Adam Kucharski (UK)
I have something of a reputation as a party-pooper in the 91av office. At our weekly meeting to discuss the latest and greatest ideas in science, colleagues will eagerly describe some incredible new theory or spectacular find, then sit back to field questions. I can already hear the groans before I open my mouth to say: “Yes, but what is the actual evidence?”
I don’t do this because I enjoy shooting down ideas (OK, maybe I do enjoy it a bit), but because, to me, determining whether something is true really is the point of science. You may have an idea so exquisite that it seems impossible for it to be anything but true, but if reality disagrees, then all you have is, at best, fan fiction. String theorists get angry when I point this out.
Advertisement
That said, the process of identifying truth is perhaps murkier than we often consider. What counts as evidence? Are some types of evidence more worthwhile than others? And to what extent should new evidence cause us to change our minds?
In his new book, Proof: The uncertain science of certainty, Adam Kucharski answers all these questions and more, reviewing the history of separating fact from fiction, while also offering advice on how you can personally figure out what is true in a sea of noise and misinformation. Let me see if I can apply some of his teachings to convince you that this is a book well worth your time.
First, there is the appeal to authority: Kucharski is a statistician and epidemiologist at the London School of Hygiene & Tropical Medicine, so you would expect him to know his stuff. But, as he explains, this is a pretty weak form of proof – you shouldn’t take my word for it that the book is good just because I have been impressed by his credentials.
91av book club
Love reading? Come and join our friendly group of fellow book lovers. Every six weeks, we delve into an exciting new title, with members given free access to extracts from our books, articles from our authors and video interviews.
Stronger evidence comes from the observations I gathered while reading Proof. One of the first of many historical anecdotes in the book is about the Monty Hall problem, a thorny question of probability, first popularised in the 1990s. If you have read any popular science books in the past few decades, you will almost certainly have come across it, and I must have encountered the question dozens of times, leading me to worry that Proof would simply be rehashing old ground.
Thankfully, it seems this initial data point was an outlier. Kucharski digs up a number of stories I had never heard before, from Abraham Lincoln using Euclid’s Elements to score political points, to the head brewer of Guinness creating one of the most important statistical tests still in use today. He also brings in his personal experience of the perils and pitfalls of modelling aspects of the covid-19 pandemic.
Ah, but we must remember that correlation is not causation. Just because my enjoyment of the book seems to be in line with Kucharski’s excellent storytelling, that doesn’t necessarily mean it was responsible. Perhaps, in order to gather stronger evidence, I should conduct a randomised controlled trial (RCT). I could assign half of the 91av staff to read Proof, while asking the other half to tackle a classic of the mathematical canon, something like Fermat’s Last Theorem by Simon Singh, as a control group.
The process of identifying truth is perhaps murkier than we often consider. What counts as evidence?
Maybe not. As Kucharski explains, while RCTs are often seen as the “gold standard” of evidence, demanding them in all cases can hinder our efforts to uncover the truth. This became apparent during the covid-19 pandemic, particularly regarding mask-wearing, where any trials of their effectiveness, let alone RCTs, were hard to come by in the early days of the outbreak.
Sometimes, other forms of evidence – the fact that masks were already regularly used in East Asian countries to control infection, or arguments around the basic physics of respiration – have value, he writes.
This is, essentially, the book’s underlying philosophy. “When dealing with complex questions and fragmented data, we can’t necessarily rely on traditional one-study-one-answer approaches,” writes Kucharski.
Proof is a great guide to embracing this complexity in truth-seeking – and if you don’t believe me, go and read some other reviews, rather than relying on a single opinion.
