Steven Pinker: RATIONALITY (2021)

Pinker, Steven. Rationality: What It Is, Why It Seems Scarce, Why It Matters. Viking, 2021

I realize I’ve never written up a book by Steven Pinker on this blog, though I’ve read several and I think Pinker is one of the very best scientist/writers around (along with, say, Dawkins and Wilson). His themes are often profound, his writing sparkles, and his illustrations are colorful. But his books tend to be long, and detailed, and so my notes are long, and thus difficult to boil down into any kind of readable blog post.

The current book may be an exception, because it easily parses into a half dozen related themes, along which I can summarize my notes. By my count there are now eight big Pinker books aimed at general audiences, and another four that seem aimed more at specialists in his fields of psychology and linguistics. Of the eight, I’ve read five, and have notes on four. So let’s start with this latest one first, a book just published three months ago.


Pinker’s introductory theme:

  • How do we make sense of all the irrational things happening all around us? Primitive mindsets beset by fallacies and illusions? Yet we are smart in many ways.
  • None of us is rational enough to consistently come to sound conclusions: rationality emerges from a community of reasoners who spot each other’s fallacies. Another conclusion: how some irrationality is about pursuit of goals other than an objective understanding of the world. [[ like fortifying the tribe ]]

Summary: The book breaks down in four parts, the longest covering 7 chapters:

  • Chs 1,2: how rational are humans? And, how do we distinguish between rationality and irrationality?
  • Chs 3-9: These cover all of varieties of critical thinking about being rational (so much is this is familiar): logic; probability; Bayesian reasoning; risk and reward; hits and false alarms; self and others (game theory); correlation and causation.
  • Ch 9, Finally author comes to crux of the book’s ideas: what’s wrong with people? Why do people seem more irrational than ever before?
  • Ch 10 offers evidence that rationality matters: it’s improved the human condition, and it’s all that can help us solve many of our problems.



Pinker says he wrote the book as a one-stop-shopping for information about rational thinking and the reasons people have problems with it. In that he likely succeeds, but at the same time much of the material is familiar. The opening chapters are good, especially for insights about when being irrational can be strategic, and how working out conflicting goals is the basis for wisdom and morality.

The key chapter, Chapter 10, attempts to explain our modern situation, and he makes some excellent points about how our “cognitive facilities work well in some environments and for some purposes but go awry when applied at scale, in novel circumstances, or in the service of other goals. The idea of motivated reasoning and myside bias are familiar; less so the ideas that humans don’t “believe” different things differently, and that many people resent that science has “disenchanted” world. But does all this really explain the current situation, in which the dumb half of the country is ready to go to Civil War? I’m skeptical.



  • Ch1 gives examples everyday problems that trick people: simple math problems (e.g. two items together cost $1.10, and one of them costs $1 more than the other. What are the prices?); simple logic problems (testing a rule that involves four cards); the famous Monty Hall problem, which even experts get wrong; how people rate specific items as more likely to happen than more general ones. We understand that some cognitive illusions are features, not bugs, to take context into account. “They lead to incorrect answers, yes, but they are often correct answers to different and more useful questions.” 33.1.


  • Rationality is the ability to use knowledge to attain goals. If you “argue” against reason, you lose; nor is reason about faith; it’s a method to follow. (Those who reject reason attempt for impose their ideas by force.)
  • Not everything needs rational arguments, e.g. falling in love, the pleasures of life.
  • Yet such passions can be in conflict with each other; working out such conflicts results in “wisdom” and “morality.”
  • There can be conflicts among goals; among time frames, e.g. our present and future self.
  • There are some situations where it’s better not to know: e.g. spoiler alerts; the sex of a child; double blind experiments.
  • And in some cases it’s rational to be irrational, or at least powerless: the game of chicken; bargaining for a car; a “man of honor”; a suicide terrorist.
  • Taboos are ideas considered evil to consider rationally, for some: organ donation; budgetary decisions about end of life. “The art of political rhetoric is to hide, euphemize, or reframe taboo tradeoffs.”
  • We can’t deduce morality in reason (Plato’s Euthryphro demolished this) but can ground it in reason. We can prefer good things to bad things, and avoid privileged positions. The result is the Golden Rule, and it’s many variations. Kant; John Rawl. The result of self-interest and sociality is morality. Working out conflicts between one’s present state and possible future states is wisdom.


  • About formal logic; it deals with forms of statements, not their contents, using connector words like And, Or, Not, All, Some, If. A valid argument is not the same as a sound argument; the latter requires the premises to be true.
  • Logical fallacies are now well-known—author mentions a bunch, and has a 2-page index at the back just of biases and fallacies.
  • Why can’t we avoid these errors and apply logic to see who’s right? Three reasons: there are disagreements in how to define certain concepts, like “bachelor”; logic is irrespective of content and doesn’t necessarily prove anything about the real world; and people care about family resemblances more than crisply defined categories. (A whole category of fallacies arises from thinking in black and white.)
  • So what is rationality if not logic? The answer we think in cognitive models like pattern associators, e.g. where examples are graded to the degree a property is present, intuitively and without conscious thought. A computing breakthrough in the last decade has led to “deep learning networks” allowing things like speech recognition; but they can perpetuate biases because we can’t examine their inner workings.


  • Probability and Randomness. Examples of weather forecasts, how we assess probability using the availability heuristic [[ leading to the examples of how people think wrongly about the world in Rosling and Duffy ]]; how the press skews our impressions of the world; about different kinds of probabilities, disjunctive events and conditional effects.


  • Belief and Evidence: Bayesian Reasoning. The goal of the “rationality community” is to be “less wrong.” We walk now of “priors.” The idea is simple enough: we update our provisional conclusions as new evidence comes in. But the equations and examples can be tricky. It requires taking into account base rates, leading to conclusions that miracles can be dismissed; yet considering base rates can be considered profiling, as in the insurance industry. Base rates can be different among different groups.


  • Risk and reward—rational choice, the idea of maximizing the “expected utility.” Economics used to assume people were rational actors. The theory of rational choice is complex, includes ideas of marginal utility and being risk-averse. People violate rational choice decisions all the time.


  • Hits and False Alarms. Signals and noise; false-positives and false-negatives. We need to analyze the costs and benefits to make useful decisions. These ideas apply especially in the courtroom, where our tolerance for errors varies based on the crime. Scientists also are prone the false assessments, as when considered Type I and Type II errors and apply the p < .05 rule.


  • Self and others – game theory. How our choices can depend on someone else’s choices. Some are non-zero games, like scissors-paper-rock or more sports, where the best strategy is not to play in the first place. More complex examples are the Prisoner’s Dilemma and the Tragedy of the Commons, the latter an example of Public Goods games. They require various methods of detecting and punishing cheaters, “free riders.” Locally: gossip; threats; in larger communities: taxes, a court system, jail. These ideas are part of the social contract, and undermine the concepts of anarchism and libertarianism


  • Correlation and Causation. Everyone learns this but quickly forgets it; confusions are common in public discourse. Long discussion of correlation, with regression to the mean; of causation. Aligning the two might involve third factors; real world experiments might be unethical. Public discourse assumes everything has a single cause, when often many interacting causes are involved.


  • So—why does humanity appear to be losing its mind? Vaccines, conspiracy theories, climate denial; beliefs in supernatural beings have not diminished.
  • Author dismisses several glib explanations: the effects of all those fallacies discusses prior; that social media is to blame (but fake news goes back at least to the scriptures); or that fake beliefs offer comfort (why should this be so?)
  • “To understand popular delusions and the madness of crowds, we have to examine cognitive facilities that work well in some environments and for some purposes but that go awry when applied at scale, in novel circumstances, or in the service of other goals.” P288b.
  • Author discusses motivated reasoning (we evolved as intuitive lawyers, not intuitive scientists); the myside bias (how different sides will interpret the same data differently; To disagree with one’s group is to be branded a traitor; the most outlandish beliefs are the strongest indicators of tribal identity. We are trapped in a tragedy of the rationality commons.)
  • People entertain two kind of belief: reality and mythology. The local reality in which things get done; and the zone beyond immediate experience, beliefs that form a social reality to bind the tribe and give it a moral purpose.
  • The idea that everything can have grounds for being true is a modern luxury, a benefit of the Enlightenment. But this isn’t the natural way of human thinking. Those who resent the “disenchantment of the world” brought by the Enlightenement become resentful, trying to replace reality with the land of mythology.
  • Pseudoscience engages deep cognitive intuitions; we’re intuitive dualists, essentialists, and teleologists. Urban legends and fake news are entertaining; conspiracy theories flourish because sometimes they’re real and it could be more costly to miss one that to detect a false one.
  • But false beliefs can be dangerous. The arc of knowledge bends towards rationality; many superstitions of the past are no longer taken seriously.
  • We can reinforce the norms of rationality; censor those who lie. 313.7: “Since no one can know everything, and most people know almost nothing, rationality consists of outsources knowledge to institutions that specialize in creating and sharing it, primarily academia, public and private research units, and the press.” These issues have become in urgent in 2020, with two alarms: misinformation about Covid, and about the presidential election.


  • Why rationality matters. Does it really improve people’s lives? Is progress the story of problem solving, or the struggle between the downtrodden and the oppressors? Likely some of both, but the former is key.
  • Material progress has increased: health, safety, life expectancy; hunger reduced; wealthy increase; war reduced, due to democracy, international trade, and international organizations like the UN.
  • Moral progress has been made. Many cruel and unjust practices have declined over history, and often the first domino was a reasoned argument, e.g. a pamphlet or manifesto. Religious persecution; torture; criminalization of homosexuality; cruelty to animals; slavery; rights of women. Such arguments make the difference between “human progress and breaking things.” 340.4.
  • Final lines: So rationality guides moral progress and material progress. The faculty of reason, the formulas and institutions that magnify its scope: “They awaken us to ideas and expose us to realities that confound our intuitions but are true for all that.”


Note how often Hume and Tversky & Kahneman are given credit for many of these ideas.

This entry was posted in Book Notes, Steven Pinker. Bookmark the permalink.