An understanding and defense of rationality is crucial to the defense of liberalism. The whole premise of a free society is that we should make our decisions on the basis of discussion, persuasion, debate. But this implies that there is something to talk about, that evidence and logic will and should change minds. And it requires a confidence that this reasoning will lead to answers—to valid conclusions that, when acted upon, will improve human life.
Defenders of the intellectual legacy of the Enlightenment have done some excellent work in recent years to show us, from real-world results, that all of these propositions are true. Steven Pinker’s Enlightenment Now exhaustively showed the massive improvement of the human condition, both materially and spiritually, that has resulted from the embrace of reason, science, free inquiry, and persuasion.
Having demonstrated that reason has improved human life, it is appropriate that Pinker’s follow-up book, Rationality, is dedicated to the question of how reason works, and why.
Pinker begins by pushing back against the notion that there is some kind of inherent barrier to human rationality. This is the currently popular idea that we are all basically unfrozen cavemen equipped with a primitive tribal-emotional default setting, so that modern rationality is an unnatural imposition on human nature. As Pinker puts it, “In social science and the media, the human being is portrayed as a caveman out of time, poised to react to a lion in the grass with a suite of biases, blind spots, fallacies, and illusions.”
But Pinker argues that rationality is human nature and can be observed in less formal, more rudimentary forms, in the earliest human activity. “Hunter-gatherers—our ancestors and contemporaries—are not nervous rabbits but cerebral problem-solvers.”
He illustrates this with extended examples from some of those contemporary hunter-gatherers: the San bushmen of southern Africa. He shows how their famed skills as game trackers require the same powers of abstraction and inference that are the basis of modern rationality. “They reason their way from fragmentary data to remote conclusions with an intuitive grasp of logic, critical thinking, statistical reasoning, causal inference, and game theory.”
It would be surprising if this were not the case. We evolved our large brain precisely because its powers of abstraction and inference gave us a massive advantage in survival. That advantage would have been moot if we had not been using those faculties. So, as Pinker puts it, “If contemporary humans seem irrational, don’t blame the hunter-gatherers.”
Pinker’s overview of the different aspects and methods of rational thinking reads a bit like extensive, detailed notes for a college course, and that is pretty much what it is. “This book grew out of a course I taught at Harvard that explored the nature of rationality and the puzzle of why it seems to be so scarce.”
Such courses, in my experience, are surprisingly rare.
My own education in philosophy featured exactly one course on logic, and then only a very narrow kind of formal, symbolic logic that is almost completely useless for thinking about problems in the real world. For an understanding of rationality and thinking methods, I found myself relying much more on my coursework in science and mathematics. The usual sales pitch for a major in the humanities is that it will “teach you how to think,” yet that is a topic that is rarely addressed directly, explicitly, or systematically—despite thousands of years of accumulated and still accumulating knowledge on the subject.
The strongest aspect of Rationality is that Pinker draws from the breadth of that knowledge—not just from the history of philosophy but from the methodologies of science and mathematics. Rationality is not something fully discovered and explored by the Greeks two thousand years ago. It is something we have been constantly augmenting with the discovery of new fallacies to avoid and new methods of analyzing data and checking conclusions.
So in addition to the basic syllogisms and the experimental method, we get sections on statistics, “Bayesian reasoning,” and game theory, all of them ways of predicting outcomes and making decisions in the face of uncertainty. If you doubt the value of this approach, consider the example Pinker gives: a 90% accurate medical test for a relatively rare disease. Sounds great, right? It’s 90% accurate! But if the prevalence of the disease is 1%, and the false positive rate is 10%, then if you test 1,000 people, the number of false positives—100—will swamp the number of true positives: 10. Put another way, someone who tests positive for the disease has a 90% chance of not actually having it. It’s a counterintuitive result, but it’s an important one if you are a doctor trying to assess the results of the test and make an accurate diagnosis.
This sort of reasoning is an important addition to our cognitive tool kit, because one does not have to always have perfect knowledge and complete certainty in order to make a rational decision.
In a brief but interesting section, Pinker also grapples with the application to rationality to morality, taking on the so-called “Is-Ought Gap” in which moral conclusions supposedly cannot be drawn from purely factual or logical arguments.
Pinker’s response is not as thorough or as rigorous as one might expect if this were a work of formal philosophy, but it is along the right lines. Reasoning, he argues, implies a reasoner—a human whose whole existence is predicated on the need for food, shelter, companionship, and so on.
How do rational agents come into existence in the first place? Unless you are talking about disembodied rational angels, they are products of evolution, with fragile, energy-hungry bodies and brains. To have remained alive long enough to enter into rational discussion, they must have staved off injuries and starvation, goaded by pleasure and pain. Evolution, moreover, works on populations, not individuals, so a rational animal must be part of a community, with all the social ties that impel it to cooperate, protect itself, and mate. Reasoners in real life must be corporeal and communal, which means that self-interest and sociality are part of the package of rationality. And with self-interest and sociality comes the implication we call morality.
Morality, in this view, can be derived from the needs of the individual’s survival—self-interest—combined with “sociality,” the rules required to pursue those interests through cooperation with others. This view has also been called “enlightened self-interest” and “self-interest properly understood,” or better yet for the present purposes, “rational self-interest.” It is an approach to morality that is widely accepted in practice, often acknowledged in passing, as Pinker does here, but is rarely taken with the seriousness it deserves or developed to its logical conclusions.
This is also crucial to a defense of reason—since what is the use of reasoning if it has nothing to do with moral action?
Yet there are crucial spots where Pinker’s defense of the validity of reason is undercut by a strange kind of metaphysical shyness. This is surprisingly common today and I have noted it in other recent defenses of reason and persuasion: an unwillingness to assert that reason is valid simply because this is the way the world is. It is valid by virtue of the nature of reality itself.
Instead, today’s defenders of reason feel the need to fall back on appeals to social consensus and what you might call an epistemic social contract.
At the end of the day the discussants have no choice but to commit to reason, because that’s what they committed themselves to at the beginning of the day, when they opened a discussion of why we should follow reason. As long as people are arguing and persuading and then evaluating accepting or rejecting the arguments—as opposed to, say, bribing or threatening each other into mouthing some words—it’s too late to ask about the value of reason. They’re already reasoning, and have tacitly accepted its value.
So Pinker arrives at two conclusions: “(1) There is objective truth, and (2) I don’t know it (and neither do you)…. But the conviction that [objective truths] are out there licenses us all to develop rules we can abide by that allow us to approach the truth collectively in ways that are impossible for any of us individually.”
It is only later that he concedes, “The cooperation of the world when we apply reason to it is a strong indication that rationality really does get at objective truths.” I found myself wishing he had started right off the bat by taking a cue from Ayn Rand—or from Aristotle by way of Ayn Rand—and simply asserting that A is A. Reason and logic work because things are what they are, because of the basic, observable nature of reality.
An inability to say strongly and forthrightly why rationality is valid is a fundamental flaw in Pinker’s defense of reason, but at least this does not necessarily warp one’s understanding of what rationality is. In this respect, the more serious error in Pinker’s account is his acceptance of crucially wrong—but conventionally accepted—ideas from philosophers, particularly from early 20th Century philosopher Ludwig Wittgenstein.
Following Wittgenstein, Pinker concedes that “the concepts that people care about” cannot be given the precise definitions required by “classical logic.”
Just try, he said, to find necessary and sufficient conditions for any of our everyday concepts. What is the common denominator across all the pastimes we call “games”? Physical activity? Not with board games. Gaiety? Not in chess. Competitors? Not solitaire. Winning and losing? No ring-around-the-rosy or a child throwing a ball against a wall. Skill? Not bingo. Chance? Not crossword puzzles.
Wittgenstein’s conclusion was that, in pace of definitions, most concepts share only a vague, approximate “family resemblance” that makes firm conclusions impossible.
But here’s the thing. I have tried to come up with a clear definition of “game,” though I can’t claim credit for the success. it was a young acquaintance of mine, who unfortunately passed away many years ago, who came up with a simple one: A game is “a form of goal-directed action in which the goal as such is not a value.” What do we mean by “the goal as such is not a value”? In American football, everyone is trying to move an oddly shaped ball into a rectangle of turf designated as the “end zone.” Why does the ball need to be there? It doesn’t—not for any reason outside the context of that moment in the game, when it needs to be in the end zone just long enough to count as a “touchdown” and add to one team’s score. So we have goal-directed action devoted to the achievement of an end, but that end does not necessarily have any meaning or significance beyond what is assigned to it, usually arbitrarily, within the game.
Think through this definition with all of the examples Pinker cited above: board games, chess, solitaire, ring-around-the-rosy, bingo, a crossword. Goal-directed action in which the goal as such is not a value—it checks out for all of them. It even covers metaphorical usages of the word. For example, the sort of person who “plays games” in a romantic relationship is someone who generates unnecessary conflicts not because the issues are really important but just for the sense of power that comes from putting their partners through the drama. The goal as such is not a value.
No, being able to define one word is not a full answer to Wittgenstein, and there are other, older answers. For example, a hard-to-define word may have several different “senses” with different definitions that need to be untangled and disambiguated, which is precisely what you will find in any dictionary. Even earlier, this was all hashed out by the Greeks. Those who have read Aristotle will recall how much time he spends differentiating a concept’s various shades of meaning, on the premise that we can solve a problem simply by identifying precisely what we are talking about.
So perhaps it is not necessary to throw up our hands and abandon the search for clear and precise definitions that are amenable to logic.
Pinker’s decision to do so leaves a giant hole in his account of rationality, leaving his readers (and his students) with no guidance on the question of how to form a useful concept or create a clear definition. It is not merely an omission of crucially valuable information. It also causes Pinker to make a weird semi-concession to the advocates of the irrational, speculating that “neural networks” could approximate the “fuzzy family resemblance categories that make up so much of our conceptual repertoire” and in so doing “demystify the inarticulate yet sometimes uncanny mental power we call intuition, instinct, inklings, gut feelings, and the sixth sense.”
It's a bit odd to find a book on rationality that ends up rejecting clear definitions and attempting to vindicate “gut feelings.”
It goes without saying that Steven Pinker is not attempting to undermine his defense of reason. He shows how far you can go with conventional 20th-century philosophy—and also shows its limits.
The defense of Enlightenment values cannot assume that we can simply return to a firm and stable status quo ante. The Enlightenment’s defense of reason was incomplete and has since been encrusted with two centuries of philosophy that was often motivated by a desire to preserve certain beliefs that would not withstand the scrutiny of reason. The reason Wittgenstein could propose that common concepts like “game” were incapable of definition and people just accepted this reflects the fact that there were many people who really wanted to escape the restrictions imposed by clear definitions, and they were just looking for an excuse.
I have heard some recent talk about the idea of “Enlightenment 2.0,” a conscious attempt to create a new Enlightenment with new institutions for the dissemination of knowledge. What I especially like is the idea of thinking of this as a second, revised edition of the Enlightenment, with new attempts to place the defense of rationality on a firm philosophical foundation and clear away previous attempts that undermined our understanding of reason.
But this book reminds us that there is an audience for this quest and a lot of knowledge that we can bring to bear. Pinker himself suggest the path forward.
The principles of cognitive psychology suggest that it’s better to work with the rationality people have and enhance it further than to write off the majority of our species as crippled by fallacies and biases. The principles of democracy suggest that, too.
Building on this base, Enlightenment 2.0 will need more discussion and defense of rationality, and more innovative thinking about its foundations.
See also my discussion with Steven Pinker about his book.
Enjoyed the review, but take exception to an aspect your "game" definition, though I'm not in fundamental disagreement. You offered “A form of goal-directed action in which the goal as such is not a value.” Not a value? Perish the thought. The arbitrary ends of games become passionate values celebrated as deeply as any in life for some people. Sit through a Yankees game with Andy Bernstein some time, if you dare. Perhaps "optional value", or non-essential value. They're not metaphysically important outside the arbitrary context of the game, but "not a value" is not what you're getting at. Think of Ayn Rand's humor discussion: Humor is the denial of metaphysical importance to that which you laugh at. The ends of games are similar - if the end of a game has metaphysical importance, it's no longer a game.