Scientific debate around Covid-19 on Twitter and beyond has become increasingly polarised & unpleasant in recent weeks. It’s not a great look. Ironically, I think it’s less down to scientific disagreement, & more down to the limits of scientific knowledge. A few thoughts. (1/16)

#AcademicTwitter is rarely the civilised, courteous affair that people outside academia might expect of university researchers, but since the pandemic started it’s been particularly rough-and-tumble. (2/16)

I think some of the ill-tempered flare-ups are down to the same pressures facing everyone. Long hours and lack of face-to-face interaction with colleagues take their toll. Recently, though, I think the quality of Covid-19 debate in particular has got worse. (3/16)

Examples include people like David Spiegelhalter—usually revered for his calm, objective and insightful reflections—getting lambasted for daring to share a BBC article offering an iconoclastic critique of the UK approach to pandemic management. (4/16) https://twitter.com/d_spiegel/status/1307935809996234753

The antagonism seems to have come to a head particularly in relation to the ‘rival’ open letters to government and scientific advisors by two groups of scientists, w. rather different views of how best to manage the pandemic over the coming months. (5/16) https://www.bmj.com/content/370/bmj.m3702

Often this is put down to scientific differences; there’s much debate about the validity or otherwise of the groups’ studies and assumptions. But we’ve accrued lots of knowledge about (some aspects of) the virus, and our understanding has advanced greatly since March. (6/16)

We now have a pretty decent grasp of likely population-level mortality and morbidity arising from the virus—much better than when lockdown began, at any rate. But there are two important caveats. (7/16)

First, there are large confidence intervals around those estimates. We’re only likely to reduce those intervals through ongoing epidemiological study over time. In the meantime, our knowledge leaves huge room for optimism and pessimism. (8/16)

Second, we have much less certainty about how particular sub-populations are hit—most obviously, exactly what exact impact certain demographic and comorbidity risk factors have on coronavirus mortality and morbidity. (9/16)

Methodological debates notwithstanding, I don’t think there’s too much disagreement about what studies say in aggregate. What we’re seeing (e.g. in the two open letters) is less disagreement about science, and more about what to do given this scientific uncertainty. (10/16)

Thus we have scientists debating values—not (verifiable) facts or (falsifiable) theories. The differences they express reflect their views on things like risk appetite, the relative importance of different facets of health and wellbeing, political standpoints, and so on. (11/16)

The scientific community struggles enough with civility when it’s debating science, and when there’s some hope that the scientific method, or reasoned debate and ‘the power of the better argument’, might lead to resolution. (12/16)

When it comes to differences of opinion that are incommensurable—for example, on the importance of collateral economic damage, or the relative value of the lives of older people or the wellbeing of younger people—it finds it impossible to cope. (13/16)

What’s the answer, then? Probably greater acknowledgement of the limits to scientific knowledge, and to how much even a perfect scientific knowledge base (even more our current, still-limited one) can ever provide a recipe for action. (14/16)

Also, epistemic modesty, and acknowledgement of when scientists are straying from their area of expertise into an area where they have a legitimate contribution—but not necessarily one that should be privileged over others. (15/16)

And kindness. Simple recognition of the importance of listening carefully, disagreeing respectfully, and valuing different perspectives. Not just because it’s a good thing to do: also because it helps everyone to get beyond the bluster and into the substance of debate. (16/16)