Whenever Steven Pinker, cognitive scientist at Harvard, publishes a book, one should buy it and read it. This is part of living a good life. His books are not easy, although the prose is clear and concise, but they provide the reader with a breath of knowledge that is frankly extremely difficult to find elsewhere. Rationality, his latest book, is no exception.

In this book, Steven Pinker talks about rationality, or reason, why we are able to act rationality, why sometimes we do not, and how we apply rationality to different problems in life, such as establish causality, separate signal from noise, and learn. Contrary to many other books on the topic, it is not about how irrational humans are. There is some of it, of course: when one talks about rationality one has to deal with our heuristics and the situations where they fail. But it is more about how we apply rationality in so many different aspects, and principally a defence of doing so as the only tested way to understand reality and live better.

There is a type of experiment mentioned in the book that I found particularly interesting. First, there is a version of the experiment typically used to show the limits of human rationality. It is an experiment that asks the subjects to check a rule of the type “If P, then Q.” It goes as follows. Suppose that the coins of a country portrait a sovereign on one side and an animal on the other. They have the following rule: if the coin has a king on one side, it has a bird on the other. Now suppose that there are four coins on the table: one has a king, one has a queen, one has a duck (which, just in case, is a bird), and one has a moose. The question is: how many coins should you flip to check whether the rule is broken?

Most people say that the one with a king and the one with the duck. The answer is, of course, not that one: you need to flip the king and the moose. The one with the duck could have a queen on the other side and the rule would not be violated. If the moose had a king, however, the rule would not be correct. This seems yet another example of human folly. But does that mean we are unable to properly comprehend a rule such as “If P, then Q”? Not really. Another experiment demonstrates this.

Imagine that we have four envelopes in front of us from the Post Office. The rule goes as follows: If it is Express Mail, it needs a 10-dollar stamp. We can see the following: one envelop is Express, another one if normal, the third one has a 50-cent stamp, and the fourth one has a 10-dollar stamp. Which envelopes should we flip to check the rule? Now most people get it right: the Express one and the 50-cent one. Same type of experiment, but with a very different situation: we are trying to find a “cheater” in the second case. In this framework, our evolved brains use rationality properly to find whoever might be breaking the rules. We wouldn’t want this person in our tribe!

Rationality is a tool. We apply it, or not, depending on whether it is in our interest to do so. Sometimes it is, such as when some politician of the other team is distorting statistics about a sensitive issue. In other situations—for instance, when this politician is in our team—we are way less likely to apply rational thinking. “We evolved not as intuitive scientists but as intuitive lawyers”, as Pinker puts it in a wonderful Chapter 10 titled “What’s Wrong with People?”. Even if that is the case, he says, we should strive to use rationality more often. “We children of the Enlightenment embrace the radical creed of universal realism: we hold that all our beliefs should fall within the reality mindset.” Rationality is a sort of public good: “Each of us has a motive to prefer our truth, but together we’re better off with the truth.” For this reason, then, rationality “is not just a cognitive virtue but a moral one.”

Buy it and read it. Learn about rationality, about us, and about the case for rationality. It is very difficult to find so much knowledge, and so well written, in under 400 pages anywhere else. I attended an event a couple of months ago with him and Richard Dawkins, which is a great introduction to the book: it has been uploaded here. Enjoy.