In 1939, with his native Germany under totalitarian rule, Bertolt Brecht wrote a poem. Called “In Praise of Doubt” it encourages readers to always question established orthodoxies. Brecht mocks “the thoughtless who never doubt”, and who “don’t believe in the facts”, relying instead on their own opinions. But the poet’s main target is those who tell the downtrodden that their place in the world is inevitable, preordained, and impossible to challenge. As Brecht says, revising “eternal wisdom” can bring real-world change too — by shifting our beliefs about what is possible.
Today, we live in a curious age of scepticism, not merely towards the authority of expertise, but also towards promises of social change. Unreasonable attachment to our own ideas often sits alongside an unwillingness to consider the arguments of others, even as we’re often cynical about their motives. In this context, Adam Kucharski asks an awkward question. What, he wonders, in his new book Proof: the Uncertain Science of Uncertainty, does it mean to be convinced of something? More to the point, what does it take to convince you — what kind of evidence, and what kind of reasoning?
Not everything can be proved with mathematics and logic. Well done, Euclid, for using proof by contradiction to show that, if two angles in a triangle are equal, the sides opposite the angles will be equal too. But does that help me decide if a defendant is guilty, whether to get vaccinated, or how likely a social media post is to be true? Unlikely.
Proof is not the argument-ending rubber stamp that internet pedants lust for. Standards of proof, in science, medicine, or the law, did not fall, ready-made, from heaven. Ideas like “beyond reasonable doubt” or “statistical significance” were developed by humans to answer important questions within the constraints of what they had to hand: evidence, people, technical resources and time.
Today, we take for granted that any new medical treatment will have been through a standard array of tests, including randomised controlled trials (RCTs), analysed with statistical methods now mostly performed by computers. All these tools were invented by researchers and statisticians to answer specific questions. Do babies thrive better on cow’s milk or mother’s milk? Does streptomycin cure tuberculosis? Which type of barley makes the best Guinness?
Nature, like a guilty defendant, does not generally crack under interrogation and deliver a full confession. Measuring the effectiveness of a treatment, or identifying the factors that influence outcomes, is equally a job for detectives: piecing together evidence, eliminating misleading or irrelevant clues, and working out ways to test a hypothesis. Even then, researchers may conclude that one thing seems to cause another, but they don’t really know why.
Does this matter? After all, as Kucharski points out, we use very effective medical techniques, such as defibrillation, every day without understanding why they work. We board aeroplanes without worrying that mathematicians still can’t explain exactly how the flow of air over the wings keeps them aloft. At some point, we just accept the evidence of results — without understanding causes. That’s especially true when familiar statistical methods are turbocharged by AI and fed by industrial-scale data, meaning social as well as medical interventions are often based on this kind of correlation, rather than a clear understanding of causation.
Even leaving aside the real problem of unintended consequences, however, there’s a drawback to this approach. Observing correlation lets us predict what happens if we do more of the same. But understanding causality allows us to conjure something new. Kucharski cites the philosopher Judea Pearl’s “ladder of causality” — rising from association, through intervention, to counterfactuals — to imagine an alternative version of reality. What would happen if we did X instead of Y? And, perhaps more controversially, what would have happened, had we not done Y?
As a medical mathematician, involved in advising the UK government during Covid as part of SAGE, Kucharski was at the sharp end of this kind of “what if?” modelling. He has more practical experience than most would want in using incomplete data to build scenarios of what might happen in the future — or, rather, in different futures. For Kucharski, after all, variables included not only natural factors, like how contagious different Covid strains would turn out to be, but also how people would behave, and how different government policies might affect that behaviour.
Decisions made on the basis of these models, as Kucharski and his colleagues were well aware, would affect millions of lives. There was no laboratory in which to test alternative policies, only mathematical models with many assumptions baked in. The stakes were high, time and information were limited, and public debate was highly politicised.
Kucharski is careful to distinguish factual disputes from mere policy disagreements. As he puts it, “deciding what is true and false is not the same as deciding what is socially right and wrong.” He rightly criticises the disingenuous claim that governments were just “following the science” approvingly quoting Austin Bradford Hill, best known for discovering the link between smoking and lung cancer. “‘It was no part of our job to tell the public how to behave with regard to smoking…To become propagandists would ruin us as scientists and make us “biased” presenters of further material’.”
Still, given the profound disagreements that emerged about the nature of Covid, and how to prevent and cure it, you might expect Kucharski’s book, like Brecht’s poem, to attack those who “don’t believe” in the facts. You’d be wrong. After all, if we are aware of what it takes to be convinced that something is true, should we not also think about what it takes to convince others? We ourselves are not machines, absorbing data and outputting hypotheses. We weigh evidence and argument, testing new theories about the world against all our existing knowledge and assumptions. It’s reasonable to assume that others do the same.
Kucharski’s personal experience of vaccine-hesitant friends and family was that they had varying reasons for doubt. Some felt they didn’t know enough. Others were deeply sceptical of pharmaceutical companies in general. The author found he could only convince them that vaccination was a good idea by “taking the time to understand why they doubted, then finding ways to address these specific doubts.”
This should be obvious to everyone, in a pluralist, democratic society. But we live in a world where “proof by intimidation (‘the evidence is clear’)” is too often invoked by those who regard the public as beneath rational argument. Ironically, research shows this kind of approach is not effective — and indeed tends to backfire. “Public trust,” warns Kucharski, “is eroded as problems are brushed away with appeals to authority.”
This points to a deeper problem with the widespread attitude that evidence, or proof, is merely ammunition for propagandists of every stripe. Those on the receiving end lose faith, not just in specific pieces of evidence, but in the very methods and institutions they should be using to weigh and test the arguments.
Kucharski quotes Jules Henri Poincaré, the French mathematician and philosopher, writing in 1908: “To doubt everything or to believe everything are two equally convenient solutions; both dispense with the necessity of reflection.” To rephrase, healthy scepticism gives way to fatalistic cynicism, what Brecht calls “the doubt which is a form of despair”.
Despite its title, “In Praise of Doubt” also targets those who doubt “not in order to come to a decision but to avoid a decision”. Writing in 1939, exiled from Nazi Germany and disillusioned by Stalinist Russia, he points the finger not only at those who fail to ask questions, or who suppress dissent — but also at those who don’t want to face the reality of their situation, or take responsibility for action.
Fair enough. Public discourse today too often falls into lazy sloganising, personal character attacks, and sloppy invocation of highly selected or dubious facts. We should all expect better standards of reasoning and evidence, and in turn deploy better arguments when trying to convince others. But political argument is generally not susceptible to the kinds of proof in Kucharski’s book. Types of evidence and reasoning appropriate to a laboratory, a court of law, or a philosopher’s musings, can only go so far.
Abraham Lincoln, fascinated by Euclid’s methods of mathematical proof, used “proof by contradiction” to show that one person could not legitimately enslave another. His capacity for logical argument against his opponents probably helped his election as President, but it did not end the practice of slavery. In the end, it took more than evidence and proof to change that social reality. It took both doubt that the existing state of affairs was the best (or only) possible world — and positively imagining an alternative reality. Yet it also meant convincing others that change was possible. And that, as Kucharski says, “is not just about data or research; it is a matter of psychology, politics and prior beliefs.”
In the end, nobody can prove in advance that it’s possible to convince enough people to change their minds. At some point, you just have to act on the strength of your conviction.
Disclaimer
Some of the posts we share are controversial and we do not necessarily agree with them in the whole extend. Sometimes we agree with the content or part of it but we do not agree with the narration or language. Nevertheless we find them somehow interesting, valuable and/or informative or we share them, because we strongly believe in freedom of speech, free press and journalism. We strongly encourage you to have a critical approach to all the content, do your own research and analysis to build your own opinion.
We would be glad to have your feedback.
Source: UnHerd Read the original article here: https://unherd.com/
