Why Steven Pinker is wrong about rationality

Rory Sutherland on the failures of rationality and science

Rationality leads to better choices in our lives and is the ultimate driver of moral progress claimed famous author and cognitive psychologist Steven Pinker. Yet many, including groundbreaking advertising executive and public intellectual Rory Sutherland, disagree. Sutherland was interviewed by the IAI to discuss his views on the failures of rationality, science, and the need to embrace irrationality and magic in our decision making.

You had an exchange a couple of years ago about the topic of rationality with Steven Pinker. Pinker wrote a book, ‘Rationality’, arguing that we need more rationality, and that its adoption would lead to better outcomes. You took issue with some of the conclusions that he reached. Why?

I think we have all been indoctrinated in some ways by high school science, in particular the hard sciences, to believe that there's an absolutely perfect correlation between the quality of reasoning that goes into a decision and the quality of outcome. I think that's simply an unsafe assumption for all kinds of reasons. It's a perfectly safe assumption in GCSE maths or physics when the opposite of a right idea is wrong. But in major questions, the opposite of a good idea may be another good idea. Furthermore, rationality probably didn’t evolve to drive human decision making. If you think about it, every other animal on the planet managed to survive and reproduce perfectly satisfactorily without a major faculty of reason. They did it simply on instinct.

The argument is that we evolved rationality as a social species to win arguments or to defend our decision making, not necessarily to inform our decision making. And if you think about it, contra Steven Pinker, there are lots of reasons why being rational may not be a good idea.

First, the information you may have is almost certainly incomplete. And therefore, by optimizing on unrepresentative information, which is almost always the case in a real world decision-making, you may hideously upweight certain quantifiable factors, and completely ignore other factors which are not unimportant, that are simply not numerically expressed at the time.

The second thing is that of course all data comes from the same place, the past. Gravity is a pretty good example. You can make fairly solid predictions based on a rational anticipation that gravity will operate in much the same way tomorrow as it did yesterday.

However, if you're dealing with something to do with human behavior, for example, or on a more extreme level, fashion, it's completely unsafe to believe that. And Aristotle, as well as other thinkers like Hayek, absolutely cautioned against what he called scientism, which is the deployment of the scientific method to solve problems in fields far removed and far less certain than those in which the scientific method was developed.

So it’s possible to be entirely rational and make bad decisions in your view?

Yes, that’s what G.K. Chesterton said. He talked about how wonderful it was to be a reasonable man, because you can find a reason for pretty much anything you care to do. And this is why I think in a way, education is making us stupid. People are finding it easier and easier, particularly on the back of very, very bad use of statistics, to justify terrible decisions.

Some of your listeners will have heard of the work of Iain McGilchrist who believes that the brain has in some ways become imbalanced. That the part of the brain that actually should be a kind of finance function, a kind of compliance function, has come to dominate decision-making to an unhealthy extent. Mr McGilchrist’s hypothesis is that this has happened at an individual level.

A further hypothesis, however, is that that has happened even more at a collective, institutional and political level. And the reason is, I would argue, the real scope for solutions is not in being right. It ‘s in widening the solution space. I would argue that the insistence on induction and inductive reasoning automatically limits the number of solutions at which you can arrive. Particularly if you're forced to use a single discipline like economics or something of that kind.

I think this is an interesting sentence: There are far more good ideas that you can post- rationalise than you can pre-rationalise. This partly explains the incredible role of luck in scientific discovery. But it also defends the use of creativity, the instinctive objective, inference or leap which arises through some valuable part of the brain which is just good at imagining alternative realities.

There are obviously people who have brilliant instincts, but knowing whether someone has brilliant instincts or not is less tangible than knowing whether they have followed some kind of rational process. How do you tell whether someone has good instincts?

I love working in the advertising industry where there are people, some crazy people, who by the way, are hugely intelligent, intellectual, educated and so forth. Some of them are, to all intents and purposes, apparently stupid, but usually they're a weird mixture of the two. I include myself in that, I suppose, in a sort of strange mixture of amateur, academic, combined with lunatic.

But the interesting thing about those people is they can arrive at solutions to problems that no faculty of reason could ever arise. They're obviously really good scientists who've done this as well. I think we can say that either through fluke or fluke combined with inspired observation or inspiration, some mixture of those three, the really significant scientific discoveries, the really significant entrepreneurial advances have really come about.

There’s an Australian economist called Nick Gruen. And one of the essays Gruen wrote is that the seemingly rational, wannabe Newtonian assumption in economics that there are always trade-offs is a massive constraint on human progress. One of the things you learn in a creative domain is that there actually is such a thing as a free lunch. There's money for nothing. There are win wins all over the place. What creativity can do is look at things that you assume are in opposition and come up with an idea or a design thought or something which actually resolves the contradiction to turn it into a positive; not a tradeoff. This very narrow idea which I think constrains people's thinking is that there's always a kind of trade-off that utility always comes at a cost of X or Y. But this often not right.

For example, what advertising would say is that you can conjure up utility out of nowhere by telling a story. If you tell an interesting story about a product and make the product either more comprehensible or more valuable or more noteworthy in some way, you ‘ve made it more valuable effectively at no cost.

Furthermore, one of the lessons of complex systems is there are also butterfly effects. Very small things can have very, very big effects. And yet most people the sort of Newtonian mechanistic, reductionist mindset leaves people effectively to go, it's a £10 million problem, so it needs a £7 million budget. What if the £10 million only needed a £20 budget though? I think there are loads of cases where actually that's absolutely true, but no one's looking for them.

Do you think science has suffered from this problem as well?

So Richard Feynman had this one right, which is basically outside his narrow field of expertise the scientist is just as capable of being a total moron as everyone else.

There is a particularly pertinent example of this. To describe people who thought there may well be a lab origin to Covid-19 as conspiracy theorists is absolute nonsense. No one was suggesting it was a conspiracy. They were suggesting it was a cock up, followed by a cover up.

But while conspiracies are comparatively rare, cock-ups followed by cover-ups happen all the bloody time. It's absolutely not implausible that Covid might have had a lab origin. And the fact that scientists were so ready to basically discount this, in many case, with absolutely no specialist knowledge themselves, is a problem.

Latest Releases
Join the conversation