Life in the Pandemic X: Exacerbating uncertainty

 Many things in life are uncertain (apart from death and taxes obviously). And many things are uncertain in science. Indeed identifying, controlling and quantifying uncertainty is a key aspect of the practice of science. We’re so keenly aware of uncertainty that we try to dissuade students of talking about science “proving” things, as though in any given situation absolutely all uncertainty can be removed. We don’t think that it can be, and we can therefore never be “certain”. What we seek to do is accumulate evidence supporting a particular explanation for a given phenomenon so that it moves from being highly provisional (a hypothesis), to being fairly probably the correct explanation (a supported hypothesis), to being the best and most highly supported explanation we have (at which point it’s  usually elevated to the status of a theory). This takes time and effort. Even so, we also accept that the most accepted theory, with apparently lots of supporting evidence, can always be superseded by a new theory. This might be an extension of the original theory, or indeed a contradiction of it. But this whole approach raises  problems. It is tricky to explain (as you may have noticed), and it’s not the way most people think or speak most of the time. These problems (and why they matter) have been amply exposed by the pandemic.

Let’s start with the language problem. There are situations where certainty is conflated with clarity. In a startling reversal of form for the particular bunch of politicians currently running the UK, the pandemic mantra has been “We’re following the science, therefore….”. This is a reversal because it suited them in a previous situation (ie the Brexit debate), to downplay the view of “experts”. But as I’ve noted before, in the pandemic, this has changed. Experts are in; but uncertainty is not out.

Politicians and the media, are very keen on what they call clarity. But COVID19 is a virus new to  humans, and therefore new to science. Nothing was known, indeed could be known, about it (although things could be inferred). Early in the pandemic, at the time when many key decisions were being taken, the science was more than usually uncertain, and therefore the scientific advice to politicians had to be highly caveated (this is an assumption on my part, I wasn’t privy to it). But this doesn’t make for snappy press conferences. And it almost certainly guaranteed that the advice would change, and therefore the instructions issued by politicians would have to change (example: face masks). The media don’t particularly help in such situations. Their stock in trade is the language of u-turn and climb-down. It might have been wise to clearly communicate from the start that the course of action being embarked upon was based on a consensus of what, given the evidence at the time, was reasonable. Not certain, but reasonable. Problem is, would any of us reacted as we need to if the politicians had spoken this way?

To be fair to them, there have been some sceptics and deniers who have been happy to jump up and down and accuse them of exaggerating the danger of the situation for nefarious political ends. They have pointed out that for all the talk of half a million UK dead and the NHS overwhelmed, this was not the disaster that developed. But this is to miss the point. The one experiment that could not be done was the one that involved doing nothing and essentially letting COVID19 run its course. So on the basis of (suitably caveated) advice, we had our lockdown. And while we can’t be certain (that is, after all, the point I’m making), the difference in case and death curves (eg see here) between most EU countries (including the UK) and others like the US and Brazil, suggests that this was indeed a sensible course of action. As an aside, we have to now hope that we don’t blow it, and revert to the earlier trajectory that could lead to disaster. However, at least some of the critics seem to suggest that with all the uncertainty involved, essentially nothing should have been done. Action should only have been taken once all doubt had been removed. But then that would have meant nothing would have been done. And many thousands more would have died, deaths that we have almost certainly avoided. It will perhaps be possible to demonstrate this statistically, once more  evidence has accumulated. But at the point the big political and economic decisions had to be taken, actual evidence was scarce.

We have heard this sort of call to wait for certainty before, both in another contemporary context and historically. And it’s here that the language problem, and the complexity problem intersect. Climate change, its cause, effects and what we should do about it (if we can do anything about it), is undoubtedly complex. The idea that it is caused by human activity (primarily the burning of fossil fuels from the industrial revolution on, increasing atmospheric CO₂) has been a matter of overwhelming scientific consensus for decades ie we’ve gone beyond hypothesis, supported hypothesis, and theory to consensus. Even still, scientists in this area will probably be unwilling to say they have no doubts, that the relevant theory/theories have been “proved” in some absolute sense. That’s just not the appropriate language of science. But that allows others to come along and say that the science is uncertain, there are alternative explanations or the whole thing is just a hoax. Here, a legal analogy might help.

I served on a murder jury some years ago. We were faced with the weighty decision of whether the prosecution had proved its case beyond reasonable doubt. Notice that you can still convict and have doubt. The question is whether the case is proved beyond reasonable doubt. One can always come up with lots of “could be’s” and “might have beens”. But if they fly in the face of the evidence, or are not supported by evidence, then they are not reasonable. And if they are not reasonable, they is no reason to pronounce the defendant “not guilty”. If the scientific consensus around climate change were a defendant in the dock, although there are doubts and uncertainties, they would be ruled out by the evidence as unreasonable, a guilty verdict handed down, and the jury would go away and sleep soundly, their duty done. And yet the uncertainty, complexity, and the language of science conspire to provide a space for those who say we should do nothing because we are not 100% certain, precisely at the time when action has to be taken.

At least some who operate in this space are following in a fairly inglorious tradition that has been exposed several times. They seek to foment doubt and increase complexity, obfuscate evidence and exacerbate uncertainty. They explicitly seek to sow doubt, of the unreasonable sort. The approach was famously summarised by a cigarette company executive in the 1960’s in a now infamous memo which stated “Doubt is our product since it is the best means of competing with the ‘body of fact’ that exists in the mind of the general public.”(1). What followed was essentially a well funded disinformation campaign of epic proportions. Meanwhile, cigarettes continued to be manufactured, sold and consumed and contributed to the early deaths of millions. The story of this and similar campaigns is expertly revealed in its gory detail by David Michaels in his books (2,3). And there’s evidence that there are commercial and other interests playing the same game with climate change. Stir up doubt, exacerbate the uncertainty, and the public will conclude that either the issues are so complicated and unclear that it would be premature to take action (like ban smoking or increase tax on gas guzzlers), or that the inconvenience of action is not worth uncertain benefits.

This kind of thing is happening in the pandemic. Reasonable people are not taking reasonable actions because, particularly in the US, misinformation is being spread and uncertainty is being exacerbated. The scary bit is that when the much hoped-for vaccine becomes available, we all know it’s likely to start over vaccination against COVID19. But, to resort to some unscientific language, you can be sure that wearing a mask and washing your hands frequently at the moment, and getting vaccinated once one or more vaccines have passed through the requisite trials, is a really good idea. I don’t doubt it.

 1. Michaels D (2005) Doubt is their product. Scientific American 292(6):96-101 (available on Research Gate: https://www.researchgate.net/publication/7806937_Doubt_Is_Their_Product)

 2. Michaels D (2008) Doubt is Their Product. Oxford Univ. Press

 3. Michaels D (2020 )The Triumph of Doubt: Dark Money and the Science of Deception Oxford Univ. Press

The insufficiency of science

There are scientists who talk about a “theory of everything” although it turns out they do not literally mean a theory of “everything”. There are others who have claimed that science can basically supply the correct answer to any correctly formulated question (at least any question worth asking). This is sometimes tempered to the view that science provides, at least in principle, an approach that can rigorously establish the truth about a given state of affairs even if in practice it’s currently difficult to see how. At one point it looked as though this was becoming a dominant view. Proponents of this sort of view, passionately and (usually) elegantly expressed, were the likes of Dawkins, Hitchens, Dennett and Harris. Let us call them collectively Ditchkinetteris (with apologies to Terry Eagleton who coined the term Ditchkins to refer to two of them; 1). As an aside, the power of this sort of view seems to be in decline, as I have discussed previously. In general, Ditchkinetteris’s take might be termed the sufficiency of science (SoS for short). It would be wrong to assume that SoS was ever a majority view even among scientists, although such things are hard to establish, erm… scientifically. It was certainly a minority view among philosophers (eg see Kaufman’s review of Harris’ “The Moral Landscape”;2). But SoS has now been implicitly undermined but one of its former (if only tacit) supporters, the journal Nature.

Nature published an editorial on the 27^th February entitled: “A code of ethics to get scientists talking”. This reports on a document produced by a group of scientists convened by the World Economic Forum and heartily recommends it. As the editorial points out, such codes are not new in science. Many funding and governmental bodies have their own codes. Interestingly the editorial claims that there’s a problem getting scientist to take them seriously and adhere to them. But what intrigues me is the question of what kind of thing is this code?

If SoS is true, then presumably such codes will be scientific. That would mean they would consist of hypotheses, predictions, experiments, results and conclusions. Or if not hypothesis driven (because not all science fits this pattern comfortably) they would consist of observations, measurements and conclusions. But there will be measurements and data, there will be stats, there will be theory; all the familiar elements of science. Right? Wrong. Actually what the particular code referred to consists of (and this would be true of all the other codes) are well meaning, sensible and pretty obvious advice about the kind of things we expect of responsible science. For example, responsible science seeks to minimise harm to citizens. Such a rule doesn’t appear to be scientific rule. It’s sensible, it’s the kind of thing tax payers expect, but it is not itself a scientific statement or a scientific rule. It’s the kind of thing I’d be happy to adhere to, as would all my colleagues, and practically any scientist anywhere I know of. But it’s not science.

The reasons given for why such a code is necessary are also interesting. It is valuable because “the code contextualizes natural sciences in a time of rapid technological change and popular questioning of expertise.” Not sure I understand the first point, but the questioning of expertise is familiar enough. The proponents of the code want to meet such questioning by “infusing research with “the most irreproachable behaviours”. But again, these are not scientific statements or aims, laudable though they may be. They depend on historical, sociological and ethical analysis, not science. So to properly practice science, we must look outside science, indeed our conduct must be ruled by principles which are not themselves scientific principles. This seems to be a blow against SoS.

Of course SoS never was true. Science always stood on foundations that were not themselves scientific. Principles, assumptions and commitments always lurked in the background that were rarely talked about. We all have them, use them and depend on them, and we’ve always known it. It was Bacon who suggested that we ought to purge ourselves of such “idols” in 1620, only for Kant to argue in the 18^th century that some of them are built into the very structure of our minds, they are wired in. Better to be aware of them, and control them, than deny that they exist at all.

Personally, I’ve always tried to be clear about my prior commitments. I’m drawn to science because it tackles an ordered universe in an ordered way. That order flows from the God who made the universe, and has sustained it ever since. He is the ultimate source of truth, so I only progress because He reveals His truth as I employ the tools that science provides, allied to the tools that He has provided. He also reveals His truth to others, even although they do not recognise Him or acknowledge Him in any way (indeed many of them are much better at this science game than me). I study the book of His works, and “think God’s thoughts after Him” (to slightly misquote Kepler).

While I’m actually running an experiment, collecting and analysing data, drawing inferences from it, accepting or rejecting hypotheses, I behave (and probably look) like a naturalist. I explain my results, accept or reject my hypotheses, in terms of mechanisms that are familiar in the field. But ultimately, on reflection, I know it is Him I’m studying. Because of that, I want to do it in way that honours rather than dishonours Him,  just like the Christian plumber, carpenter, bus driver, dentist or lawyer. I don’t work to please my boss, or the head of my Institution, or really for the good of the community or for the honour of science. All of these things are good things to do. But they are secondary. My aim is to “serve wholeheartedly as if (I) were serving the Lord, not men” (Ephesians 6:7). All these are prior, outside commitments. But it turns out it’s not just me that has them, indeed needs them, because science is insufficient. At least I’m (reasonably) coherent about it.

1. Eagleton, T. (2009) Reason, faith and revolution: reflections on the God debate. Yale University Press.

2. Kaufman, WRP (2012) Can science determine moral value? A reply to Sam Harris. Neuroethics 5:55-65.