Seeing like an epidemiologist? Mobilising people against COVID-19

Abstract

Diaries and other materials in the Mass Observation Archive have been characterised as intersubjective and dialogic. They have been used to study top-down and bottom-up processes, including how ordinary people respond to sociological constructs and, more broadly, the footprint of social science in the 20th century. In this article, we use the Archive’s COVID-19 collections to study how attempts to govern the pandemic by mobilising ordinary people to see like an epidemiologist played out in the United Kingdom during 2020. People were asked to think in terms of populations and groups; rates, trends, and distributions; the capacity of public services; and complex systems of causation. How did they respond? How did they use the statistics, charts, maps, concepts, identities, and roles they were given? We find evidence of engagement with science plural; confident and comfortable engagement with epidemiological terms and concepts; sceptical and reluctant engagement with epidemiological subject positions; use of both scientific and moral literacy to negotiate regulations and guidance; and use of scientific literacy to compare and judge government performance. Governing the pandemic through scientific literacy was partially successful, but in some unexpected ways.

The uses of scientific literacy

During the first year of the COVID-19 pandemic, before mass vaccination had been achieved, the UK government – like many governments around the world – tried to get ordinary people to see like an epidemiologist. In daily press conferences watched by millions (Lawson, 2020), slideshows of charts and maps encouraged people to think in terms of infection rates, hospitalisation rates, and mortality rates, including how these varied over time, how they were distributed across space, and the differential risks posed to different social groups. Public health messages such as ‘Stay Home, Protect the NHS, Save Lives’ encouraged people to behave not only for their own benefit or that of their family and friends, but for the benefit of the population as a whole, or at least high-risk groups within that population.

Behind these messages and their circulation was a complex network of actors including government ministers, their scientific and political advisors, civil servants, journalists, and citizens. The reasoning used in these messages was also complex. In the example provided above, people were instructed to stay home not to protect themselves from infection (which might happen outside the home), nor to protect their loved ones from infection (which might happen outside the home and then within the home), nor even to protect strangers from infection (which might happen outside the home and result in direct hospitalisation or death). Rather, people were instructed to stay home to help lower the infection rate in society, which in turn would lower the hospitalisation rate, which in turn would lower the pressure on hospitals and health care services more generally, which in turn would lead to fewer preventable deaths (whether caused by COVID-19 or not).

During the pandemic, then, ordinary people were encouraged to think in terms of populations and (risk) groups; rates, trends, and distributions; the capacity of public services; and complex systems of causation. We might say that they were encouraged to see like an epidemiologist. We say this recognising that epidemiology is constructed and contested, and so has multiple gazes, but noting that classic epidemiological texts (e.g. Coggon, Rose, and Barker, 2003) tend to emphasise the study of how and why diseases occur; populations at risk, to which disease outcomes are related; groups of people, which are compared in terms of disease rates and levels of exposure; and disease distributions, trends, and changes, which epidemiologists monitor and surveil. Of course, this image of seeing like an epidemiologist takes inspiration from Scott’s Seeing Like a State (1998). We might also say that during the pandemic, ordinary people were encouraged to see like a state, to engage with ‘instruments of statecraft’ designed to make subjects and environments legible – surveys, maps, statistics, and so on – and, by doing so, to improve state capacity (which included their own capacity) and the possibility of state intervention (which included their own behaviour). More specifically, they were encouraged to see like a state looking through an ‘epidemiological risk’ rationality (Dean, 1999); looking at rates of morbidity and mortality in populations, and looking to achieve good health outcomes for populations by making risk calculations and controlling risks through public health interventions. In this respect, seeing like an epidemiologist may be thought of as one mode of seeing like a state. Furthermore, seeing like an epidemiologist, or a state looking through an epidemiological risk rationality, encourages a certain practical ethics – what Keane (2015) terms a ‘third-person stance’. So we might also say that during the pandemic, ordinary people were encouraged to see from a third-person stance; to leave their subjective, situated, emotional position where the ethical way forward might be to visit a loved one (the first-person stance), and adopt a position where the ethical way forward was to stay home for the benefit of the greater good (the main focus of the state or epidemiologist).

If this is one way the pandemic was governed, or governing of the pandemic was attempted – by trying to get ordinary people to see like epidemiologists – then it is worth asking how this played out. How did people respond to such encouragement? How did they use the statistics, charts, and maps provided to them? Did they come to think in terms of populations, groups, and risk? In the words of Montgomery and Engelmann (2020), did they become ‘epidemiological publics’ with ‘epidemiological imaginations’? Some commentators have assumed they did. For example, Bratton (2021b) defines ‘the epidemiological view of society’ as incorporating a view of each organism as a transmission medium, defined by who and what it is connected to; a view of risk as a plural and intersubjective set of relations, in which the individual may be of risk to the collective, while herd immunity reduces the individual’s risk; and a view of people as nodes in a biopolitical network to which they are responsible. At various points in The Revenge of the Real, he asserts that ‘people learned to see society as epidemiology does’ during the pandemic (ibid.: 33), that ‘we became amateur epidemiologists and learned to see ourselves accordingly’ (ibid.: 48), and that ‘the COVID-19 pandemic made everyone into an epidemiologist’ (ibid.: 164).

For Bratton, this ‘mainstreaming of the epidemiological view of society’ (2021b: 164) gives cause for optimism that people will accept the ‘positive biopower’ central to his ‘post-pandemic politics’. Bratton sets his optimism – that the reality of our technical and biological circumstances will be recognised, and sensing and modelling will be used to prevent and mitigate unjust deaths and misery – against Agamben’s denouncement of biopolitical responses to the pandemic as coercive, manipulative, totalitarian, despotic (Bratton, 2021a). However, empirical research on past engagements between epidemiology and publics gives cause for scepticism regarding both such (rather stylised) positions: Bratton’s ‘positive biopolitics’ and Agamben’s negative biopolitics. Using the case of public health in post-war Britain, Mold et al. (2019) show how epidemiological tools and interventions, from the epidemiological survey to mass vaccination, played important roles in imagining publics as populations, masses, and wholes (in addition to collections of groups and individuals). However, these publics often ‘spoke back’ to public health authorities. They resisted, complained, and reinterpreted or reappropriated public health recommendations and communications. They developed their own ‘lay epidemiology’ (Frankel, Davison, and Smith, 1991) – their own ‘folk knowledges’ about diseases, causes, and risks, based on ‘official’ medical and public health sources, but also mass media and the lived experiences of people they knew (including ‘folk figures’ like the elderly relative who smoked and drank heavily throughout their long and happy life).

The success of attempts to mobilise people against COVID-19, then, should not be assumed. Indeed, mobilisation for Callon (1986) is but one phase in a broader process of ‘translation’, which can fail at any phase. Problematisation can fail when actors refuse their defined identities and associations. Enrolment can fail when actors refuse their defined roles in the network, or refuse to coordinate in the network. Mobilisation can fail when actors resist displacement from one position to another. As Rose (1999) emphasises, translation – or ‘government at a distance’ (alignment of the objectives of authorities and the personal projects of the subjects of government) – is always contested, fragile, risky.

During the COVID-19 pandemic, such a process of translation was likely to be especially contested and fragile for at least two reasons. First, ‘the science’, which both governments and ordinary people were meant to be following, was uncertain and plural. According to Freedman’s (2020) reading of minutes from SAGE meetings in early 2020, ¹ this was a new and poorly understood virus, there were many evidence gaps, scientists in the UK were dependent on reports from China and Italy, and ‘the science’ was constantly revised in response to such reports. Over time, the scientific community in the UK and beyond split into factions (Horton, 2021). There was SAGE, but also Independent SAGE. ² There was the Great Barrington Declaration and the John Snow Memorandum. ³ Like governments, citizens could choose their science (ibid.). Boin, McConnell, and ‘t Hart (2021) describe a ‘framing contest’ in which different scientists, politicians, journalists, and others tried to frame the crisis, picking and choosing from ‘a cacophony of scientific advice’, most of which was also available to ordinary people.

A second reason why translation may have been difficult in these particular circumstances is that, ethically, pandemics are not straightforward events in which roles and responsibilities are clearly understood and agreed upon. An exchange in Camus’ The Plague illustrates this point (Camus, 2020[1947]: 67–71). ⁴ The argument is between Rieux, a doctor, and Rambert, a journalist who seeks permission to leave Oran and join his wife – who is not ‘strictly speaking’ his wife, but ‘the same thing’ – in Paris. In the argument, Rieux represents ‘abstract reason’ and is concerned with questions of ‘administration’, ‘regulation’, ‘law’, and ‘the general good’. By contrast, Rambert represents the ‘individual’ concerned with his own ‘happiness’, asking for ‘pity’ and to be treated as an exceptional case. Camus summarises the exchange – and the moral complexity of the situation it dramatises – in the following terms: ‘a struggle between the happiness of each individual and the abstractions of the plague’ (ibid.: 71). In pandemics, governments might encourage people to see like an epidemiologist (like Rieux), but people might defend their first-person stance on the grounds of love and individual happiness (like Rambert).

In the rest of this article, we consider how this process of translation – this attempt to get ordinary people to see like epidemiologists – played out in the UK during the first months of the COVID-19 pandemic (roughly spring and summer 2020 – see below). In doing so, we focus in particular on the uses of scientific literacy. Hoggart (1957) focused The Uses of Literacy on the period from the end of the First World War to the middle of the 20th century, when literacy improved among working people in Britain. He was concerned with how such improved literacy was used, especially in mass publications where elites encouraged and made appeals to working people. Inspired by Hoggart, in this article we focus on a moment of heightened scientific literacy prompted by the pandemic. Whereas Hoggart focused on the uses of literacy by elites, our focus is more on the uses of scientific literacy by ordinary people. What did they do with all the statistics, charts, maps, identities, and roles – ‘the science’, but especially the epidemiological way of seeing (focused on rates, trends, distributions, populations, risk groups, transmission) – they were offered and encouraged to use? We find answers to this question – and the broader question of how attempts to govern the pandemic through scientific literacy played out – in Mass Observation’s COVID-19 collections.

Mass Observation’s COVID-19 collections

Mass Observation (MO) has a long and fascinating history (Hinton, 2013; Hubble, 2010), but is probably best known for the material it collected on everyday life in Britain during the Second World War. Today, MO consists of the Mass Observation Archive (MOA), established in 1975 and located at the University of Sussex in Brighton; the Mass Observation Project, established in 1981 by the MOA, which sends ‘directives’ three times per year to a panel of around 500 volunteer writers from across the UK, asking panellists to write about particular topics for the Archive; and the 12 May project, established in 2010 by the MOA, which publishes annual calls asking anyone resident in the UK to keep a ‘day diary’ on 12 May for the Archive (inspired by the original MO’s first book, May the Twelfth – Jennings and Madge, 1937).

In 2020, MO asked panellists and diarists to write about life during the COVID-19 pandemic in its regular spring and summer directives, a special directive sent on 17 March at the start of the UK’s first lockdown, and the call for 12 May day diaries. ⁵ The writing sent to MO in response to these various calls makes up the first part of MO’s COVID-19 collections (with other parts to be added over time). In the rest of this article, we focus on this first part, which covers the period of the UK’s first lockdown (March to May 2020) and the ensuing period of opening up, before restrictions were brought back gradually from July 2020 (with variation between the devolved nations of the UK and also the regions within those nations).

What can writing for MO tell us about the uses of scientific literacy by ordinary people and, more broadly, attempts to govern the pandemic by constructing epidemiological publics? There is a large literature on how to approach and interpret writing for MO. One lesson from this literature is that writing for MO is dialogic. It is part of a dialogue with MO (Bloome, Sheridan, and Street, 1993; Salter, 2010), with writers best thought of not as respondents, but as correspondents (Sheridan, Street, and Bloome, 2000) who exchange writing with MO as part of an ongoing conversation with an imagined audience of archivists and future researchers (Kramer, 2014). Moreover, it provides access to a dialogue between writers and various authorities (Sheridan, Street, and Bloome, 2000) and ‘worlds of discourse’ (Hinton, 2010). Correspondents write about standards and codes, but also what they do with them, and their dis/comfort with them (Langhamer, 2016). They use sociological constructs, but do so selectively – interpreting, appropriating, and incorporating them (Wilson-Kovacs, 2014).

Writing for MO, we might say, provides evidence of ‘positioning’ (Davies and Harré, 1990). We see in the diaries and directive responses how people exercise choice in relation to categories, storylines, and subject positions – positioning themselves using the subject positions made available to them (as resources), recognising themselves as members of some categories and not others, accepting certain subject positions made available to them (but not others). In this view, MO’s COVID-19 collections look to be an ideal data set for investigating how attempts to get ordinary people to see like an epidemiologist – to recognise and accept certain categories and subject positions – played out during 2020.

Indeed, materials in the MOA have been used successfully for similar purposes before. Savage (2010) used them, alongside other materials, to study the ‘footprint’ of social science during its period of professionalisation between 1945 and 1970. He found that popular understandings of society changed during this period. Social science became embedded in everyday life and provided ordinary people with a language and imaginary of social group, relationship, and change. By 1970, in contrast to the late 1940s, many people were routinely and confidently doing ‘class-talk’ and positioning themselves – or knowingly refusing to position themselves – in relation to social classifications. Importantly, Savage notes that MO not only recorded this process (in writing it encouraged and then archived), but also participated in it, mobilising diarists and panellists, providing them with resources, as one of many research organisations active in the post-war period. This last point has been developed by Harrison (2014). The original MO around the time of the Second World War was a popular social movement that mobilised people by encouraging them to write, providing them with templates, and sending them examples of the writing of other correspondents and summaries of all writing received on a particular topic or period. In doing so, it provided people with models for self-observation, helped to normalise mass surveillance as a collective habit, and helped to develop new conceptions of population and culture (e.g. ‘mass’ and ‘morale’), which became new means of knowing and governing.

MO today is not quite the organisation studied by Savage and Harrison. Nevertheless, in 2020 the MOA did mobilise people to write about life during the COVID-19 pandemic. And it prompted people to write dialogically about authority (e.g. government regulations and advice), standards (e.g. of hand washing and social distancing), and subject positions (e.g. the family member with caring responsibilities, or the furloughed worker with time for new hobbies, to learn new skills, for more exercise). In response, the Archive received over 500 directive responses (average length: 11 pages) and over 5000 day diaries (average length: 2 pages).

The people who responded to MO’s invites and prompts are not formally representative of the UK’s population. Among correspondents with MO, certain groups have been over-represented consistently in recent decades, including women, the elderly, the middle classes, and residents of England (Hinton, 2013). Moreover, people who write for MO, regardless of their demographic profile, tend to be particularly dutiful and engaged citizens – almost by definition, since they are volunteering for a social history project (Hinton, 2010). On the other side of the balance, such people also tend to be particularly reflexive and critical regarding the world around them and their place within it (ibid.). They tend to be ordinary and to think of themselves as people without power, success, influence, or fame – people whose voices often get missed from official histories and characterisations of Britain, and delegates of such people (Sheridan, Street, and Bloome, 2000). Furthermore, given our purpose of investigating the uses of scientific literacy by ordinary people, a data set including the most engaged of such people should be viewed positively as an opportunity. Adapting Brenner (2003), MO correspondents are best approached partly as ‘stereotypes’ (‘typical cases’ of ordinary people) and partly as ‘archetypes’ (‘extreme cases’ of people who are particularly engaged with governing the pandemic through science and scientific literacy). As such, they constitute a data set well suited to the task of establishing the various ways ordinary people used ‘the science’ to participate in governing the pandemic.

Aware of the size and potential variety within the data set, we sampled 60 writers per collection – a number Clarke et al. (2018) found sufficient to provide both descriptive saturation and filled quotas for gender, age group, region of the UK, and occupational classification (see Appendix A). In total, we read 180 pieces of writing from 60 day diarists and 60 panellists (two directive responses per panellist). Together, these made up almost 800 sides of A4 (typed and single-spaced). We read this material for engagement with government messaging and science communication – so various forms of engagement with regulations and guidance; statistics, charts, and maps; the language and imaginary of epidemiology (populations, rates, trends, distributions, capacities, risk groups); and defined identities, roles, positions, responsibilities, and stances – working independently at first, then comparing notes, then, where agreement could be reached on interpretations (a version of intercoder reliability), sorting the material accordingly.

In summary, our methodology combined elements from the different methodologies debated in Twentieth Century British History’s recent round table on ‘historians’ uses of archived material from sociological research’ (see Hilliard, 2022). We selected a ‘critical case’ – a group who might be expected to be mobilised against COVID-19 – and carefully sampled and coded the writing of this group in the tradition of more quantitative sociological research (Goldthorpe, 2022). But we also drew from the tradition of more interpretivist historical research. Like Lawrence (2022), we sought to make full use of personal testimony, however it was collected. We sought to identify patterns in such testimony and to develop plausible interpretations for such patterns. Most importantly, given our focus on governing, we sought to make claims about individual subjectivities, whether typical or not, but also intersubjectivity (ibid.); to establish the subjectivities it was possible to inhabit and articulate in the context of the pandemic, and so to infer the context for those subjectivities.

Our focus on governing meant we read the material for one set of patterns – engagement with government messaging and science communication – when other readings may have been possible. For example, we did not prioritise reading the material for difference by gender, age, region, or occupational classification (a proxy for class in our sampling). These alternative readings did not impose themselves on us either, though we do note a relationship between age and vulnerability – how older people were positioned as ‘vulnerable’ and how they responded – in the next section. Given the reading we imposed on the material, and the readings it imposed (or failed to impose) on us, our approach in the rest of the article is to quote from a variety of diaries and responses. Here, we take inspiration from histories of MO (e.g. Highmore, 2002; Hubble, 2010; Jardine, 2018; Marcus, 2001). If the original M-O of the late 1930s was concerned a little with representativeness – the statistics of how material gets collected or sampled for analysis – it was concerned a lot with representation: the aesthetics of how material gets written up. Everyday life was depicted using compositional techniques. Ordinary people were given voice sometimes as images or close-ups, sometimes in juxtaposition, and sometimes as part of panoramas, montages, or collages. A list of correspondents given voice in the rest of this article is provided in Appendix B.

Following the science?

Most writers watched the daily briefings by ministers and their scientific advisors, or at least the highlights of them, at least when they began in March 2020. Some went so far as to look at the statistics, keep their own data sets, produce their own charts. A woman in her 60s from Northumberland (K7522) wrote: ‘I follow the daily briefings and listen to the relevant minister and the science and medical reps and then I look at the statistics.’ A retired local government officer in his 70s from Leicestershire (L6048) wrote: ‘Every day I watched the press briefing broadcast from Downing Street. I keep statistics so I can have my own dataset with which to examine the disease in this country.’ A woman in her 60s from Scotland (H6675) reported: ‘Husband used information from BBC website to make his own graphs of infections and deaths in UK and in Scotland.’

While only a few correspondents went so far as to produce their own charts, many used the language of charts in their diaries and directive responses (curves, spikes, etc.) and the language of epidemiology more generally (e.g. ‘the R rate’). A secretary in her 70s from Dorset (P1796) reported: ‘The rise in deaths in the UK has continued, it was nearly 1000 yesterday. Let’s hope this is the start of the top of the curve.’ At the end of the first lockdown, when people were being encouraged back to work, a business executive in his 20s from London (O7365) wrote: ‘I’m not convinced social distancing can properly be observed in industries like construction and manufacturing, while not everyone has the luxury of being able to avoid public transport. As such, I think there could yet be a further spike that drives the R-rate above 1.’ In their diary of 12 May 2020, looking forward to 12 May 2021, a student in London (MT 2020 37) wrote: ‘I do not think that life will simply be back to the way it was before this virus, as there may not be a vaccine by this point. I dearly hope that by this point the “R” rate will have decreased enough so that large gatherings will be allowed.’

Many writers used this language of rates, curves, and spikes – and seemed confident and comfortable in doing so. Could the same be said for the subject positions offered by epidemiology during these early months of the pandemic? Correspondents certainly used terms like ‘vulnerable’ and ‘high/low risk’. ‘Mum is classed as vulnerable’, wrote a designer in her 30s from London (MT 2020 200). A retired banker in his 70s from West Sussex (S3035) reported: ‘As I am deemed to be a vulnerable person (through age and health), my wife has been doing most of our shopping.’ Similarly, a woman in her 60s from Lancashire (A7000) reported: ‘I am deemed “extremely vulnerable” and so have been shielding from the end of March until the end of July.’

Using such terms, however, is not the same as fully accepting their referents (in these cases, the subject position of the vulnerable person). In the previous paragraph, the last correspondent puts ‘extremely vulnerable’ in scare quotes. None of these writers says, ‘I’m vulnerable’ or ‘Mum is vulnerable’. They say ‘deemed’ or ‘classed as’ vulnerable. They are aware that epidemiologists, or those advised by epidemiologists, are positioning people as vulnerable. They are not quite sure whether to accept such a position for themselves or those they care for – though in some cases the awareness itself appears to have led to action (I’m deemed vulnerable so ‘my wife has been doing most of the shopping’, or I’m deemed vulnerable ‘so have been shielding’).

If these writers appeared unsure regarding such classifications, there were other correspondents who seemed more sure (that such classifications should not apply to them). Responses among these writers varied. An artist in his 80s from Yorkshire (P3209) reported: ‘I’m regarded as a member of the “at risk” group because I’m 80’ (note the ‘regarded as’ and scare quotes formulation). He went on to assert: ‘My health is very good, as is that of my wife.’ He reported compliance with government advice for members of at risk groups (‘We have adhered to the advice to stay at home’), but also frustration: ‘We feel frustrated if only because we are fit enough to lead active lives and would have liked to help others in our area.’ So here we have someone who contested their designation (‘at risk’) and felt frustrated by it (given their own assessment of their own health), but nevertheless adhered to the advice (if only reluctantly). A similar response was provided by a teacher in his 70s from Hampshire (W2322). He described himself and his wife as ‘both in our 70s, and in good health, physically and mentally’. He followed government policy and self-isolated – as an ‘older’ person, ‘currently defined as over 70’ – but felt ‘irrelevant’, ‘neglected’, ‘pushed to one side’, ‘locked away’, ‘forgotten’. He wrote that he ‘hated being referred to as “vulnerable”, since I am able to keep myself physically and mentally fit, and the label, constantly used, simply disabled me’.

This correspondent expressed the now familiar scepticism (with scare quotes around ‘vulnerable’, which he referred to as a ‘label’), but also strongly felt discontent (not just frustration, but ‘hatred’) – yet he still followed the guidance for his group and self-isolated. There were some writers, however, for whom contestation over terms led to rejection of positions and, ultimately, behavioural consequences. An example is the journalist in her 70s from Durham (W633) who wrote at length about ‘the over 70s classification’, which she saw as ‘a sweeping, ageist category’. She wrote: ‘Most of the over 70s I know are fitter than many younger people. Categorising by sex, income, ethnicity etc. would be regarded as offensive. So is ageism.’ After a government press conference in mid May, she continued: ‘The Prime Minister has spoken.… The over 70s are still expected to stay at home. I have emailed my MP.’ She argued: ‘The over 70s with no underlying conditions … are no more at risk than an obese 50 year old.’ Then she reported: ‘I have made three masked and gloved short 8am supermarket trips since lockdown two months ago … and we have taken a brisk walk every evening of between 1½ and three miles, carefully social distancing from the few we pass.’ These trips out were ‘probably against the “rules”. We would tell any politician who asked that the fit over 70s were probably no greater risk than the obese 50-somethings and they were not being told to stay in all the time (ageism is OK, fattism is not).’

Here, we have someone who found the ‘classification’ or ‘category’ they were put in during the pandemic ‘offensive’ for the way it reduced them to their age. They contested the logic used to justify this classification. They rejected the position they were given (the vulnerable, at-risk ‘over 70s’) and broke what they understood as the ‘stay home’ rule for people in such a position (while nevertheless taking other precautions: wearing masks and gloves, selecting quiet times for trips out, keeping their distance from other people). So we have evidence that at least some people used their scientific literacy – e.g. their understanding of risk as it relates to different groups (‘the fit over 70s’, ‘the obese 50-somethings’) – to contest and ultimately reject the identities and roles assigned to them by epidemiology during the pandemic, with behavioural consequences and so consequences for governing the pandemic through scientific literacy.

We also have evidence that at least some people struggled to see from a third-person stance, as the epidemiological lens encourages people to do. Epidemiology, at least when translated into a risk rationality of the state, asks people to think abstractly in terms of populations and groups, and to act for the greater good (as opposed to individual happiness, which might be the concern of situated, exceptional, emotional subjects). Some writers found this difficult, or reported that others close to them found it difficult. Indeed, from particular situations, some of the rules and guidance seemed absurd. In his 12 May diary, a retired teacher in his 60s from Somerset (MT 2020 119) identified what he called ‘anomalies’ and gave the following example: ‘From 13 May we will be able to meet with one person from another household in a public space at a distance of two metres. So our daughters can only meet with one of their parents at a time. Yet, should we be employed as their cleaner we can meet with them in their house.’ In his lockdown diary, submitted in response to MO’s Special and Spring directives, an air traffic planner in his 50s from Hampshire (D4736) reported: ‘Our vicar is still spitting communion wafers at not being able to go into the church to pray, but he has found a way around that. He is allowed to go in as “janitor” to check the pipes. So he does, regularly.’

In both these cases, rules that make sense when seeing like an epidemiologist seem absurd from the perspective of the individual (situated and seeing from a first-person stance). In the latter case, this absurdity – or at least the many alternative moral reasons presumably available to the vicar – appeared to justify breaking the rules. Whether breaking the rules could be justified from a first-person stance was discussed at length by a designer in her 30s from Oxfordshire (T5903). She had just had a baby and her mother wished to visit, care for her daughter, and see her new granddaughter. The panellist wrote: ‘For some reason *she* should be the exception to the literal law about social distancing.’ Later, she continued:

Mother still complaining about being a stranger to our … baby and asking when she will be allowed to see her … My mother asked whether she could drive over here for a visit under the guise of needing to take me to the GP which would constitute providing care. I said NO! I do not want her to come! Lockdown means lockdown! I’m not trying to find ways around it.

And later: ‘They [mother and father] asked in our latest video call when they could come over.’ And later: ‘My mother is still whining about not being allowed to visit and trying to come up with reasons why she should be allowed an exception.’

This writer’s submission to MO recalls the argument staged by Camus in The Plague and discussed previously. The daughter could be Rieux, concerned about the law – the ‘literal law’ – and seeing from a third-person stance (though also, perhaps, keen to avoid her mother for other reasons – reasons we can only guess at). The mother could be Rambert, concerned about her family, her loved ones, and perhaps her own individual happiness; seeing from a first-person stance and requesting an exception for her particular situation.

Looking across all the submissions discussed so far, we find plenty of engagement with government messaging and ‘the science’, and plenty of scientific literacy – all of which is perhaps unsurprising, given the (engaged, literate) character of most MO correspondents. However, we also find plenty of variety in how such engagement and literacy was used. Sometimes, it was used to see like an epidemiologist. People used the language of rates, curves, spikes, and so on – doing so confidently and comfortably in most cases. Sometimes, though, it was used to contest such a way of seeing. Subject positions – for example, member of a vulnerable or at-risk group – were treated with scepticism or just refused. Regulations and guidance were contested using alternative risk calculations or ethical stances (especially the first-person stance). Sometimes, but not always, behavioural consequences followed (in the form of rule-breaking). We now turn to another way that scientific literacy was used by ordinary people in the first months of the pandemic: to judge government performance.

Lay comparative analysis

For many correspondents, ‘the science’ was plural and they found resources for engaging with government messaging and generally making sense of the pandemic in multiple locations. A business executive in his 20s from London (O7365) wrote: ‘The UK has consistently disregarded the advice of the WHO [World Health Organization] in terms of lock downs, testing and contact tracing. Time and again, their advice has proven correct and we have found ourselves playing catch up on these requirements.’ A teacher in his 70s from Hampshire (W2322) wrote: ‘I’ve been very interested in international comparisons so I’ve made regular use of the Johns Hopkins Coronavirus Resource Center website.’ On the basis of these resources, he concluded: ‘Britain has handled the pandemic disastrously, with only a little more honour than the USA … We were ill-prepared, and did not take action at once, which would have saved lives … We have become the laughing-stock of Europe, certainly compared with the determination and prompt action shown by Germany.’

We’ll return to the figures of Europe and the USA shortly. First, there is more to say about multiple sources of scientific advice and information. One panellist who responded to MO’s spring and summer directives with a diary covering those months was a scientist in his 70s from North-West England (M3231). While not a typical case, he was situated particularly well to notice and reflect on ‘the science’ available to ordinary people during the pandemic. He wrote of ‘an email from my friend in Sweden’ that ‘included an article on the Guardian website’ that ‘included criticisms of the Swedish approach from Swedish academic epidemiologists’. He looked ‘on the BBC news website to check the number of new deaths from the virus’. He ‘read an article which states that the rate of infection is getting close to one, which could lead to an increase in cases’, and wrote: ‘I do believe that the government lifted the lock down in England too soon.’ On another occasion, he looked ‘at the number of corona deaths for the day’, then ‘heard a news item on BBC Radio 4’s news programme that said the R₀ number is about 1 in the North West’, then commented: ‘As I live in the North West this does not encourage me to dash out to any of the shops that are open, or going to open soon.’ On another occasion, he reported:

Last night I watched a C4 [Channel 4] programme on the corona virus which had been broadcast a week previously. There were a number of scientists being interviewed and what became apparent was they were well aware of the dangers of the virus, and that the importance of a lock down was recognised, and recommended, well before it was instituted. The conclusion drawn was that many more lives could have been saved if the lock down had been started earlier. It does seem that the government was slow or reluctant to act.

On another occasion, he reported listening to ‘the 6 o’clock news on Radio 4 as I prepare a meal’ and hearing ‘a report on a comparison of the death rates from the corona virus between this country and others. The unsurprising conclusion was that this country had the highest number of deaths from the virus.’ He also ‘listened to two BBC podcasts’ that ‘examined claims about the corona virus’. He also ‘read in the paper … that 6% of the population have had the corona virus’. On yet another occasion, he ‘was listening to the news on BBC Radio 4’ and heard ‘a man being interviewed, and discussing testing for the corona virus, and the importance of false positives and false negatives and the need for multiple testing for accuracy’. He reflected: ‘I didn’t ever think that I would hear such a topic discussed on a news programme.’

In the previous paragraph, we see multiple sources of scientific information and knowledge: newspaper articles, websites, radio programmes, television programmes, podcasts, emails from friends. These mediate between ordinary people and multiple epidemiologists and other scientists (not just government advisors providing the official line). They provide concepts and terminology (‘false positives and false negatives’). They provide statistics including infection rates by locality and mortality rates by country. They allow people to make judgements about their own behaviour (‘this does not encourage me to dash out to any of the shops that are open’) and the UK government’s response to the pandemic (‘the government lifted the lock down in England too soon’).

This last point is important. For many correspondents, the science communication they received, especially from sources beyond ministers and government advisors, allowed them to judge their government’s response to the pandemic by comparing it with other responses by other governments around the world. One popular storyline, repeated frequently across the submissions to MO, was that other countries like Germany were doing better, while the ‘UK now has the highest number of deaths in Europe’ (H6004); ‘Our death rate is the worst in Europe’ (A7000); ‘We have the highest death rate in Europe’ (M4463). A similar storyline, just as common, was that other countries like New Zealand were doing better, while ‘we are looking at one of the highest death rates in the world’ (M6737); we are ‘the country with the worst death toll apart from the USA’ (T6654); ‘we have the highest death rate in the world, after the US’ (P7209).

These last two panellists mentioned the performance of the USA, and some panellists did use the figure of the USA to cast the UK’s performance during 2020 in a better light. A lecturer in his 30s from Surrey (K7050), who found himself ‘quite sympathetic to the government’ because ‘the situation was new and very difficult’, wrote: ‘In comparison to the buffoon Trump in the US, Johnson was at least listening to the scientific advice.’ Comparing the UK government’s daily press conferences to those of Trump, a retired local government officer in his 70s from Leicestershire (L6048) wrote: ‘One sees straight away how more superior was our government’s approach.’ An administrator in her 70s (L7499) favourably compared ‘Boris’ (‘sensibly taking advice’) with ‘idiotic Trump who just bullies everyone, ignores their advice and does what he wants’. Many writers, however, did not regard the USA as a standard for how to govern the pandemic. Like this writer in her 40s from North Ayrshire (M5645), whose father argued that ‘Johnson was bad but not as bad as Trump’, many correspondents viewed the performance of the USA as ‘a pretty low bar’. By contrast, for this panellist, commenting on reports from New Zealand, ‘Jacinda Ardern should be made Queen of the world’.

Whatever standards people used, and whatever conclusions they reached, one thing seems clear from the last few paragraphs. During the pandemic, ordinary people were provided with scientific resources allowing them to make international comparisons and use their scientific literacy to judge the performance of their own government in comparative terms. This was not perhaps what governments had in mind when they encouraged citizens to participate in governing the pandemic by seeing like an epidemiologist. To this extent, we agree with Krastev (2020) that COVID-19 empowered governments by making people dependent on them for help, but also empowered citizens by subjecting governments around the world to similar metrics, allowing performance to be measured and compared.

Doing like an epidemiologist?

Our starting point for this article was an observation: during the first few months of the COVID-19 pandemic, before pharmaceutical interventions became widely available, the UK – like many other countries around the world – sought to govern the pandemic by getting ordinary people to see like epidemiologists (to think in terms of populations, groups, rates, trends, distributions, risk) or to see like a state (to think in terms of populations, state capacity, the general good). We wanted to know how this attempt to govern the pandemic through scientific literacy played out. Was translation successful? Did people become ‘epidemiological publics’ with ‘epidemiological imaginations’ (Montgomery and Engelmann, 2020)? Did ‘the epidemiological view of society’ become ‘mainstreamed’ so that we all became ‘amateur epidemiologists’ (Bratton, 2021b)? To answer these top-level questions, we asked a series of more modest questions appropriate to the diaries and other writing collected by MO during spring and summer 2020. How did ordinary people use their scientific literacy? What science did they follow, how, and to what effect? In what ways did they respond to provision by various actors – government, science, media – of resources for making sense of the pandemic and their role within it? These resources included regulations and guidance; statistics, charts, and maps; the language and imaginary of epidemiology; and defined identities, roles, positions, responsibilities, and stances.

Reading materials in the MOA, we found at least five things. First, there were many opportunities for people to engage with ‘the science’, including government press conferences, newspaper articles, radio and television programmes, podcasts, and many other websites. Second, this engagement was with science plural, i.e. the science of government advisors, but also other scientists and scientific authorities. Third, for many people this engagement led to seemingly confident and comfortable use of epidemiological terms and concepts (rates, curves, spikes, etc.). However, fourth, engagement with some of the subject positions offered by epidemiology tended to be less comfortable. Engagement with positions like ‘vulnerable’ or ‘at risk’ was frequently sceptical and reluctant. Sometimes these subject positions were refused, with consequences for compliance with government regulations and advice. Importantly, such refusal, reluctance, or scepticism did not appear to result from ignorance, lack of information, or scientific illiteracy. Rather, they appeared to result from scientific literacy, with people mobilising alternative data and logic (e.g. alternative risk calculations). And they appeared to result from a kind of moral literacy too, with people mobilising alternative models of practical ethics (e.g. Keane’s first-person stance). Finally, we found that many people used their literacy and the resources made available to them during the pandemic – especially statistics – in a way perhaps not intended or anticipated by governments: to compare and judge government performance (and, in the case of the UK during 2020, to judge performance as comparatively poor).

In conclusion, we might say that governing the pandemic through scientific literacy was partially successful, but in some unexpected ways. Multiple scientific resources were used in multiple ways. They were used to justify compliance, thus contributing to governing of the pandemic. But they were also used to refuse or at least contest subjectification, making governing the pandemic more difficult. Then they were used to hold governments to account – measuring their performance, comparing it, judging it – which in theory should have improved governing of the pandemic. This last point is especially important because governing behaviour, which is central to most non-pharmaceutical interventions, rarely works just through information provision, messaging, and subjectification. It depends on infrastructures (Barnett et al., 2011) or systems (Drury, Reicher, and Stott, 2020; Reicher, Drury, and Stott, 2020), which provide opportunities and constraints for selected practices, and so determine capacity to act. Governing behaviour, then, depends on effective governments and their regulations, policies, and investments, which in turn should all benefit from improved accountability.

A different way of putting this is to say that governing pandemics involves getting ordinary people not only to see like an epidemiologist, but also to do like an epidemiologist (or at least to do what is recommended by epidemiology, or required by a state looking through an epidemiological risk rationality). Practically speaking, what were people being asked to do when they were asked to social distance, self-isolate, or practice good hygiene? What opportunities were they given? What constraints did they face? Was their capacity to act supported by infrastructures and systems? These are another set of questions that researchers could ask of MO’s COVID-19 collections.

Returning to MO’s collections brings us to some final thoughts on this archive, historical approaches to the COVID-19 pandemic, and our title: ‘Seeing like an epidemiologist’. Using these collections encouraged us to consider previous uses of the MOA and previous histories of MO, including characterisations of materials in the Archive as intersubjective and dialogic; uses of these materials to study top-down and bottom-up processes; and uses of these materials to study how ordinary people respond to social scientific constructs and, more broadly, to study what Savage (2010) termed ‘the footprint of social science’ in the 20th century. These uses and histories in turn encouraged us to approach the COVID-19 pandemic in particular ways: as an ‘epidemiological moment’ when epidemiology may leave its footprint on the early 21st century; as a moment both dependent on and contributing to heightened scientific literacy among ordinary people (not least via the charts and maps of government press conferences and associated media coverage); and as an emergency potentially governable through scientific literacy, but only potentially governable in this way. As we have seen, literacy is plural. Scientific literacy exists alongside and is sometimes confronted by moral literacy. Furthermore, literacies have multiple uses. We saw this demonstrated in the 20th century by Camus on the moral complexity of plagues, Hoggart on the uses of literacy by elites, and Frankel, Davison, and Smith on lay epidemiology.

These ways of approach in turn encouraged our theme of seeing like an epidemiologist. During the first year of the COVID-19 pandemic, before mass vaccination had been achieved, ministers and their networks of advisors, civil servants, and journalists attempted to govern the pandemic by getting ordinary people to focus on rates, trends, distributions, populations, risk groups, transmission, and their own position and agency in this vision. The historical context for such a project was that, for much of the last century, government had been understood in terms of risk (Dean, 1999). One part of this government through risk had been the epidemiological risk rationality, whereby risk is made calculable and governable by morbidity and mortality rates, and public health interventions like sanitation and inoculation programmes. In recent decades, governments had sought to mobilise relatively high levels of capacity among ordinary people – including relatively high levels of scientific literacy – and so to supplement and partially replace social government with advanced liberal governmentality (ibid.). In advanced liberalism, questions of individual autonomy exist alongside questions of personal responsibility and the role of individuals in governing emergencies. This was the historical context. Our article has focused on what resulted from this project of encouraging ordinary people to see like epidemiologists. What resulted was a lay epidemiology that contributed to governance of the pandemic in complex ways. Seemingly confident and comfortable use of epidemiological terms and concepts was accompanied by sceptical and reluctant engagement with (or refusal of) epidemiological subject positions, and also judgement of national governments by way of lay comparative analysis. This lay epidemiology could have important behavioural, democratic, and other consequences in the coming years, as the COVID-19 pandemic proceeds and is joined by other emergencies.

Note on authorship

Nick Clarke (principal investigator) and Clive Barnett (co-investigator) acquired the funding and played an equal role in conceptualising the research, developing the methodology, and collecting and analysing the data. They developed a plan for the article together, including an outline of the overall approach and argument. Nick led on writing up a first draft of the article from this plan. Clive died in December 2021 before he was able to respond to this draft. The article was completed by Nick, who submitted the draft to History of the Human Sciences and revised it in response to helpful comments from two anonymous reviewers and the editors.