As a lockdown-sceptical teacher of Philosophy, I tell my students daily that I am currently seeking what I call a ‘Greta in reverse’: a youthful person, ideally of innocent and appealing countenance, who would be willing to face the media and to lead a student strike not against school, but rather demanding the right to be educated, in school, maskless, and face to face, should a second corona-related closure of our educational establishments be proposed.
Since, however, I was never wildly enamoured of the original Greta, who seemed to me strong in irate accusation but depressingly weak in argument and positive proposals, I am aware that my alternative activist – let’s call her Gerta – will need, in the interim, to hone her intellect so that she is armed against the arguments that will be advanced against her. And if her studies in Philosophy are to serve any practical purpose, it must surely be their aim so to equip her.
Examples are the life-blood of Philosophy, and the silver lining to the dark cloud of lockdown, from my perspective, has been that it has offered a wealth of instances of dire logic and related philosophical failings, from which teacher and student alike may learn and profit. For the benefit of fellow teachers; for any parents who may be nurturing potential candidates for the Gerta role; and for the possible interest of any independent observers without such skin in the game, here are a few examples.
Begging the question
In popular parlance, and to the annoyance of those linguistic conservatives who have not already given the matter up as a lost cause, ‘beg the question’ is nowadays often used synonymously with ‘raise the question’. In Philosophy, however, ‘begging the question’ retains its older, technical sense, in which it denotes a particular kind of logical fallacy (also known by the Latin tag petitio principii): that of assuming without argument, and taking as a premise, the supposed truth of one’s preferred conclusion as to the matter under discussion.
A nice example of question-begging can be found in a recent article from the ‘Reality Check’ section of the BBC news site, in which the journalists – Simon Maybin and Josephine Casserly – set out to pour cold water on concerns about false positives in COVID-19 testing.
While acknowledging the truth that false positives can seriously exaggerate the numbers of cases, especially when rates of actual infection are low, Maybin and Casserly downplay the problem by insisting that it is of no significance once one takes into account the fact that current community testing is not of randomly-selected members of the population, but of people who have symptoms, are in care homes, or are in hot-spot areas.
Now, it is of course correct that where there are independent grounds for certainty that the prevalence of the disease among those tested is high, the significance of the false positive rate will be reduced: it will make less difference to ‘reliability’ in the sense that matters – the likelihood that one is actually positive, given that one has tested positive. That fact is understood and accepted by all those who have expressed concerns about the lack of serious attention being paid to the risk of false positives.
The journalists’ aim, then, is to establish that incidence of infection is high among those tested – sufficiently high that the public need not worry about false positives. The allegedly small minority of false positives is likely (the journalists imply, but do not say explicitly) to be rendered trivial or negligible, especially once the possibility of some false negatives is also taken into account.
What do the journalists go on to say to support the view that actual testing is in fact of sectors of the population in which incidence of SARS-CoV-2 is high? Astonishingly, this:
Figures for late September from Public Health England show that 7% of community tests were positive. That means that of every 1,000 people tested, 70 were positive. Even with a false positive rate of 0.8%, seven of those would be false positives, but 63 would be true positives – the vast majority.
So the daily case count is not being skewed significantly by false positives.
Yes, but what is the primary source of the relevant figures from Public Health England? You guessed it: test results. At that point, the journalists’ argument collapses: since it is the reliability of the test results that is precisely what is at issue here, one cannot legitimately appeal to those results in support of the conclusion that the vast majority of the test results are correct, and hence that false positives represent a small minority of results.
Now, it is true that test results do not appear to be the sole source of the Public Health England (“PHE”) figures. The weekly PHE report cited also mentions the existence of what are, potentially, relevant pieces of supplementary evidence, in some of those tested, in the form of symptoms, deaths, and, as the journalists suggest, the residence of test subjects.
But there is no indication, and no evidence, of the relative value of such supplementary factors as evidence for what the journalists assert. You can read, as I did, through all 46 pages of the PHE report in question, and you will find within it literally no evidence that offers genuine support to the view that a large proportion of those tested are likely to be positive – let alone a proportion large enough to render concerns about false positives trivial. That is because there is not a single item of quantified evidence in the report that supports this view. There is simply the assertion, on the first page, that indicators supporting the view that COVID-19 activity increased during the relevant period include, as well as test results, the following:
- The number of patients in hospital with coronavirus
- Modelling data
- ONS and React surveillance data
- NHS indicators such as GP attendance, calls to the NHS 111 service and hospital admissions.
Even if such ‘additional’ indicators were independent of PCR testing (and of course, not all of them are); even if, furthermore, they were all faultlessly reliable guides to SARS-CoV-2 infection (which of course they aren’t) and even if all of these indicators were of incontestable trustworthiness (which they aren’t): in the absence of quantitative data for these indicators, they cannot possibly be thought to support the conclusion that the ‘reality check’ journalists draw.
It is for others, with better knowledge than I have of the empirical facts, and who are better able than I am to analyse their relative importance, to estimate the relevance of these different factors to the reliability of testing. But even a non-specialist, who has not studied the relevant data in detail, knows that it is just not true that all tests are of symptomatic individuals, or of people whose residential status renders them especially susceptible to COVID-19, or of people whose tests have been confirmed by some other method. As a schoolteacher, I know that numbers of asymptomatic students of my own have been tested; and since tests are available on demand, it is obvious that there will be many more asymptomatic individuals, without any special reason to suspect that they are infected, who are being tested under the present arrangements, and whose test data will be contributing to the figures. Without quantification of these elements, you simply cannot draw any firm conclusion about whether positive test figures reflect a relevant prevalence of infection or not.
I’m a philosopher, and I don’t claim to know, or to be able to estimate, the relevant figures. But what even a layman like me knows is that it is an irresponsible misrepresentation of the facts to assert, or imply, that the existence of some supplementary evidence that some who are tested have a higher likelihood of being infected means that we can ignore the problem of false positives.
In their attempt to reassure the public that concern about false positives is a storm in a tea-cup, Maybin and Casserly are therefore accepting as matter of faith, rather than of reason, that the reliability of the test results is backed by independent evidence; and they insult their readers by expecting them to share their faith on the basis of a supposed intellectual authority that lacks foundation. Street preachers, and Jehovah’s Witnesses on your doorstep, may in some respects be more annoying and intrusive (you don’t have to look at a BBC web page, after all) but at least they are normally honest enough to admit the non-evidence-based status of the belief that they invite you to adopt.
In an argument about whether incidence of the virus is sufficiently low as to present a risk of major distortion of test results owing to false positives, it is obviously question-begging to place any weight at all on the very test results whose reliability is at issue. It beggars belief that such blatantly fallacious reasoning should be published as a ‘reality check’ by a supposedly responsible broadcaster.
Why so sure?
The seemingly faith-based conviction that false positives don’t matter raises a further question: What is really driving the determination to downplay the sceptics’ concerns?
Suggestions are beginning to be made on these pages (and doubtless elsewhere – I wouldn’t know) that we may already be witnessing, on the part of politicians and their advisers, behaviour that is driven by a self-interested concern to avoid blame in whatever form of public inquest is eventually carried out on the Government’s response to SARS-CoV-2. My classroom, however, is not (yet) a zone in which such speculation is encouraged. So, putting such suspicions to one side, let us instead invite our putative or prospective Gertas to consider some other, more strictly philosophical, errors in lockdown enthusiasts’ thinking.
In the number one position must, of course, be that most popular of all fallacies relating to causation: the post hoc ergo propter hoc fallacy – that of assuming that since B follows A, B must have been caused by A. “The first lockdown worked: look at how the number of cases and deaths declined after 23 March!” “Masks work: look at how the number of cases and deaths have declined since people started wearing masks!” Or, negatively: “Look: it was when we opened up society again that the case-numbers began to rise again. So it was cafés, and shops, and pubs, and tattoo parlours that killed Granny.”
The logical error in question is well known. And others, better qualified that I am, have done valuable work to show how the evidence points positively against the causal connections assumed by the faulty reasoning (for example, by pointing out that the decline in deaths was already well under way before the onset of lockdown).
What is depressing is that one hears such remarks not just from members of the general public and students deprived of, or resistant to, what the Philosophy classroom has to offer, but from people who are presented as experts, and who should know better. Professsor Christina Pagel of University College London, for example, interviewed on Channel 4 news on 9 October, and speaking in support of an immediate lockdown, told viewers (with syntax tailored to match the quality of her argument), “It’s a choice between do you get to see your friends and family and have the virus grow, or do you not and have it shrink. … We learned that back in March.”
This is the same Professor Pagel who was interviewed on BBC Radio 4’s More or Less programme on 7 October. On that occasion, after expressing reluctance to assert that we were in a second wave, on the ground that that would be a matter of definition, she continued, “but we’re definitely in a second surge. We’re not that far from where we were when we locked down.”
I invite students to consider the plausibility of the professor’s suggestion that, of these two obviously non-technical metaphors, ‘wave’ might be precisely defined in a way that ‘surge’ is not. I have yet to meet anyone who can point to the supposed second wave on the excellent CEBR graphic published here on 21 October. I wonder whether Professor Pagel thinks that she can nonetheless point to a ‘second surge’ there. (Readers of Mike Yeadon’s recent article will not be surprised, incidentally, to learn that Professor Pagel is yet another mathematical modeller, rather than a virologist or immunologist).
A copy of the CEBR graphic is a present blu-tacked to my classroom door, with an invitation to students and colleagues to mark the position of the second wave, as in a spot-the-ball competition. This is intended as a way of sweetening the philosophical pill more traditionally administered by inviting students to consider statements about ‘the current King of France’: the pill delivers the (Russellian) lesson that not all definite descriptions (expressions of the form ‘the F’) are referring expressions: for there may be nothing at all to which they refer.
As yet, no one has submitted an entry; that is a sign, I trust, of a strong cohort. Perhaps I shall find my Gerta yet.
Bliks and the second
Most readers of articles on this site self-select as careful, critical, readers and thinkers who need no lessons from me about the manifold ways in which specious argument can be developed.
There is, however, one notion with which relatively few non-philosophers are likely to be familiar, but that increasingly strikes me as relevant to the way in which policy is formulated by the UK Government (and slavishly repeated by a lazy and uncritical public broadcaster). Belief in the so-called ‘second wave’, and in the importance of lockdown restrictions to control it, has acquired, I think, the status of what the late Oxford philosopher R M Hare called a ‘blik’.
Hare introduced this concept in a published discussion with Antony Flew about the status of religious language, well known to A-level students of Philosophy of Religion. Flew contended that statements made by typical believers fail to qualify as genuine assertions, on the ground that their adherents modify them, without limit, in order to accommodate apparently countervailing evidence. Thus, for example, Flew alleged that when the believer’s assertion that ‘God loves us’ is challenged by the observation that a loving, omnipotent, and omniscient God would surely not permit horrors like childhood leukaemia, the typical believer replies that God’s love is of a special divine kind that we cannot necessarily understand; and similar challenges are met in the same way. According to Flew, the original assertion suffers ‘death by a thousand cuts’: anyone who is willing endlessly to qualify the content of his/her assertion so as to accommodate no matter what evidence to the contrary is ultimately shown to have made no genuine assertion at all: for (s)he is not really committing her/himself to anything that her/his opponent denies.
In response to this challenge, Hare suggested that religious statements are better understood not as assertions of fact, but rather as what he called ‘bliks’. Hare offered, in this context, the memorable example of an imagined university student with an irrational conviction that all the dons are out to get him. As Hare puts it:
A certain lunatic is convinced that all dons want to murder him. His friends introduce him to all the mildest and most respectable dons that they can find, and after each of them has retired, they say, ‘You see, he doesn’t really want to murder you; he spoke to you in a most cordial manner; surely you are convinced now?’ But the lunatic replies, ‘Yes, but that was only his diabolical cunning; he’s really plotting against me the whole time, like the rest of them; I know it I tell you’. However many kindly dons are produced, the reaction is still the same.
Exactly how Hare’s remarks should be interpreted remains to some extent a matter of philosophical debate; but the central idea is clear in outline: religious language consists of expressions of ‘bliks’, as opposed to ordinary assertions that are responsive to empirical evidence. A blik has meaning, for its subject, in that it has an important influence on how (s)he views the world and how (s)he behaves. Even so, a blik (according to Hare) does not stand in need of evidence-based justification; nor will it (normally) be rejected on the basis of empirical evidence (no matter how strong) against it. A blik is (roughly) an article of faith: it is a belief, or an attitude, that isn’t answerable to evidence or reason, but that plays an important role in the life of a person who has it.
What, then, is the status of SAGE’s (or the BBC’s, or the general public’s) conviction that a second COVID-19 wave is imminent or already under way? The belief that the NHS is under threat; the conviction that in the absence of further restrictions on personal liberties and economic activity, we risk hundreds of thousands of excess deaths? Clearly, it’s a blik: these are expressions of an unshakeable faith in the most pessimistic outcomes – one that is retained and allowed to govern argument, attitudes in social life, and decision-making, regardless of (almost) no matter what evidence to the contrary. The virus – just like the dons, in the lunatic student’s view – is out to get us; and any evidence that suggests otherwise just shows how cunning it is, and how cautious we need to be to protect ourselves against its malevolence. These articles of faith are ‘meaningful’ to those committed to them, in that they affect the way they live their lives; and they are retained regardless of all rational considerations that count against them.
Some may be inclined to the view that it is no real accident that belief in ‘the second wave (or, perhaps, second coming)’ should resemble religious belief. They may suggest that it accords frighteningly well with certain other aspects of the phenomenon as we experience it: the cult of supposed experts – the scientifically-informed ‘priesthood’ controlling and interpreting the models, whose identities were kept secret for so long (as a guard, presumably, against some form of jealous magic that might undermine, them were their identities publicised). For me, that’s going a bit too far: my view is simply that these people can’t think straight.
Whether or not belief in the Covid threat is a blik, the big problem as things stand is one of changing popular belief on a grand scale. We have a mainstream, BBC-endorsed, uncritical version of reality that has been inculcated in the minds of an equally uncritical majority by a largely uncritical (but with a few admirable exceptions) wider media; and it is a majority that unquestioningly assumes the reality of such phenomena as the second wave, for no reason other than that reference to such phenomena is constant in the news media and elsewhere. Evidence for any alternative view is by its nature complex, and it has to get over the major hurdle of natural incredulity that goes with ordinary acceptance of the majority view.
It’a hell of a challenge. And I’m still looking for my Gerta. I hope she is up to it.