Proof by Adam Kucharski: Milk or tea in first? Confirmed, you really can tell the
PROOF by Adam Kucharski (Profile £22, 368 pp)
In the early 1920s, a scientist named Ronald Fisher poured a cup of tea and offered it to his colleague Muriel Bristol.
She refused it on the basis that she liked her tea with milk added first. Bristol was sceptical that she could tell the difference, but Bristol insisted, and so they came up with a test.
Eight cups of tea were prepared – four with milk added before the tea, the rest with milk added after – and Bristol blind-tasted them.
Remarkably, her fussiness was shown to be well-founded. She guessed correctly which order the tea had been made on eight out of eight cups.
The method used to investigate Bristol’s claim about her tea preferences might seem intuitive to us now, but as Professor Adam Kucharski shows in his illuminating book Proof, it’s only fairly recently that statistical analysis has been used to sort fact from fiction.
In medieval England, people accused of crimes would be plopped in a pond. If they floated, they were guilty. If they sank, they were innocent. It wasn’t an ideal method given the guilty would be put to death, but the innocent would die too – from drowning. Things improved when justice began to be meted out by juries, but there were still problems.
In 1994, four jurors holed up in a hotel tried to discern whether an insurance broker was guilty of double murder by creating a makeshift Ouija board to communicate with one of his victims. ‘Vote guilty tomorrow,’ the spirit moving the Ouija board advised.
Kucharski acknowledges that his area of inquiry is vast, and at points, the book’s leaps from the history of mathematics to, say, the difficulty of analysing Covid variants, do induce intellectual whiplash. And the more he warms to his theme, the more unwieldy the concept of Proof appears.
But in a way that’s the point. What we consider to be proof is slippery. Witnesses frequently identify the wrong culprit and scientific principles that seem unassailable come tumbling down, as happened when Einstein showed that Newton’s theory didn’t apply to massive objects such as planets.
Taste test: Cuppa and biscuit
In recent years, scientists have worried about the ‘replication crisis’ – the problem that many studies’ conclusions collapse when they are tested afresh.
The most famous of these is the marshmallow test, which claimed that children who had the self-discipline to defer eating a marshmallow until later grew into slimmer, richer and better behaved adults.
Other studies that have been shown to be shakier than initially thought include the idea that standing in a power pose can boost testosterone and thus reduce levels of the stress hormone cortisol. Hogwash, alas.
At the heart of the problem is that scientists have an incentive to produce whizz-bang results, not least because those are the kinds of studies the public like.
Still, as Kucharski shows, efforts are being made to tackle the crisis by encouraging academics to declare what they’re going to study, and how, before they undertake their research.
This is a serious book by a serious person. It’s full of gnarly theorems and concepts. But it’s also a virtuosic look at the varied and sometimes amusing ways human beings have over the centuries used their intelligence to make sense of the world.
And it’s a reminder to be inspired by the motto favoured by the Royal Society: ‘Nullius in verba’ – Take nobody’s word for it.