All PhDs are offered without subject area and are effectively the same, someone once observed. This is because the topic of a PhD is really just a vehicle for exploring the limits to knowledge. The fundamental question: how definitely can you know something?

This is similar to Sir Richard Doll’s work on the link between smoking and lung cancer. Today the link is beyond dispute. Yet it was difficult to prove at the time — and not just because of the opposition from tobacco companies. Almost all patients had strong exposures to smoke, be it from pea soup smog from burning of coal or the near ubiquitous scourge of nicotine addiction meant. It is often difficult to separate cause from effect.

It is similarly not proving easy to show a link between infection with COVID and various forms of exposure. Based on previous coronaviruses, we feel confident that this germ is spread by aerosols generated by talking and coughing that ultimately find their way onto a naive host where the infection cycle starts again.

A scale exists for how we can measure the strength of medical evidence, where 1 is evidence from a meta-analysis of multiple high quality randomised control trials through to 5 which is simply an expert opinion or individual case reports. Much of the current COVID research is reliant are on level 5 opinion, often based on non-transparent datasets and judgement.

Yet for all of this apparent knowledge about transmission, it is obvious that even dedicated scientists are struggling to show the effectiveness of even simple interventions. Perhaps the most obvious is face masks, where we have seen multiple revisions of the scientific advice. We also see confusion about whether we should all be 1 metre (as the WHO recommends) or 2 metres (as the UK Government recommends) apart, or indeed the number of people who can meet safely. Perhaps the most extreme example is that of the lockdown itself, which was always going to be a blunt and socially and economically costly tool. The computer modelling used to justify the lockdown has proven a complete mess, highlighting the difficulty of modelling unpredictable of human behaviour.

The scientific community is perhaps more comfortable with the constant chopping and changing nature of a debate under a blizzard of new facts and figures. Indeed, raised voices and the ability (pretence?) of having the latest research on the tip of your tongue is in sharp conflict between the PR consultants — who are thriving on 3, 4 and 5 point plans, where all points are a tag line that starts with a verb. It is interesting to see both sides also come unstuck.

The scientists – including a professor at Imperial - do not seem to appreciate that messages must be simple, unambiguous, and not for negotiation or exception. Ultimately all of science has a strong random walk element, where the point of inquiry meanders as different results become available. In this way, it is difficult for scientists to provide leadership – especially over the hours and days time frame that characterises the news/political cycle – and stay true to their recognition of doubt.

The politicians are struggling to find a style of leadership that survives deep uncertainty, and potentially rapid changes in direction. The initial strategy has been outed as a rather crude attempt to hide behind scientists and effectively blame shift if it goes wrong. Much of politics is about the art of doable – and politicians here are in somewhat uncharted territory. Methinks they could rapidly become passengers rather than drivers.

I wonder if what has been portrayed in the broadsheets as a rejection of authority and expertise is in reality not some sort of modern Luddism. Instead, perhaps the population has been inured to level 5 evidence, having heard people in authority during the great financial crisis and the Brexit debate being grossly overconfident (wrong?).