Do animals know when they don’t know? In 2003 a team of psychologists led by John David Smith attempted to answer this question. They trained a bottlenose dolphin to press a lever when it heard a low tone, and another lever when it heard a high tone. The dolphin loved this game, kicking up swirls of water as it raced towards one or the other lever.
Now and again, the researchers would do something different: raise the low tone gradually until it approached a high tone. That gave the dolphin pause for thought. It would “creep in”, because it didn’t know what to do, before opting to pull a third lever, indicating uncertainty.
Knowing when to reach for the lever of uncertainty is a rare talent in the animal kingdom. The researchers did a similar experiment with rhesus monkeys, who also know when they don’t know, and with rats, who, it turns out, do not. Rats are not the only ones, of course.
Elon Musk, the billionaire inventor and entrepreneur, is fluent in advanced mathematics, theoretical physics and rocket science. But he seems to be lacking in “meta-rationality”: the ability to recognise the limits of your own knowledge. That makes him say some very stupid things.
On 6 March, after thousands of deaths from Covid-19 were reported in China and Italy, and experts predicted similar in the United States, Musk tweeted: “The coronavirus panic is dumb.” He later mangled the data on US cases to claim, falsely, that the growth rate of the virus was falling.
Once upon a time, Musk probably knew he didn’t know everything. If you have a five-year-old child to hand – and over the next few weeks many of you will, every hour of every damn day – I suggest recruiting them for a scientific investigation. Cut a banana into two pieces and ask them to take the bigger piece. She will do so without a second’s thought, especially if it’s before teatime. Now do the same with another banana, but this time, cut it in half. She will almost certainly hesitate, for at least a few seconds, because she is questioning her own judgement. She has meta-rationality. For some reason, many humans lose this capability as they get older, perhaps because our society rewards bluffers over shruggers. We are more likely to assign status to those who project unjustified certainty than to those who say “I don’t know”.
I’ve never been more acutely aware of this as in recent weeks. Every time I check social media, somebody is confidently proclaiming that they know exactly how to manage a global pandemic, having spent a few minutes looking at data online and scanned a Wikipedia page. They explain, with implacable conviction, why they are right, and why Chris Whitty – chief medical officer, epidemiologist, infectious diseases specialist – is quite obviously wrong.
Unearned confidence can infect anyone, though some individuals are more vulnerable to it than others. If you are in one of the at-risk groups – columnists, barristers, breakfast-TV presenters, anyone who went to Oxford or Cambridge – you may already have displayed some of the symptoms, such as the inability to shut up about how Singapore’s social-distancing strategy differs from South Korea’s. In which case, I advise that you go into full opinion lockdown.
A couple of weeks ago I started to make my own “deep dives” into the data, poring over rates of transmission and graphs of cases and deaths. I found myself telling others what I thought about it all, until some residual sense of my own ridiculousness, combined with rising anxiety, made me stop. I cut down my consumption of information about the outbreak altogether, focused on not touching my face and felt better for it.
This crisis has presented a challenge to those of us who habitually share our opinions on current affairs. At least when we talk about the state of the health service, or immigration policy, we can appeal to some kind of personal knowledge, however sketchy. But while all those topics are complex – more so than we pretend – this one is dizzyingly so. It requires a deep well of scientific expertise that simply can’t be acquired in a few weeks, or even a few years.
It’s strange to see many of the same people who have been harping on this government’s supposed contempt for “experts” telling us that the government’s experts have it wrong. Trust in experts does not mean trusting only those who agree with you. It means admitting that on certain subjects you don’t know what the hell you’re talking about. It means that sometimes it’s best to leave the opinionating to others, as painful an act of self-denial that might be.
The scientific project itself is meta-rational, founded, as it is, on the assumption of ignorance until proven otherwise. Its logic is very different to that of politics, where everyone must be either right or wrong, and condemned or celebrated accordingly. When the government announced that it was shutting schools after all, it was perceived as an embarrassing “U-turn” – the experts had been wrong about something and had now become right. The meta-rational way to see it was that they had taken one decision with very incomplete data and another with slightly less incomplete data, and nobody will know if either was right until there is much more data, months or even years from now.
We value conviction and revere intelligence, but we greatly under-value meta-rationality. When I’m assessing decision-makers, what’s most important to me is not their intelligence or self-assurance, but their meta-rational ability. I look for people who are acutely aware of what they don’t know; who listen intently to other views; who know where they want to get to but are comfortable with changing course when conflicting information comes in. People who are prepared to pull the third lever.
This article appears in the 25 Mar 2020 issue of the New Statesman, The crisis chancellor