Studying COVID-19 in light of critical approaches to risk and uncertainty: research pathways, conceptual tools, and some magic from Mary Douglas

Abstract

This editorial is a response to the recent COVID-19 pandemic and underlines the valuable role that critical social science approaches to risk and uncertainty can play in helping us understand how risk is being understood and mitigated. Drawing on Heyman’s approach to understanding risk as a configuration of probabilistic knowledge, time-framing, categories and values, I explore COVID-19 risk in relation to each of these features while also emphasising how different features stabilise one another. I suggest lines of inquiry into each of these features and their interrelatedness. I then move to present some important insights from the work of Mary Douglas which are especially germane to studying the risk of COVID-19 and, again, I raise possibilities for future research. Emphasising the centrality of ritual to Douglas’s theory, I develop these considerations to encourage an exploration of magic and magical thinking, alongside rational approaches to COVID-19 risk.

Keywords:

• risk
• probability
• values
• categories
• time-frames
• Douglas
• rituals
• magic

Introduction

This is my first editorial since taking over as Editor of the journal in January 2018. I have delayed writing an editorial up until now largely due to my focusing on the wider challenge of seeking to follow the excellent work of Andy Alaszewski, my predecessor and founding editor (along with Jill Manthorpe). Moreover I was keen not to attempt to put forward a particular editorial commentary before having first thoroughly immersed myself in the running of the journal for a significant amount of time. In light of these concerns, two years seemed like a legitimate basis upon which I felt I could write something outlining some aspirations for the journal and the wider field to which it speaks. I would characterise this field in terms of critical studies of risk and uncertainty in the social sciences and beyond (including the medical humanties, for example) as these address matters of health, illness, care, professional work and healthcare governance.

Towards the end of 2019 I began to consider the likely foci and emphases of this first editorial. My thinking was importantly shaped by some of the themes cutting across the articles we had accepted for publication towards the end of last year and which appear in this issue (Becerra et al., 2020; Feltham-King & Macleod, 2020; Kiviniemi et al., 2020; Moore, 2020; Webster, 2020), my own renewed interest in Mary Douglas’s work after reading 6 and Richards (2017) excellent overview of this great thinker’s wider work, and the opportunity to address a workshop on risk and risk-work, organised by the British Sociological Association risk study group in London in December.

More specifically, the articles in this issue raise important themes regarding the limits of scientific bureaucratic approaches to risk and the related manner by which practices are shaped by complex webs of accountability relationships (Feltham-King & Macleod, 2020), the way contrasting approaches to the same ‘risk’ can be understood through the influence of wider cultural frames and hierarchies (Webster, 2020), the common experience of not knowing, ambivalence and fatalism in the face of risks (Becerra et al., 2020; Kiviniemi et al., 2020), and the important role of rituals for understanding our preoccupation with some risks and not others (Moore, 2020). What these findings have in common is that they all echo, while also helping us to develop further, some of the most important insights which Mary Douglas’s work has contributed to studies of risk and uncertainty.

Arguably the cornerstone of our field, Douglas remains a common reference point across many of the three hundred submissions we have received in the past two years. Yet these references typically make little more than a polite nod towards her works and very seldom could be described as thorough engagement. Even where authors do refer more directly to Douglas, this rarely equates to a sustained engagement with her work and, as I will note later, risk researchers (myself included) have tended to operate with a narrow understanding of her central theoretical approaches.

I had initially planned to elaborate some of these considerations, alongside a call for deeper engagement with a wider range of Douglas’s work, in relation to recent governance questions regarding vaccine hesitancy and particular directions in political discourse around science, authority and expertise which have been (more or less problematically) referred to under the ‘post-truth’ epithet. However the pertinence of critical insights – not least from the critical social science approaches to risk and uncertainty published in this journal in the past, and from Douglas – into recent discourses, practices and policy around COVID-19 (e.g., Cyronoski, 2020) were too compelling for me to ignore. Hence I have the focus of the main body of this editorial onto some reflections and potential lines of research in relation to COVID19, drawing partly from important work published in this journal. I then later begin to develop insights specifically drawing on the work of Douglas and related approaches to ritual and magic. These ‘beginnings’ will be fleshed out more fully in a further editorial on Douglas later this year.

Studying COVID19 through the lens of the critical social sciences of risk and uncertainty

As I write this editorial, we have just marked twenty-one years of the journal’s publication with a virtual special issue (Health, Risk & Society, 2020). The twenty-one articles drawn from across the twenty-one volumes include many important insights which are germane to studying how ‘we’ – as a global community, as national polities, as organisations and as citizens – are handling Corona and, moreover, why we are handling it differently. For if there is one common preoccupation underlying the many media articles and social media posts I have come across in recent weeks, it is the question of why are ‘we’ (usually a particular national government or governance regime) not approaching this COVID-19 risk like ‘them’? As of early March, which ‘them’ is changing on a regular basis – be it Italy (on the extent of its physical distancing approach), South Korea (on widespread and proactive testing), or Taiwan (on its multidimensional and proactive approach) – but work published in this journal helps us understand the inevitability of, and manifold reasons for, these variabilities in risk governance across national policy regimes.

These differences, often cultural-organisational in origin (Douglas, 1986; 1992), include different institutional processes in risk management which have been understood as developing particular self-perpetuating logics of their own (Rothstein, 2006), not least in the evolving relationship between authority, transparency and authenticity (Brown & Michael, 2002). These developments in risk governance impact upon, and are shaped by, changing forms of trust relations in society (Alaszewski, 2003) which – at least in relation to healthcare professionals – may be better understood as changing from more blind-passive forms towards more critical-active forms (see Walls et al., 2004 for an incisive conceptualisation), rather than a decline per se. These studies thus push us towards a more nuanced assessment than simplistic accounts of a post-trust or post-truth (c.f. Löfstedt, 2005). It is also vital that we pay attention to different sub-groups in political systems, and acknowledge how varying experiences and perceptions of government and healthcare organisations, shaped at the intersections of class, gender, race and ethnicity, will shape very different relations and approaches to risk (Finucane et al., 2000).

These more general insights can be further developed by considering specific features of risk, as we start to deconstruct the nature of COVID-19 risk. A basic starting point here is to recognise that risk is never just about probabilities but, of course, entails a particular way of handling these numbers in relation to values (Aven, 2013; Szmukler, 2003). This approach has been taken further in our two year special issue project (issues 14:2; 14:5; 15:1; 15:6–7) whereby the classic Royal Society (1992, p. 2) definition of risk – ‘the probability that a particular adverse event occurs during a stated period of time, or results from a particular challenge’ – was developed into a deeper set of understandings, through which risk can be seen as involving particular configurations of probabilities, categories, time-frames and values; following Heyman (2010). Below I move to consider recent discussions of COVID-19 in light of some of these insights. I use these four features as headings under which to organise a number of critical considerations of the COVID-19 pandemic and related risks, though in several senses it is artificial and analytically awkward to attempt to disentangle these assemblages. Indeed, as I will show, probabilities, time-frames, categories and values qualify, (de)stabilise and feed back upon one another in several integral ways.

Probabalistic knowledge of COVID-19

Central to discussions of the COVID-19 virus and its ‘risk’ have been arguments around probabilities. Two key sets of numbers have been discussed in particular – mortality rates and rates of infection – with these numbers being used to consider questions of categorisation and time-framing. The risk of death from contracting COVID-19 has become particularly germane to sense-making processes and debates over the virus. That the mortality rate in China in late February was reported at around three per cent was one important justification for differentiating COVID19, from seasonal flu and the H1N1 virus, for example. Though these understandings of mortality rates have become increasingly destabilised through a growing wariness around our knowledge of the denominator for mortality rates of both seasonal flu and COVID-19, especially as the initial numbers emerging from China started to be compared against growing numbers of cases and reported mortality rates in South Korea (0.65 per cent – Power, 2020) and Italy (close to 8 per cent – Bernard et al., 2020).

These differing numbers and comparisons have generated important epistemological and ontological questions regarding the nature of COVID-19. At the time of writing, it remains unclear how much of these differences are explainable by: wider testing in South Korea, thus increasing the presence of mild cases in the denominator of the mortality rate, thus ‘better’ reflecting the spread of COVID19; Italy’s ageing population (see Dowd, Rotondi, Andriano, Brazel et al., 2020); different types of the virus developing over time; the concentration of cases in confined time periods and local areas in Italy (and areas of Lombardia in particular) which have overstretched local healthcare services, thus compromising treatment possibilities; and/or public reactions towards knowledge of overstretched services which has delayed when some people with severe symptoms attend hospital. These epistemic challenges, which are only being tackled slowly, leave the precise ontology or nature of the virus ambiguous to some degree, with even more uncertainty over how it will impact upon different local populations, cultures and healthcare contexts as it continues to spread.

The Italian case in particular has been important in framing discussions elsewhere in Europe (in this editorial I refer more to discussion and coverage in northern European contexts as I am better placed to write about these), highlighting a number of concerns around the speed by which the virus can spread, the number of people who may require hospitalisation and intensive care, and the high fatality rates which would result from services being overstretched in particular localities. Common to wider discussions of the Italian case are charts suggesting that most European countries are facing a growth rate in cases which is similar to Italy, albeit with the start of such growth occurring a number of days later (for example, Financial Times coverage – see Bernard et al., 2020). There have thus been fears that other European countries could, like Italy, share a similar mortality rate of close to eight per cent. Concerns regarding such a rate are based on important assumptions regarding similarities in health system capacity, localised outbreaks, demographics (Dowd et al., 2020), culture and contact. Italy has the largest percentage of people aged 65-and-over in Europe and one of the lowest percentages of older people living alone (Eurostat, 2020). Yet, because of the multiple, complex and potentially countervailing factors noted above, it remains very difficult to infer likely death rates from infection rates, not least because of the different testing approaches across regions and countries.

Despite the multiple uncertainties noted here – regarding the ontology and epistemology of the present, and the assumptions and validity of models to consider the future – COVID-19 ‘probabilities’ are frequently and intensively discussed in news media and social media coverage. Earlier research on public understandings of disease outbreaks has noted that ‘frequent media coverage may introduce a disjuncture between perceptions of personal risk and estimates of population incidence, rather than simply increasing perceptions of risk per se’ (Young et al., 2013, p. 112). What has been striking in the coverage of COVID-19 in early March, in news media and the social media which increasingly inform these news media, have been regular disjunctures between accounts of front-line staff, which have highlighted death, rationing and examples of young people receiving intensive care treatment and/or dying, and narratives centred upon epidemiological models and data which emphasise the heightened relative risk to older people. More or less explicit in some of these accounts is the relative safety of the mainstream ‘normals’, and the implicit othering of those who are older and more vulnerable. In these senses, risk ‘does not signify an all-round assessment of probable outcomes’, instead it becomes profoundly political and moral (Douglas, 1992, p. 39), though the precise ramifications of media-driven understandings, and how these come to be understood and internalised amid wider lifeworlds, in turn shaped by earlier health policy discourses of ‘genuine’ and legitimate illness (Parsons, 1975, p. 259), require further investigation.

The different time-framing of COVID-19

Different epidemiological framings of COVID19 and their insidious moral effects are vitally bound up with discursive time-frames and different imaginaries of ‘future time’ (Brown & de Graaf 2013). This contrast has become especially vivid in the contrasting stances of the UK Government and that of the World Health Organisation (WHO). As Dr Michael Ryan, Executive Director of the WHO Health Emergencies Programme stated in a press conference in Geneva on 15 March 2020, ‘be fast, have no regrets. You must be the first mover. The virus will always get you if you don’t move quickly’. This emphasis on short-term speed, is echoed by discussions of interventions framed in relation to Italy and overwhelmed healthcare provision (Bernard et al., 2020).

In contrast, the UK policy up until the evening of 16^th March at least, has tended to emphasise a longer-term time-framing of risk and policy interventions, with the emphasis being placed on developing herd immunity over a period of months, alongside ‘cocooning’ the more vulnerable in the meantime. This orientation has been combined with much less of an emphasis on testing for infection. Hence graphs comparing countries on increasing numbers of cases suggests the UK has a flatter growth curve compared with other European countries (though much steeper than Japan, Singapore and Hong Kong), and yet the rises in deaths in the UK is (at the time of writing, and along with Spain) among the very steepest. The practicalities of the approach, combined with some of the earlier rhetoric of the UK Prime Minister, suggests that the longer-term framing is one where a goal of herd immunity assumes death as a ‘natural’ if regrettable part of the pandemic solution. This approach to the pandemic has been increasingly challenged in the UK, and to some degree reflected – practically and rhetorically – by the approach of the Dutch Government. First Minister Mark Rutte’s March 16^th national televised speech extoled the importance of developing herd immunity among lower risk groups, with the approach of shielding vulnerable and older people similar to that of the UK. In an earlier press conference Rutte had stressed ‘We are a level-headed people who are not waiting for symbolic measures’ (Hendrickx, 2020).

That the UK and Netherlands governments’ risk governance approaches appear to contrast with several other European countries could be interpreted as largely political – with both governments following distinctively right-leaning, economically liberal, socially conservative and individualising policy trajectories – though may also have much to do with cultural norms which shape political imaginaries regarding possible interventions of the state into the private sphere (at least in the domain of public health). Both governments also emphasise the importance of public health expertise and the need to trust in epidemiological experts.

Rather than choosing between explanations of politics, culture or trust in experts, several studies in critical social science approaches to risk characterise public health governance, culture and politics as very much interwoven – and typically connected and oriented by a distinctively neo-liberal logic of governance (Petersen and Lupton, 1996, p. 178). If science-as-culture and politics combine to orient COVID-19 responses towards longer-term horizons of possibilities, then critical social scientists should ask which more proximate realities (whose deaths, for example) are being ‘bracketed off’ or ‘looked past’ by this distant gaze; and at what cost. They should also ask how policy-making, and levels of cooperation with these policies, are configured by multiple trust relations (see Brown and Calnan 2016). These would include trust in some sources of advice (national experts or WHO) rather than others, alongside how the level of policy-makers’ trust in citizens shapes which policy approaches they deem feasible. Policies based on presumptions of low levels of trust may further undermine trust and cooperation, especially where these policies are less inclusive (Brown and Calnan 2016).

Who and what are being homogenised and categorised in COVID-19 risk?

In order to make ‘events’ (infections; deaths) countable and amenable to probabilistic projection into the future, these events are lifted out of their space-time context (Giddens, 1990) and, in doing so, become homogenised and categorised between given boundaries (Heyman, 2010). Though seemingly benign, technical or arbitrary, these processes of disembedding, homogenisation and categorisation have profound political ramifications. As already noted, risk necessarily involves a form of thinking which – by its abstract quantification and dissolving of context – has a tendency to ignore the whole person and their humanity, through its focus on ‘factors’ and a wider epidemiological logic (Castel, 1991). The categories that do emerge have deeper institutional roots and, from a Douglasian perspective, we should be attentive to ‘the way our classification system is set’ and how this relates to organisational dynamics (Reis, 2019, p. 2; drawing on Douglas, 1986). In the UK policy discussions of COVID-19, a boundary of age (especially 70 years of age) has emerged as one dominant categorisation by which risk is understood. This is not the only category by any means – various pre-existing conditions including heart and lung conditions are also emphasised – but age has become the primary basis for categorising risk in policy and wider discussions.

This emphasis on age is partly an artefact of which data are commonly collected across the many countries affected and which categories are emphasised by the WHO. International organisations thus play an important role in bringing ‘risk’ into being in particular ways (Figuié, 2014). The aggregation of numbers, the collection of data on certain categories and not others, and the ways these are compared across countries has been seen above to provide a particular understanding of the condition itself, which has come to be categorised through these framings. A common basis of comparison and debate in late February and early March has involved seasonal flu, either as a way of stating the low total mortality figures of COVID-19, either in relation to the more mundane and (currently) more deadly seasonal flu (Harvard Medical School, 2020), or as a way of highlighting the mortality rate of COVID-19 (as discussed earlier) and denoting the differential speed by which COVID-19 can spread.

Numbers and time-frames (pertaining to age and mortality rate, or speed of infection, for example) are in these ways drawn upon in delineating and stabilising categories, while these categories then become the basis of probabilistic models and the emphasising of some time-frames over others; in turn recasting wider discussions and understandings of risk. Relatively low mortality rates in younger age groups have been used, albeit often more implicitly, to down-play the risk of COVID-19, depicting it as more mundane and ‘flu-like’. This categorisation would, in turn, seem to neglect both those of younger age who will suffer and die from COVID-19, and those who are older but, by being neglected or omitted, come to be portrayed as marginal to mainstream society and everyday life.

These tendencies in categorisations, delineations and comparisons are more apparent in some contexts than others; with the UK and the Netherlands, again, standing in contrast to some southern European and east Asian countries. If, following Douglas (1966; 1986), we analyse ways of thinking (‘thought-styles’) and use of categories as being rooted in institutional dynamics and everyday interaction rituals (6 & Richards, 2017), then we might look to the relative (mis)recognition, connectedness and importance of older people in everyday life in these different cultures to begin to understand contrasting categorisations of the COVID-19, and the use of framings pertaining to age. With these underlying processes of categorisation importantly bound up with processes of (de)valuing.

Giving and denying value through COVID-19 risk

Past research into health risks has explored how categorisations of people, events and conditions are inextricably bound up with processes of (de)valuing and (de)legitimation (Caiata-Zufferey, 2012). While COVID-19 is continually evolving and has seemingly diverged into two types (Tang et al., 2020), these features of heterogeneity and change are neglected in most news media and social media discussions. Depicting COVID-19 as one stable virus may serve to legitimate its risk while, simultaneously providing some reassurance by facilitating a growing relative familiarity with this same ‘risk’. In contrast, objects which are unfamiliar, liminal (Warner & Gabe, 2004), fluid and which cross boundaries (Douglas, 1966) are more likely to be experienced as problematic and dangerous.

That COVID-19 aroused such elevated levels of concern and media coverage in Europe in the early weeks of the disease outbreak, when total numbers of fatalities were relatively low (compared to other major causes of mortality), can be understood in terms of the role of key claims-makers (WHO) and media reporting, but also in relation to the source and spreading of the virus – from a locality in a distant country which is relatively unfamiliar (Wuhan), ‘other’ and yet powerful; the fast transmission of the virus across boundaries and coming closer to ‘home’; and the limited effectiveness of attempts at curtailing its spread. When presented in particular ways – at the time of writing, President Trump is commonly referring to COVID-19 as ‘the Chinese virus’ on his twitter account – these ‘other’ and anomalous characteristics of the virus were likely to arouse strong emotions due to underlying ways in which we make sense of the world around us through attempts to categorise (Douglas, 1966; Lupton, 1999, p. 131).

Developing some of these analytical threads still further, Lupton (2013) draws on the work of (Douglas, 1966; 1992) and Ahmed (2004) to consider ‘emotion-risk assemblages’. This conceptualisation aims to capture how affective responses to objects are configured in ways which are powerfully underpinned and legitimated through historically-constituted notions of ‘pure bodies’, which are represented as in danger of being violated from other bodies from ‘outside’ (Ahmed, 2004), and which are therefore responded to as dirty – that is, as ‘matter out of place’ (Douglas, 1966, p. 50). For Lupton (2013: 640) ‘part of what the focus on space, place and the production of emotion is able to offer is recognition of the shifting dynamics that are inherent in the embodied nature of risk understandings’ and their political consequences.

Emotions thus form a vital mechanism which helps us understand the connection between categories – of objects and time (time-frames), for example – and values. Shifting categories, probabilities and their (mis)representation, as information and media discourses around COVID-19 continually evolve, thus arouse different emotional reactions – creating new forms of contestation and risk while shutting down debate on others. Slovic’s (2012, p. 409) concern that media and marketing ‘imagery’, and related imaginaries, have powerful effects in elevating or ‘suppressing perception of risk and manipulating behaviour’ is thus vital to considering the shifting processes of (de)valuing and legitimation.

The role of these images and imagining in providing us with ‘emotional repertoires’ (Cottingham & Fischer, 2017), which shape our responses towards, and valuing of, objects and ideas, may be especially influential where the ‘threshold between “dangerous” and “safe” has become nebulous and susceptible to categorical revision’ (Caiata-Zufferey, 2012, p. 427). Lupton’s (2013) conceptual work can be used to further develop understanding how objects, bodies, space and place become imbued with heightened emotional and ritual significance and the consequences of this. At least two important consequences are that everyday rituals and meaning-making generate forms of heightened emotion and magical thinking which, in turn, may come to warp and impede open discussion and critique in the public sphere (Habermas, 1987). I now turn towards a deeper consideration of rituals as one key way of widening the engagement of risk researchers’ engagement with Mary Douglas.

Broadening our use of Douglas through a focus on rituals and magic

As noted in the Introduction, my original plan for this editorial had been to develop a considered editorial plea for a more wide-ranging and in-depth engagement with Douglas’s work. Given the confines of an editorial, alongside the shift in focus onto COVID-19, this aim cannot be fulfilled in this essay. Nevertheless I will finish this editorial with some brief sketches of further insights which a Douglas-inspired research programme would afford us as we study social processes of COVID-19 in relation to risk and uncertainty.

Moore’s (2020) study in this issue, as with earlier work (Moore & Burgess, 2011; see also Burgess & Horii, 2012) picks up, importantly albeit more briefly, on Douglas’s interest in rituals. As 6 and Richards (2017) argue, everyday rituals are central to Douglas’s theorising of thought-styles and organisational dynamics yet this analytical emphasis,), tends to be neglected by risk researchers who usually only refer to Douglas's work with Wildavsky (1982) and her (1992) essays on Risk and Blame. Risk researchers have, as a consequence, generally tended to develop a somewhat impoverished understanding of Douglas’s theory, though Moore and Burgess (2011) engage in some detail with Douglas’s (1966) earlier writing on ritual.

Placing everyday interaction rituals more centrally (see 6 & Richards, 2017) shifts the analytical lens of the risk researcher towards interactions, practices and objects (Moore, 2020) – with important implications for how we might study COVID-19. As noted above, the nature, (in)frequency and value of interactions with older people in everyday society, the ritual ordering of these interactions and the hierarchies which are reproduced through these, may have important implications for wider thought-styles and the categories which become woven into our sense-making of COVID-19. As Becerra et al. (2020 – this issue) found in their study of the handling of occupational risks in Abidjan, the ordering of everyday practices and interactions reproduced age-related hierarchies, and in other contexts, fatalism, in ways which were vital to understanding risk experiences and inequalities in this context. Although these authors worked with a new methodological tool and did not engage with Douglas, their study helps us to envisage ways in which interactions and practices are shaped by, and in turn shape, categorical thinking and relating, which are highly relevant to grasping risk and inequality.

Ritualised practices and interactions can thus help affirm and reproduce existing hierarchies, ordering and cohesion, though Douglas is most interested in tendencies towards social and organisational conflict (6 & Richards, 2017). Anomalous objects and bodies often come to be especially important to driving these institutional dynamics – either as totemic objects which form a basis of unity and cohesion (see Moore, 2020) or as objects of contention, dispute or individualisation. Child vaccinations are one striking example where practices relating to an object have ritual like qualities, being seen by some as a ‘risk object’ (risky), and yet by others as an ‘object at risk’ (something valuable and under threat) (see Boholm & Corvallec, 2011). These contrasting views – of the biomedical mainstream and the ‘anti-vaxxers’ – have become increasingly opposed as each party seeks to convince and critique the other. Amid exposure to the others’ views and emotionally heightened exchanges, Douglas helps us understand why ‘it is essential for each culture to believe that the other cultures cherish wrong-headed concepts of justice, they are based on essentially immoral precepts [… this] analysis takes us far into the central theme of Cultural Theory, irreconcilable conflict’ (Douglas, 2007, p. 9).

Likewise, we see different ways of thinking about COVID-19 which have the potential to become further entrenched through interactions with others from a different thought-style or institutional position:

‘Here is a dispute between two who will never agree. No new facts will change the opinions of the pioneering individualist who cheerily asserts that all will be well, or those of the holy man who warns him of terrible dangers to be unleashed if he continues in his ways. Whatever information is tendered, their differences are irreconcilable’. (Douglas, 2007, p. 9)

Amid these rival styles of thinking about the world, anomalous objects, which do not quite fit into the existing categories of a thought-style, were of particular interest to Douglas and she saw them as especially salient – in their emerging through underlying institutional dynamics and dispute and, in turn, in driving and potentially intensifying these same dynamics: ‘rival styles encounter each other as generators of anomalies. Ritual enactments provide reinforcement for disputants, each to the point of disorganisation. Each rite is a fight. Conflict and anomalies proliferate’ (6 & Richards, 2017, p. 178). The way rituals and related understandings involving specific objects are affirmed in relation to the underlying institutional dynamics, conflict and the thought-styles these give rise to, help us understand the enduring value and power of rituals for how we think. Where ritual practices, and/or understandings of objects, come to endure in ways which seem unlikely or implausible given the evidence available, or which even impede the possibility of weighing and scrutinising the evidence in the first place, then we are usually studying thinking or practices which can best be described as ‘institutional magic’ (Alaszewski, 2015; Roth, 1957).

Far from existing at the margins of our cultures, Alaszewski (2015) argues that a combining of the rational and the magical is far more common in our everyday handling of uncertainty and risk than we often realise. If we take a broader, anthropological understanding of magic (adapted from Roth, 1957) – as a ritual practice involving an object or body, with the aim of making particular futures more likely – then magic is pervasive, including in approaches to infectious disease control in the modern hospital environments in North America (Roth, 1957). Similarly in China, the response to the virus has been one of rational efficacy, testing and intervention, and yet debates over the zoonotic source of the virus have returned to focus, somewhat improbably, on the pangolin (Cyronoski, 2020). Although there is a limited evidence for this claim, that discussions over the source of COVID-19 focus on this peculiar creature will be of little surprise to scholars of Mary Douglas, whose early work noted how the Lele people (in what is now the Democratic Republic of Congo) found the pangolin to be both abominable and having sacred, ritual powers (Reis, 2017). In understanding how people react to COVID-19 risks, it is important that we explore combinations of rational and magical/non-rational approaches to uncertainty (Alaszewski, 2015; Zinn, 2008).

Conclusion – some pathways for researching COVID-19

In this editorial I have sought to sketch out a number of existing lines of inquiry in critical social science approaches to risk and uncertainty and to apply these towards inquiring into the social processes inderpinning COVID-19 risk. In particular I have emphasised questions of probabilistic knowledge, time-framing, categories and values which cohere, more or less neatly, into (mis)understandings of risk. Far from separate parts of a risk jigsaw, these four components interweave fairly thoroughly and I have noted how these features may stabilise one another in some senses, despite wider questioning which undermines or interrogates risk understandings in other ways.

More specifically I have raised questions regarding probabilistic knowledge of COVID-19 and tensions between population-level epidemiological discourses and powerful mediatised narratives pertaining to suffering individuals; with the latter challenging the abstract horizon of possibilities of COVID-19 emphasised in the former. These tensions and challenges, in news media reporting, social media discussion and policy developments, would form one important potential focus of critical social science research around the virus and its risk.

A separate, but related line of inquiry would interrogate the time-framing bound up with these different levels (population or individual) of thinking about the virus and ways in which some time-framings come to be relatively accepted, whereas others remain more contested or shift over time. These time-framings necessarily focus our attention on some outcomes while leading us to neglect others. A valuable role for critical perspectives is to highlight which outcomes and which people may be being ‘bracketed off’ or ‘looked past’ in these framings.

I have understood time-framings and other ways of categorising objects, people, space, time and so on as related to particular thought-styles (Douglas, 1986) which, in turn, I have related to the everyday rituals which sustain and reproduce these modes of thinking. Further investigation into how different national governments adopt different approaches and make sense of risk through different master categories would benefit greatly from an orientation towards understanding the everyday workings of culture through ritual. ‘For, as a social animal, man [sic] is a ritual animal’ (Douglas, 1966, p. 62; as cited in 6 & Richards, 2017, p. 179).

These rituals and their relationship to risk are also of interest in how they configure emotions around future uncertainty. Rituals, media portrayals and other culture responses to risk are of great interest in their shaping of the values which are inherent to risk – especially in their role of providing affective ‘repertoires’ (Cottingham & Fischer, 2017) for coping with and responding to risk. I have underlined the potentially valuable role of research into emotion-risk assemblages (Lupton, 2013) and what these can tell us about the way objects, space, bodies and place are connected in ways which orchestrate particular feeling norms and dynamics.

Like Lupton, Douglas’s work also points us towards the importance of material objects and their emotional and symbolic significance for us. While Douglas borrows her interest in rituals from Durkheim, she differs in understanding ritual in a far more mundane, everyday sense. Along with a range of insights I have sketched which we might borrow from Douglas, central to research in this tradition would be a consideration of how practices in relation to particular objects – face masks (Burgess & Horii 2012), COVID-19 vaccinations when they emerge, or even toilet rolls, perhaps (which have been selling out of supermarkets in some countries) – evolve and what this tells us about changes in social organisation (hierarchy, integration) and categorical thinking. Similarly, we have discussed the centrality of older bodies for COVID-19 risk and how these are understood in relation to wider thought-styles and the everyday interaction norms which propagate these forms of thinking. An attentiveness to the way time-frames, probabilities and categorisations are combined to form meanings of risk, which connect to emotional sensibilities and valuing, enable us to pose questions such as how different the experiences and processes around COVID-19 risk might be if the main ‘objects at risk’ were children instead of older people. Given the sacred position of children in some societies (Zelizer, 1987), risk and risk governance might well look very different.

This is not to say that current COVID-19 emotion-risk assemblages are without magical thinking. Indeed my main goal in the latter half of this editorial has been to encourage a curiosity and investigation into the ‘institutional magic’ of risk thinking around disease containment (Roth, 1957) and to explore how the rational and the magical are combined in risk policies and everyday practices (Alaszewski, 2015) towards COVID-19.

While the works of Douglas which risk researchers are most familiar with tend to have a more pessimistic tone towards institution tensions and risk, a number of her later (post-1992) works employ her cultural theory in order to develop insights for ‘diffusing conflict’ and developing reconciliation (6 and Richards 2017: 146 – see chapter 4). Rituals remain central to this approach, though the focus is shifted towards developing narratives and rituals which translate inclusive values into social practices (p. 159). Developing strategies, narratives and practices which are inclusive, both symbolically and substantively – thus fostering trust, cooperation and communication across social divisions – represents perhaps the most important and tangible contribution which the critical social sciences can make to the current COVID-19 risk governance process. There is much more to be said on this aspect of Douglas’s later work however and this theme will form the central focus of an editorial essay which will be published later in the year.

Disclosure Statement

No potential conflict of interest was reported by the author.