Steven Pinker is Johnstone Professor of Psychology at Harvard, an elected member of the National Academy of Sciences, a two-time Pulitzer Prize finalist, a Humanist of the Year, one of Time’s “100 Most Influential People in the World Today,” and sits on Persuasion’s advisory board. His latest book is When Everyone Knows That Everyone Knows...: Common Knowledge and the Mysteries of Money, Power, and Everyday Life.
In this week’s conversation, Yascha Mounk and Steven Pinker discuss why common knowledge is the most important psychological concept you’ve never heard of, why authoritarian states are hostile to it, and where to find someone you’ve lost in New York.
This transcript has been edited and lightly condensed for clarity.
Yascha Mounk: I love your work. I have read many of your books. In the new book, you suggest that a deceptively straightforward concept—common knowledge—actually holds the key to explaining all kinds of different social phenomena. You take us on a really fun wild ride, both in terms of the anecdotes you provide, the illustrations you provide, and in terms of the kind of domains of social life that you illustrate through this seemingly simple concept. Before we jump into some of those points and some of those examples, what do you mean by common knowledge?
Steven Pinker: I’m using it in a technical sense, which is not the same as the everyday sense of conventional wisdom or something that people know. Common knowledge in the technical sense refers to a case where everyone knows that everyone knows something and everyone knows that and everyone knows it, ad infinitum. So I know something, you know it, I know that you know it, you know that I know it, I know that you know that I know it, et cetera.
Now, people start to smile when I explain the concept, because it sounds so complicated. It sounds impossible—although it’s tapping into a familiar feature of human nature: we’re always trying to get inside each other’s heads.
First of all, common knowledge is not impossible. The reason it doesn’t make everyone’s heads explode is that you can get a sense of common knowledge at a stroke from the sense that something is conspicuous or self-evident or public or “out there.” That is, if you are witnessing something and you witness someone else witnessing it and vice versa, that implicitly generates common knowledge at a stroke. The psychological manifestation is the sense that something is out there, including the metaphor “out there,” and other metaphors like “it was in his face,” or “the elephant in the room.” These are cases where something is so conspicuous that you know that other people are seeing it at the same time that you are.
That’s the idea. Why is it significant? Here I’m exposing readers to something that is well known in a number of sectors of academia, but mostly has not become, as we might say, common knowledge. That is, that common knowledge is necessary for coordination: for two or more people being on the same page. Making choices that may be arbitrary are very adaptive if everyone makes the same choice. The simplest example would be a rendezvous where I meet with you at a certain place at a certain time. The easiest way to do it is a phone call, which generates common knowledge. But in the absence of common knowledge, it’s not enough for me to know that you like to go to Starbucks, because you might know that I like to go to Peet’s, and so you might not go to Starbucks that day but go to Peet’s in anticipation of meeting me. But then I might go to Starbucks in anticipation of meeting you, and then I might second guess you and say, well, Yascha knows I like Peet’s, so he’s going to go to Peet’s after all. But Yascha is going to think, well, Steve knows I like Starbucks so he’s going to go to Starbucks.
Mounk: It becomes a guessing game where you just can’t accurately predict the other’s behavior, in part because they might be trying to predict your behavior.
Pinker: Exactly. That’s a simple example, but there are a lot of examples that just make large-scale social coordination possible, such as driving on the left or driving on the right. It doesn’t matter which side you drive on; but it matters a lot that everyone drives on the same side, and moreover that everyone knows that everyone drives on the same side, and knows which side that is.
Mounk: Unless I knew that everybody knows that we’re supposed to drive on the right or on the left, if you’re in Britain or India or wherever, even if I was going to drive on the right and even if could observe that a lot of people drive on the right, I’d have to drive with extreme caution, because I would think perhaps somebody isn’t in on this fact and they might suddenly drive on the left side and then we’re going to be heading for an imminent crash.
Pinker: Indeed—and in fact, in 1969 Sweden decided to switch from driving on the left to driving on the right to harmonize with the rest of continental Europe. That was a big problem. And so there was a lot of anticipation and announcement and the switchover took place at a stroke on, I believe it was midnight, December 31st. So it was a day that popped out of the calendar that everyone knew about, and that’s what made it work. Very often problems with coordination, where there actually isn’t literal common knowledge, are solved when there’s a focal point—something that pops out. It’s sometimes called a Schelling point, in honor of Thomas Schelling, who first gave the parable of a couple who was separated in New York. He tried to imagine how they would find each other. This was years before cell phones were invented. He suspected that they might each choose to rendezvous at the clock in Grand Central Station at noon, even if it hadn’t been particularly close to where they had been separated, simply because each one could reasonably guess that it would pop into the mind of the other, and more important, that the other would know that it would pop into their mind.
Mounk: That’s interesting, because if I thought of a Schelling point in New York, there’s a bunch of different ones that could come to mind. Times Square would be another obvious alternative. Perhaps the Statue of Liberty, although that’s too out of the way, so you might think, we’re not gonna go there. In some European towns, you might have more obvious Schelling points. I mean, in Munich, the town where I grew up, you would definitely go to Marienplatz, which is the central square, and there’s a central clock, and it’s sort of obvious which one it is. American cities are built in such a way that it might be less obvious.
Thanks for reading! The best way to make sure that you don’t miss any of these conversations is to subscribe to The Good Fight on your favorite podcast app.
If you are already a paying subscriber to Persuasion or Yascha Mounk’s Substack, this will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
And if you are having a problem setting up the full podcast feed on a third-party app, please email our podcast team at leonora.barclay@persuasion.community
I want to go back a couple of steps. I think this is a great preview of the terrain you cover, and now I want to sort of retread it a little bit more slowly. So to go back to this idea of common knowledge: in everyday speech, we might say that the fact that Shakespeare wrote Romeo and Juliet is a piece of common knowledge. An educated citizen should know that. Most people, even today in the United States, probably know that fact. That’s not what you’re talking about.
Pinker: It is not what I meant. I think a better term would have been mutual knowledge. And that’s what I called it in an earlier book, The Stuff of Thought. But I came to appreciate that among the experts in the academic fields that discuss it, “common knowledge” has just become the conventional term. Which, by the way, is itself an example of common knowledge. The reason that linguists are interested in common knowledge is that it’s how language works. Every word is a convention. It’s arbitrary. A rose by any other name would smell as sweet. But we can use the sound “rose” to convey the idea of a rose simply because all of us know that the rest of us know that that’s the way we do it in English. Even though it’s somewhat misleading, we’re kind of stuck with it. It got grandfathered in: words get grandfathered in, and when kids learn a word—they might hear it from a parent or from a sibling or from a friend—they instantly know that they can use it with any other English speaker, and they’ll be understood. So a word is a quintessential convention and a convention works by common knowledge.
Mounk: That’s a really fascinating point. I believe you make the point in your book on language that if you make up a word and use it in front of a child, they will assume that this is just a normal word in the language, and they are then going to use it with other people who won’t in fact know or understand that word because they have that assumption.
Let’s go to a second topic—the emperor’s new clothes. At the beginning of that story, you have a situation in which everybody knows the same thing. Everybody knows that the emperor is not wearing any clothes. They look at the emperor, they can see that he’s naked. They can see his genitals—and yet you would argue that’s not common knowledge. There’s this crucial distinction between everybody knows and everybody knows that everybody knows. What is that distinction and why is that so important? Why does it make such a difference whether everybody knows or whether everybody also knows that everybody else knows?
Pinker: Yes, and when the child blurts out, the emperor is naked, he wasn’t telling anyone anything they didn’t already know, but he changed the state of their knowledge because now everyone knew that everyone else knew, and everyone else knew that. Now, in the version of the story which most people are familiar with, that changed their social relationship with the emperor from obsequious deference to ridicule and scorn. So the second idea is that I argue that social relationships are coordination games. Whether you defer to someone, whether you are intimate friends, whether you’re lovers, whether you’re transaction partners, all of those are matters of common knowledge. What makes two people friends? It’s not as if they sign a contract, but each one knows that the other one knows that they’re friends and each one knows that, etc. And often challenges to status—in the case of the emperor’s new clothes, the dignity and status of the king, the emperor—are matters of common knowledge: questions of deference and authority and esteem.
The reason given in game theory is that very often, when there’s an encounter between two agents, they both might want a resource. Will they fight over it? Well, not necessarily, because it could be in each of their interests for one to defer to the other and claim the resource if the cost of fighting is worth more than the resource itself. The dilemma being, of course, that each one would rather be the one that is the alpha, the one with the authority that the other one defers to—but once they get locked into it, the subordinate defers to the dominant because he knows the dominant will stand his ground, and the dominant stands his ground because he knows the subordinate will give way, and the subordinate gives way because the dominant knows that he’ll give way, because both of them know the dominant will stand his ground and so on.
That’s why so much of social life consists of little games in which we save face or lose face. It’s why we have pissing contests, why it could actually come to blows, when the one who’s always on the short end of the stick decides to challenge the dominant one. And it’s why so many homicides are over what police classify as trivial altercations: road rage, fighting over who gets to use a pool table, jostling, cursing, looking someone in the eye. And one might say that on the international stage, it literally leads to war. Many wars, most obviously Russia and Ukraine, are really over nothing except esteem, standing face, rectifying humiliation, wreaking revenge. It matters, even though these are trivial altercations in one sense, because they could decide going forward who defers to whom in an unending set of confrontations in the future.
Mounk: There are obviously different game theory games that are famous and canonical. It seems to me that in some of the examples you just mentioned, the most important theoretical game is probably that of chicken. And the game of chicken is basically—there’s different ways of putting it—but one of them is that, I think in the famous movie Rebel Without a Cause, there’s a scene of chicken in which two people get into cars and they drive towards the abyss and the question is, who’s gonna jump out first? If you’re the one who jumps out first, then you lose face. But if neither of the people are willing to jump out early enough, then one or both of them may end up driving over the cliff and losing their life.
What you point out is that in certain contexts, you really don’t want to have a reputation for standing down, because how close to the cliff I’m willing to drive is gonna depend on my assessment of you. If I think that you’re a horrible coward who’s definitely sure to jump out well before the cliff, then I’m gonna say, well, I’m definitely gonna keep driving because I know that my rival is eventually gonna jump out and then I can stop the car. If I think you’re a crazy son of a bitch, that you’re gonna do whatever it takes to win this, then I might be much more scared and I might be willing to jump out of the way much quicker.
So I take it that you’re sort of applying that to something like an inner city game of chicken, where you might have somebody say, did you look at me weird? Do you have a problem with me? What’s going on? The rational thing to do is to stand down. Oh no, I wasn’t looking at you, never mind me. Keep your head down and so on. But what you’re doing by that is to signal that you’re subject to intimidation, that you can be messed with. And if you acquire a reputation that you can be messed with, then other people are going to be much more willing to play games of chicken with you, because they know you’re going to back down. And you can’t afford to acquire that reputation. So you might end up in a dangerous—perhaps a deadly—altercation over these really trivial things, because that reputation matters so much. And that’s to do with common knowledge. Why? Because you don’t want to create the common knowledge that you’re willing to back down.
Pinker: It’s because in the game of chicken—which is the same mathematically as what evolutionary biologists sometimes call the Hawk-Dove game, where you’ve got two animals that compete over a resource—the worst outcome is if they fight, they could both die. Even the winner could sustain a fatal wound. The best is if they split the resource or they resolve it by some convention, like whoever arrived first claims it and the other one leaves. But yes, in chicken or hawk-dove, where common knowledge comes in is that the stable relationship of one deferring to the other can settle it if avoiding the fight is a better outcome than fighting. Being the loser kind of sucks, but it’s better than being killed in a fight. Now, the problem is if you settle into a relationship where one dominates the other, where the other always gets his way. That can sometimes be decided by fairly arbitrary conventions in the animal kingdom. It’s why animals bellow and they rear and they strut.
All these rituals are now really understood to be hawk-dove confrontations. That is a case where it’s better for one to defer to the other than both to fight. Now what happens is, of course, animals sometimes do fight to the death and armies sometimes do meet each other on the battlefield. If the outcome is a foregone conclusion, then the rational thing for both parties to do is for the weaker to defer to the stronger without a fight. Now what happens to the beta who always gets the short end of the stick? He could challenge the alpha—risky—or he could just exit that relationship and choose a different partner to interact with if that is an option.
Another example goes back to Jean-Jacques Rousseau. It’s called a Stag Hunt. Imagine two hunters who can work together with heavy weaponry to fell a big meaty stag. There is some risk that if one of them shows up with the big bow and arrow and the other one just comes with a slingshot, then the guy with the bow and arrow won’t be able to fell the stag on his own. But if both of them show up with slingshots, they end up with nothing but puny rabbits.
Mounk: The idea here is that if both cooperate, it’s really positive for both. But if one defects and doesn’t help with the bow and arrow—the presumption is it takes two people to kill the stag—then the other person has no chance of killing the stag on their own. So that person is going to go hungry. Whereas the person who defects to just going after the smaller animal is very likely to get that, and at least have some kind of dinner. Is that the background idea here?
Pinker: Exactly. It’s not a Prisoner’s Dilemma because there’s no question of betrayal, selfishness, or loyalty. It’s really a question of assurance, of trust, of confidence. If each one fears that the other one may show up with a slingshot, they will show up with a slingshot, because it’s the safe but less lucrative option in the absence of common knowledge. Common knowledge solves the problem. If they have a phone call, it’s like, let’s hunt stag. Great, they all win. In the absence of a phone call, with each one guessing what the other one will bring, they will naturally slide to the second best, which is also the second worst alternative of each settling.
Mounk: I’m going to advocate here for listeners who may not be up on Game Theory. The crucial difference between this and the Prisoner’s Dilemma is that in the Stag Hunt, both have an interest in cooperating. As long as they’re somehow able to both cooperate, the outcome is better for both. The best outcome is that we both cooperate. So we both have an interest in getting there. I know that unless you’re just an idiot or forgetful or get distracted, you’re going to show up with your weaponry because you have no interest in going for the smaller animal.
Pinker: It is a quintessential case of common knowledge in the sense that it is not a question of loyalty or principle or fear of being cheated. It’s just you worry that the other guy may not cooperate because he’s worried that you might not, and so on.
Mounk: The Prisoner’s Dilemma is different. In that case, the basic setup is there’s two suspects. If they both cooperate, then they get a lesser punishment. But if I cooperate and you don’t cooperate, or you go to the state and you snitch on me, then I might get 10 years and you’re going to get off for free, and vice versa. So in that case, even if we have common knowledge, even if we ring each other and say, hey, you’re not going to talk to them, and I’m not going to talk to them, I promise, the problem is that in that scenario, I still have an advantage in not cooperating. Whether you cooperate or you don’t cooperate, if I defect from you, my relative outcome is better. There’s common knowledge about that—I know that you have an interest in defecting and you know that I have an interest in defecting. The most likely outcome is that we both defect, even for a suboptimal result relative to what we both could have achieved by cooperating.
But we also need to recognize that in many situations, the Prisoner’s Dilemma really doesn’t apply, and you might, I think, get an overly negative view of the social world if you assume that everything is a Prisoner’s Dilemma all the way down.
Pinker: Yes, that’s right. In the Prisoner’s Dilemma, the best thing that could happen to you is if you defect and the other guy cooperates, then he gets the worst outcome, and you get the best outcome. In the Stag Hunt, the best outcome is if you both cooperate, because then you both come home with lots of meat. The Prisoner’s Dilemma is the same as what’s sometimes called the tragedy of the commons, which is that the only equilibrium is for both to defect. And crucially it’s in a prison, so communication doesn’t help. It’s what in Game Theory is called cheap talk. If one of them says, hey, I’ll stay loyal if you will, why should the other guy trust you given that it’s still gonna be in your interest to defect regardless of what he does? And each one goes through that same soliloquy in their head. They both defect. They end up with the second worst outcome for the two of them.
In the Stag Hunt, communication is actually self-reinforcing. If I say, I’m going to show up tomorrow with a big bow and arrow, there’s no reason for me to defect. If I know that the other person hears it and can acknowledge it, then we both win. In general, here’s just a generalization about coordination, common knowledge, and game theory. Whenever you have a game with two equilibria, situations in which both of them make choices and neither one has an incentive to switch given the other guy’s choice, then it’s a question of, which of the two equilibria are they going to find themselves in? And that’s why Schelling pointed out 60 years ago that arbitrary symbols, caprice, whimsy, historical accidents—all of those can be quite decisive. For the couple meeting in Grand Central Station, there’s no reason they should meet there. There’s no way to predict that they would meet there other than the fact that it pops out as a solution.
Mounk: One example of a Schelling point in politics, it seems to me, is the strange politics of elections and dictatorships. Why is it that dictators are so intent on holding elections when it’s pretty obvious that they’re fraudulent? I would guess that a majority of dictatorships in the world hold elections, which is quite a strange, remarkable thing, really. And why is that?
Well, I think perhaps one answer that I’m medium confident about—not super confident about, but I think it’s plausible—is that if they stop holding elections, that is a really clear indication that they’re no longer a democracy. Until then, countries that are allied with this country for geopolitical interest, or people within the state who want to have an interest in pretending that they’re legitimate, can claim there is still a democracy. Even if everybody kind of knows it’s bullshit, some people can pretend that they don’t know it’s bullshit. The moment the election is canceled—it used to be that every four years there’s an election, but now suddenly there hasn’t been an election for five or six years—people can’t pretend anymore. That may be one of the explanations why people continue to hold elections.
Pinker: Particularly if there was an announcement that elections were canceled, where everyone could coordinate—because then everyone knew that everyone else knew that they were living in a dictatorship. And certainly, regimes holding power operate by common knowledge phenomena. This is an observation I owe to Michael Chwe, who wrote about 25 years ago a book with a similar theme to mine called Rational Ritual, where he talked about the role of common knowledge. He gave the example of public demonstration. Why is freedom of association enshrined in democratic constitutions? Well, each individual in a dictatorship may know that the dictatorship is inefficient and corrupt and cruel. They may even suspect that everyone else believes it. But they can’t be sure that everyone else knows that everyone else believes it. Without the safety of numbers, they might fear that any one of them standing up could be picked off. If everyone stood up at once, then there’s no regime on Earth that could stand it. For example, telling a British officer, in the end you will leave because it is simply impossible for 100,000 Englishmen to control 350 million Indians if the Indians refuse to cooperate.
That’s a basic fact of arithmetic. He could have said, chose to coordinate. So dictators clinging to power have to ensure that the people don’t coordinate. That even if every one of them hates the regime, they can’t rise up at the same time to storm the palace or to have a work stoppage. And so they try to prevent freedom of speech, freedom of the press that would generate that. I tell a recounted joke from the old Soviet Union in which a man is handing out leaflets in Red Square and sure enough the KGB arrest him, they take him back to headquarters only to discover that the leaflets are blank sheets of paper. So they confront him and they say, what is the meaning of this? He says, what’s there to say? It’s so obvious. Now the point of the joke is he was generating subversive common knowledge—that by accepting the pieces of paper, now everyone knew that everyone else knew, by the very fact that he was distributing the paper, that there was a reason to resent the regime, which the authorities wanted to stop. And in the case of life imitating a joke, Putin’s police force have arrested protesters for carrying a blank sign. Because the blank sign is subversive. Even with no content, it generates common knowledge.
Mounk: I kept thinking as I was reading your book about one of my favorite essays, The Power of the Powerless by Václav Havel. And in that essay, I think he grapples—without using that terminology, or perhaps quite having that frame—with the power of common knowledge in two ways. The first is, he describes that if you’re a greengrocer, you need to put in your store some slogan like, workers of the world unite. That’s meaningless, it’s an empty political slogan. Everybody kind of knows that it’s an empty political slogan. But if you refused to put that sign in your store, you would somehow mark yourself out as subversive. I wonder whether this is meant to do two kinds of things. First, I think it is meant to introduce common knowledge about the fact that everybody is choosing to comply with the regime. But even if people know that you’re in some ways coerced to do this, the fact that you look around and you see all of these signs gives you the common knowledge that all of my fellow citizens are in fact choosing to obey the diktat of this regime, which I may not like, which I may suspect is unpopular. But there’s this daily visible common knowledge that all the other citizens are choosing to obey it. So perhaps I’d better do as well.
This is why, secondly, Havel argues that ordinary citizens have this tremendous power in a totalitarian regime like that, because your act of refusal to put the sign in your store can take on this incredible political power. But even though it seems rather innocuous to not put that sign in your store, if many people choose not to comply with the small, not that important daily demands of the regime, that can perhaps be the beginning of a source of common knowledge. I’m not the only person who feels that way. I’m not the only one who wants to resist these kinds of instructions. My fellow citizens actually have the same thoughts that I suspected previously they have.
Pinker: It works best if everyone does it at the same time, if somehow they know what day it is that everyone takes down the sign. Which is why freedom of the press, freedom of speech, freedom of assembly are repressed in dictatorships—it’s less obvious than you would think, because you might think, well, a dictator can just let the people bitch and moan all they want. He’s got the guns, he’s got the power. And the reason that they don’t is that there is safety in numbers, but only if there is coordination, if everyone resists at the same time so they overpower the regime, they can only do that with common knowledge. By the way, a case of a public protest like that is a stag hunt, where it’s in everyone’s interest if everyone acts together.
Mounk: That takes us back, by the way, to the last point I was going to make a few minutes ago about the politics of elections and dictatorships, which is that on the one hand, you don’t want to announce that you’re cancelling elections because that creates a Schelling point. That creates a moment for people who are unhappy to say, we just found out there’s not going to be an election. Let’s go out and protest that. But on the other hand, of course, if you do hold an election, even if everybody in advance knows that it’s fake, even if everybody expects the results to be fraudulent, the moment the official results are announced and the dictator gets, depending on their ambition, 55 or 85 or 99% of the vote, that is also a natural coordination point for people to come out and say, this election was stolen. So you can see this sort of managing of common knowledge and managing of trying to avoid creating these Schelling points as a key part of the politics of dictatorship.
Pinker: And you get Schelling points often with decisive historic roles sometimes when there is what I call communal outrage. If there is an attack on a well-known individual, that gets publicity, that person symbolizes an entire group, and then that often galvanizes the entire group that can serve as a Schelling point for them to engage in massive aggression against what they perceive as the aggressor group. Because the event is the Schelling point, the common knowledge generator, that allows them to coordinate their resistance. Examples are 9/11, or Pearl Harbor, the Maine, or Lusitania—cases where there seems to be a disproportionate response, but it’s as if the assault on one of us, because it is common knowledge, that is everyone knows that it happened, if we accept it, we are accepting second class beta subordinate status going forward. We have to challenge it also in highly public ways to regain or claim dominant status.
In the rest of this week’s conversation, Yascha and Steven discuss what the madman theory of global politics shows about common knowledge, how to overcome polarization caused by different bubbles of common knowledge, and why inviting someone up to “see my etchings” is such an appealing euphemism. This part of the conversation is reserved for paying subscribers…
