from a risk communication standpoint, because health experts are well placed to understand scientific uncertainty and establish facts based on evidence. There could otherwise be a gap in mediation that could be filled with others who are less well-placed to understand public health measures and COVID-19.

At the same time, doctors and scientists outside of the oral community are not well positioned to be mediators because they do not have access to the oral language. To this point, trust within oral communities in the UK also operates locally, which makes sense given the important role of the messenger in how information is shared. The perception of being a ‘community insider’, combined with perceived medical and scientific expertise, are thus intersecting axes of trust that explain how medics from local communities have had such success as mediators.

At the same time, it is important to acknowledge that trust works differently within these communities depending on individual roles in the informational ecology. The Bangladeshi elders we interviewed in our study, for example, seemed to have retained trust in national institutions over a six-month period of the pandemic (October 2020 until April 2021), but for some very different reasons.

Older men, many of whom had taken an active role in the community’s original development, felt the onus was on the community itself to adhere to public health measures; for them, lack of compliance was typically judged as the fault of the individual and the community, not the government or other institutions for communicating poorly. In contrast, among older women (who were wholly or mostly reliant on others for information about the pandemic), their trust in government and power structures was widely based on a view that these institutions are mandated by the ultimate will of God, far beyond the influence of either scientific expertise or local communities.

Meanwhile, the younger information-sharers in our sample expressed strong distrust in national institutions, which were perceived as being complicit in the perpetuation of the social inequalities faced by their community and highlighted by the pandemic. By the time the first lockdown was ending, we found that the people most likely to access and share government advice (because of their digital connectedness) were the least likely to trust it.

Mediation – a solution to building trust on a local level?

Our projects suggest that, while information-sharing practices founded on the right principles can build trust on a local level, making scientific outputs fully accessible for oral cultural communities is not possible without mediation. Those who are primarily reliant on friends, family members and acquaintances for advice about the pandemic need linguistically and culturally framed advice which they can apply in managing daily COVID-related risk encounters.

Narratives are a powerful and persuasive knowledge structure, particularly for those reliant on spoken ways of knowing and learning. We would therefore argue that good ‘evidence communication’ still needs a narrative framing – one that is driven by the data, but also relevant to the situations faced by local communities.

For instance, it is possible to frame vaccination within a narrative in which the act of getting vaccinated empowers the community by helping businesses reopen, people to see loved ones again, and the easing of lockdown restrictions. Anecdote is also important: for example, experiential examples from trusted members of the community that have had the vaccination and are happy they took it. As so much rests on the messenger for these groups, messages also have their own story that is part of the persuasion – for example, ‘where did you hear this, and who did this person hear it from?’

Yet mediation is no easy feat. Sifting through the facts and uncertainties of science, media conjecture and fabrications, then selecting and reframing appropriately for their communities is a challenging task. An even greater challenge is to ensure consistent messaging by different mediators – a task that no doubt requires national and local coordination. Ultimately, mediation is likely to be most effective when it is personal and involves repeated contact over a period of time.

The pandemic has seen some exemplary outreach initiatives, such as those run by our partner the Asian Resource Centre Croydon (ARCC), where volunteers and community workers are trained to provide information and support to members of their communities. What would boost the success of these mediator initiatives is a consistent programme of training – both in the emerging science, and in evidence-based strategies for making information (as Chataway and Morgan Jones put it) ‘accessible, understandable, usable and assessable’ for their communities.

By way of example, our pilot training in Croydon was developed and run as a collaboration between a health expert (a GP) and experts in health communication. By employing a ‘train the trainer’ model, our trainees can themselves operate as trainers for volunteers embedded in their local communities. Our project team is currently building a framework for mediation so that such programmes can run on consistent, evidence-based principles to address the shifting dynamics of the pandemic.

Dr Chris Tang is Lecturer in International Education & Applied Linguistics at King’s College London (KCL). Dr Mark Cabling is a Research Associate in COVID-19 and Social Learning at KCL.