A brief framing note: in the current emotionally charged context it’s probably wise to first articulate some framing thoughts and try to avoid potential misinterpretations. It’s clear that SARS-CoV-2 and the disease it causes are significant threats. It would be a grave error and immoral to ignore its effects on the health and lives of so many people. So, what follows doesn’t ignore this. Moreover, like most people, I’m extremely concerned. However, I also know that it’s also a threat we still have limited knowledge of, that it’s a rapidly changing situation, and that we’re constantly learning (and unlearning) new things about this virus: we need to be aware of the limitations of available data and existing knowledge and seek to embody an appropriate level of epistemic humility. I also try to keep this in mind. If we consider earlier pandemics it typically took years to get a good understanding of their dangers (e.g. a solid estimate of the fatality rate). Similarly, as Tim Harford recently noted, “if the scale of our ignorance about coronavirus may seem hard to swallow, bear in mind that the fatality rate for the H1N1 swine flu pandemic in 2009 was still being debated years later”. A recent episode of the Munk Debates podcast offers what seems to be a good primer on what is and isn’t being debated about the pandemic. However, despite this context, I recognise that governments need to act to address a new health threat (I support this). Social scientists may be told to “stay in their lane” and not comment on such things (i.e. leave it to relevant experts), but t here likely are aspects that social scientists can comment on whilst recognising the importance of responding… (The post begins below).
This is a weird, slightly surreal time. How are we to make sense of the current coronavirus pandemic event, or to consider (or analyse) responses to it? What perspectives are relevant to considering a major novel threat like a new infectious disease and responses to such threats? How are we to cope with the torrent of information (the “infodemic”) coming at us on a daily and sometimes hourly basis and not get overwhelmed? As we engage in various forms of social distancing of unknown duration, and deal with other aspects of the situation, many folk (including me) have much more time to contemplate such questions.
I have found myself flip-flopping somewhat between different perspectives or ‘lenses’. One perspective is the instinctive fear response – particularly regarding the potential dangers to people in at-risk groups like my parents, though we see stories of people of various ages needing medical care and, in some cases, dying. Health systems are under enormous pressure in many countries and seemingly every day there are worse stories from all over the world. Faced with a novel threat like the SARS-CoV-2 virus it’s perhaps rational to fear the worst until proven otherwise and to act accordingly. That’s one lens. I also find myself shifting to other lenses or perspectives shaped by my exposure to critical social theory, and the diverse studies I’ve completed over the past decade or so in the sociology and philosophy of science. I find that this study makes it impossible for me to ‘uncritically’ read the news and follow government directives, though – like most people – I’m washing my hands much more frequently, keeping the stipulated distance from other people in public, and generally following the rules. I also find myself becoming a pseudo-anthropologist – watching, and being fascinated by, the behaviour of other people. I watch in amazement as many people who might, for example, normally have a much more ‘skeptical’ orientation towards things like modelling exercises and the future projections that they produce suddenly react intensely to the latest projections as if the models that produce them are “truth machines” and modellers can reliably predict the future (e.g. see this article in The New York Times, or see the reaction of American journalists to the latest epidemic projections).
In such a context complete clarity is probably an unrealistic ideal. As one writer at The Atlantic puts it, we’re in the “fog of pandemic” – a situation analogous to the “fog of war” where “militaries do not fully understand either their enemy’s threat or their own capacity to combat it”. As many others have suggested, it’s probably wise to consider how cognitive biases could be shaping our views under such uncertain and dynamic conditions, to evaluate the quality and reliability of sources of information, and be wary of misinformation. This is good advice, however I also know that we live in a time of stretched newsrooms and pressures on journalists that mean that even high-quality news outlets may be publishing poor quality information (at least occasionally). I also know from my own studies of scientific controversies and failures that even prestigious news organisations publish material that turns out to be misleading and sometimes entirely false – particularly in the “heat of the moment” when deadlines are tight, emotions are intense, and data is unreliable or still emerging regarding the focal topic. Journalists are human beings like anyone else, even if their training is intended to help them to navigate such moments.
So, given this context and my own somewhat chaotic flip-flopping between different lenses or perspectives, what else can or should be said about the present situation? I have to say that I find myself doubting that I have anything useful to contribute or say. My own areas of partial expertise aren’t directly relevant to the study and management of pandemics. Moreover, if my interactions with others are any guide many people currently find ‘critical’ takes on the present crisis to be either a distraction or unhelpful (at best), and view the only kind of valid response to COVID-19 to be those that help to provide solutions.
Nonetheless, I’ve compiled a few early thoughts on the coronavirus pandemic:
1. A fast-moving pandemic situation reveals (or reinforces) the importance of scrutinising, and reflecting upon, knowledge practices: Though we still have partial evidence, which likely suffers from multiple biases, the SAR-CoV-2 pandemic is data heavy. Every day we’re shown case counts and body counts (deaths), as various statistics are tracked, and other disease-related statistics for fatality rates are frequently reported. My local newspaper, The Age, newspaper even notifies me (via ‘push’ notification on my smart phone) of every death in Victoria. It’s also analysis heavily. For instance, seemingly every day we’re shown new projections which decisions-makers are expected to act upon. However, in the “fog of pandemic” it’s often not clear what the available data and evidence means and we need to be wary of reaching strong conclusion on partial and/or biased data. Even seemingly straightforward statistics like death counts can be misleading (e.g. link, link), with medical specialists being asked to determine and report the cause of death under rushed and difficult circumstances. Broader challenges and dynamics were reported by Derek Thompson in his piece in The Atlantic: “The officials tracking COVID-19 are swimming in statistics: infection rates, case-fatality ratios, economic data. But in these early stages of the fight against the coronavirus, these figures each have their own particular limitations. We are already seeing how, in the haze of confusing data, political leaders are trying to marshal that uncertainty to override the advice of public-health experts. Indeed, President Donald Trump seems eager to seize on anything that can justify his push to reopen public life in mid-April”.
The quote above points to two key things: 1) the question of how actors appraise ambiguous evidence and deal with the “infodemic” (as well as the pandemic itself), including how they reach judgments about the quality and significance of evidence; and 2) the way that actors mobilise such evidence to justify their policy preferences and/or decisions.
It also points to the importance of the social context in which such practices are enacted. A key objective for producing the latest scary projections in the US appears to have been trying to persuade President Trump to adopt a different view, though the American media ignores such contextual factors in both their questions and reporting.
We also see some interesting practices emerging in this pandemic, such as rushing to get modelling to key decision-makers before it’s been scrutinised in the usual ways (e.g. peer review by relevant academics/specialists). This can be seen as required in the context, but it potentially is also a highly consequential deviation from established norms. Thus, this crisis is also a context in which to observe and scrutinise practice innovations.
2. Pandemic management as the attempted creation of ‘self-denying prophecies’ and related performative acts: what I mean by this is the basic observation that we see the prominent articulation of future projections (usually based on modelling exercises) as a means of stimulating behaviour that seeks to prevent such scenarios from ever coming to pass. If this is successful – i.e. if it stimulates behavioural change – then the prediction (broadly speaking) won’t come to pass. On the one hand, this is understandable and reasonable. On the other hand, there can be clear incentives to paint exaggerated pictures of the future which may not be realistic (or are extremely unlikely to occur even if mitigating action wasn’t taken). Scientists risk being accused of fear-mongering, as some climate scientists have been, for using scenarios which are alleged to be unrealistic worst case scenarios. Perhaps this is reasonable – even unavoidable – in order to inform and/or motivate timely action. But the experience in other fields suggests a need for scientists and others to be wary given the fraught relation between science and politics in the early 21st century.
Additionally, we perhaps shouldn’t ignore the views of well-credential epidemiologists who note that “in the past most claims for major epidemics that would be devastating […] were not validated” (link). I take such statements to be well-intentioned efforts to encourage good, evidence-based decision-making that tries to avoids panic, though others might understandably see them differently in the present context. Indeed, sometimes mitigating actions will be what prevents worst case scenarios from coming to pass; in other cases they won’t occur because early claims and fears about a novel pathogen turn out to be incorrect or exaggerated (to some extent). In the latter case, there is also the risk of impairing the public’s confidence in science and adding more fuel to related fires.
3. Considering and alternating between ‘objectivist’ and ‘constructionist’ perspectives on the coronavirus crisis: I find it useful to consider both of these perspectives and try to alternate between them (the terminology comes from social problems theory). The objectivist viewpoint is that the ‘COVID-19 crisis’ is objectively real, having a fundamental reality which is independent of our thoughts and discourses about it. It unambiguously and accurately reflects something real in the world which is objectively dangerous – a harmful new pathogen (the SARS-CoV-2 virus), with specific harmful properties and consequences, which we are measuring, tracking and managing. In contrast, a constructionist viewpoint takes seriously the potential ways in which the COVID-19 crisis may be a socially constructed phenomenon. This viewpoint doesn’t simplistically suggest that the pandemic is merely an irrational concern – a baseless moment of hysteria, or something like that. But adopting such a viewpoint can entail carefully locating claims and other behaviour within social context(s), and considering how this might influence both our understandings of the COVID-19 crisis in consequential ways (e.g. having the effect of amplifying or downplaying the threat) and our responses to it. At their most ‘extreme’, the constructionist viewpoint argues that social problems (like the COVID-19 crisis) are, in essence, what people think are problems, rather than some condition and/or phenomena that is objectively harmful (but, again, I’m not suggesting that this more extreme version should be adopted here).
The objectivist viewpoint is essentially what we see expressed in the news every day, with little critical interpretation of the statistics and other evidence that is being reported.
The ‘constructionist’ viewpoint is rarer (outside of social scientific contexts) and more difficult to consider. If this viewpoint seems hard to conceptualise here is a relevant assessment articulated by a retired British medical specialist: “Much of the response to COVID-19 seems explained by the fact that we are watching this virus in a way that no virus has been watched before. The scenes from the Italian hospitals have been shocking, and make for grim television. But television is not science” (link). I’m not saying this view is correct, but it moves analysis and thinking in a constructionist direction.
Another way that you can begin to adopt a more constructionist viewpoint is to consider what Joel Best terms the social problems process: the processes and day-to-day activities of asserting, claiming and alleging, etc engaged in by actors who are involved in an issue coming to public attention (or not) and then being responded to (or not). Best argues that an issue must be socially constructed in order for it to be addressed – moreover, sociologists argue that both valid and invalid social problem claim have to be constructed (Yearley, 1992).
Some sociologists have also adopted such a perspective when studying and explaining policy diffusion (e.g. the global diffusion of approaches used by national governments to slow transmission and/or contain the virus). of COVID-19). These sociologists argue that what’s considered to be legitimate ends and appropriate means (to achieve them) are shared social constructs which are shaped by causal mechanisms such as mimicking other countries (e.g. following those perceived to be the leaders) and expert theorisation of the effects of policy solutions (either actual or predicted effects), and they often ground their arguments in the ‘bounded-ness’ of policy-maker rationality. Consideration of this perspective encourages a different view of policy diffusion that suggests its less driven by the impartial consideration of empirical evidence and highly influenced by social causes. Indeed in the case of COVID-19 policy-makers are acting under conditions of considerable uncertainty.
4. The coronavirus case reminds us of the importance of a key social process: the resolution of contingency (and uncertainty) through ideational and other social constructs. Think about the ideational constructs of partial or full ‘lockdown’ (widely expected to be in place for many months): widespread convergence on these constructs as the correct response, or perhaps the only possible response, appeared to quickly remove any sense of contingency regarding potential responses to the crisis. Here in Victoria, Australia, where I live progressively stricter police-enforced forms of ‘lockdown’ have been implemented. Craig Parsons has analysed how such constructs often act as interpretive filters through which people perceive, think and feel, and I think we can see this in the case of the coronavirus case. This is important for at least a few reasons: such resolution of contingency may happen somewhat unconsciously (the chosen course of action comes to seem like the only possible course or the ‘natural’ way to do something); it’s often the case that more options were available than were seriously considered; and, third, existing social constructs can be a barrier to subsequent innovation and agency. Additionally, some infectious disease experts argue that some of the strict police-enforced restrictions which are currently in place in some locations (e.g. where I live in Melbourne, Australia) “appear more as a result of panicked political decisions rather than based on biological plausibility or evidence” (link).
In the fullness of time we’ll have a better idea of the wisdom of current government actions (e.g. whether they were a rational response to material conditions). Some people demand more strict forms of ‘lockdown’; others point to contrasting alternatives.
5. Finally, if we are going to consider sociological perspectives then it could be useful to consider a sociology of scientific knowledge perspective: In his recent book Forms of Life: The Method and Meaning of Sociology Harry Collins provocatively suggests that one should study scientists’ beliefs and practices just like one might study religious beliefs. His approach – which he terms methodological relativism – is outlined as follows:
In studies of science, methodological relativism demands that the world is treated as though it in no way affects what scientists come to believe about it. It is understood best by thinking of the counterpart in studies of religion, where it is obviously the right approach. Suppose you wanted to do research on why Southern Irish Catholics believe the wine in the Mass turns into blood while Northern Irish Protestants believe it is only symbolic. The last thing you would want to bring into the explanation is the question of whether the wine actually turns into blood!
[…] you had better be a relativist about it and accept that whether it turns or not does not affect people’s beliefs about whether it turns or not […]. The right approach is to work as though the belief determines what comes to count as truth, not that the truth determines belief” (pp.66-7).
Collins goes even further: the sociologist should “ignore what nature may or may not be telling” scientists (p.67), and focus solely on the social causes of what counts as truth.
One reason this social scientific perspective is potentially interesting is because currently, for most questions about COVID-19, no one knows the absolute truth. At best, for some aspects estimates and hypothesised ranges have been generated (e.g. for key variables such as how contagious or deadly this disease is) – but there’s limited confidence in current numbers. Regarding other aspects, scientists are basically guessing. So, Collins would probably say that if no one knows the truth then it’s not determining belief.
I’m curious about how such a perspective could be adopted regarding COVID-19 and the related pandemic. For scientists and related communities, one could investigate different groups of scientists who hold different views – such as those who are convinced that the current situation is a once-in-a-century pandemic, and those who are convinced it isn’t (see Ioannidis, 2020). According to this approach, whether COVID-19 is or isn’t a once-in-a-century pandemic doesn’t affect their beliefs about whether it is. Rather, social forces are hypothesised to be at work. The sociologist would intentionally ignore empirical evidence regarding the pandemic (since it is assumed to not affect what scientists come to believe) and instead solely focus their attention on the possible social causes of beliefs about the pandemic. This might be a useful way to better understand why different views about the truth of the pandemic have emerged despite scientists (and medical experts) having access to the same information on it. As Collins suggests, such scholarship could also be useful for probing the ‘socialness’ of actors (e.g. scientists) and its wider consequences.
For non-scientists, one could similarly consider social causes. Many important social factors like trust and reputation can have an influence as well as social networks. Here I’m thinking about research groups like the modellers based at Imperial College London and their influence on the beliefs of politicians and the public. The New York Times noted related aspects: “With ties to the World Health Organization and a team of 50 scientists, led by a prominent epidemiologist, Neil Ferguson, Imperial is treated as a sort of gold standard, its mathematical models feeding directly into government policies”. Other modellers at Oxford University later questioned the work of this group and raised concerns about their influence on the coronavirus pandemic policies being developed by the British Government, but they were too late – many major policy decisions had already been taken.