Agnes:

Um, so I thought we could talk about cancel culture today, um, because, uh, Scott Alexander had an interesting post, um, sort of referencing the difficulty of defining and also of drawing a line between like, at what point do we call some activity cancel culture, um, uh, versus not, um, and, and he references an old, um, essay of mine on, in which I explained why I thought philosophers shouldn't try to use petitions to adjudicate their disputes with one another, that that was sort of anti-philosophical. And I then drew on that in response to the controversy that arose when the New York Times was going to reveal his name and sort of undermine his blog thereby. And I thought they shouldn't do that. But I also thought the attempt by the sort of set of people who support him to try to use petitions to stop the New York Times from doing that. also seem to me like a certain form of cancel culture. Anyway, that's the context. That's why I was thinking about this. And so I myself would like to have more clarity on Yeah, on just what are the evils in play? And a question I have that comes up at the end of Scott's post is he's sort of assuming that there's a set of allies who share interests in protecting one another's free speech, and that there's a kind of coalition, a kind of possible coalition of all the people who really care about free speech. And I'm not sure that that is a possible coalition. So that's just a framing question. Okay.

Robin:

So I reviewed the topic a little in the last few days because it's been in the news, we should mention. The Telegram is a website that lets you have secret messages and the CEO of Telegram was just arrested in France because France wants them to reveal some of the secrets of people's messages and he doesn't want to do that. And then in the last month in the United Kingdom, there've been protests and then many people who have been supporting the protesters have been jailed for several years on the basis of saying things that in US would be very innocuous, like it's okay to be white or something. And so that also raised the issue in the press recently. But I think the how to limit government question is easier than the question you're highlighting, which is more about private action and social norms. And so, yes, let's focus on that. And the traditional idea of free speech, I think, is the idea that we should just let people talk and say what they want, and that you don't wanna hear what they wanna say, well, don't listen. But you shouldn't try to stop them from talking to people who would want to listen to them. Now, we have some, of course, legal constraints on free speech, i.e., you know, But they're pretty extreme. We might have secrets in wartime that we wouldn't want people to reveal or reveal private information about somebody that was their credit card numbers or say a naked picture of them. So I don't think we're focused on those sorts of issues that we're willing to accept that be some sorts of things we might want to prevent people from saying. They don't have much to do with big things. So the usual intuition for free speech rationale was this idea that we could all together learn more if we argued and debated instead of shutting people up. Because it's hard to learn from people who are shut up. And we can more learn from people who, when they say things that other people say things contrary to them, et cetera. But I think the place it gets harder is that we do have the sense that sometimes people want to coordinate to achieve outcomes. We do that in, say, politics, where we form political parties, we have movements where people are trying to coordinate. And one of the common actions of a movement or a coordination is to try to get a group of people and to recruit them to all sort of act and service of some cause. And then if you identify your enemies, then a potential tactic and attempting one is to basically push, repress them on all fronts. Basically, get all of your people to sort of do everything they can to discourage those people, i.e. don't date them, don't visit with them, don't buy their products. and don't listen to them. And that's more what's the key thing at issue here is some people have said some things that other people disliked, and then instead of just not listening, not buying the product, unsubscribing, whatever it is, they will coordinate to get a group of people who weren't paying attention to that particular person to say, you should do everything you can to also oppose them. And you should unsubscribe to any platform they're on. If they have a job, you should tell their employer to fire them, et cetera, as a way of using the power of a big group of people willing to coordinate to further the ends of your group. And that's more where I want to draw a line. That is, I'm pretty okay with people just not listening to somebody they don't like. You know, not going to a talk of theirs, not subscribing to their sub stack, et cetera. And that if nobody liked what somebody says, well then, yeah, they wouldn't have an audience. And I'm willing for there to be that consequence. That is, nobody will listen to you if nobody's willing to listen to you and nobody thinks you're worth listening to. But the idea that you would form these groups and then coordinate with these groups to further what you see as your group's ends by identifying someone as you see as your group's enemy, and then using everything you have, all the influence you have over all the shared institutions, to shut them up, that's when I worry more. But I can see that in some extreme circumstances that might be called for, that's where I'm trying to figure out how to draw the line. But at least that's how I would talk about what the key problem is, the key, the hard issue, as I think it's pretty easy maybe to limit what government can do, pretty easy to limit what, you know, individual people could do IEC. I let people say what they want and let them listen to what they want and, you know, evaluate what they want. And I let small publications have their editorial policies, but, I mean, in an intermediate difficult case, if you have a large platform, like say Twitter or YouTube or something, you know, somebody can use their influence over that, also shut people down. And, you know, how can I think about that when I'd rather think about either the big government and limiting what they can do in many small parties and letting them have editorial freedom? What position would these large platforms play in that space? Anyway, I've said enough, but that's how I frame the issues.

Agnes:

So the idea that we should just let people talk, and maybe we don't listen, but we don't stop them, seems to me to apply to very, very few contexts. So if I'm in a class, for instance, and there's some student and they're talking, and they're going on and on and on, the rest of us can't just leave the classroom? Or you often don't let me talk. And so the norm, let people talk, is not a norm that we generally follow.

Robin:

Let me clarify.

Agnes:

Let me finish. Let me finish. Let me talk. And I think that often the kind of conceit that there is such a norm goes along with the idea of something like a special magical place called the public sphere where we let people talk out there. And I'm not sure there is any such space as that. That is, there are spaces like Twitter. There's something like the New York Times. So there's bigger groups. But I yeah. So that's that's the first question.

Robin:

So in the background, we have a society we live in, and then we have different places that are controlled by different people. So you can control your home and you can decide who comes in and out of your home. As a teacher, you are nominally in control of your classroom. As a publisher, Like a newspaper, you're in control of what you publish. As a mall, you decide what stores are there, what customers can come in. So the ordinary idea would be that the usual people who control each space should be able to decide who speaks what there, and that other people shouldn't come in and stop them. So if you want to have a discussion in your home, other people shouldn't prevent people from coming to your home to discuss something.

Agnes:

I thought the norm you were talking about is let people talk. Maybe you don't listen, but you don't stop. What I'm saying is if you take a shopping mall or a family or whatever, I don't think we adhere to that norm. That's not a norm.

Robin:

I'm just trying to say you misunderstood what I intended. So I'm trying to be clearer just to say the usual norm would be to let people who control whatever spaces there are decide who says what there. That's what I mean by let people talk.

Agnes:

Just isn't it just analytic that whoever controls the space is going to decide who talks? If you can make them do something, they don't control it.

Robin:

So that's why I'm trying to highlight the coordination. So let me give you some other examples, because I think in your reluctance to sign petitions, I think you are sensing this concern and weariness people have of certain kinds of coordination. So for example, if we think of a labor market, we might think lots of small employers, lots of small employees, a competitive market for jobs, each employee should Each employee should be able to pick whatever job somebody will offer them. Every employer should be able to offer jobs for people. And then it's very decentralized and there's not much coordination. And then people have noticed that sometimes there's a big company, maybe in a small town, and then they can coordinate more to set a policy of the company that will then be at the disadvantage of local workers because of its ability to coordinate. And that seems illicit. But of course, we have other reasons why big companies should exist, so we're not wanting to stop big companies. But then people say, oh, well, we need a counter effort of a union so that they can also coordinate to be an opposing force to the coordination of the firms. And now maybe it's better because we have two forms of coordination, but they're both kind of worrisome. So if we didn't have the big employers, lots of employees all coordinating would be more of a problem, you know, in terms of, you know, so, you know, in the past there were cobblers would form a cobbler, you know, and butchers would be in a butcher's basically coordination and they would all decide what together to do and those were problems. I guess another example is in politics. Many have observed that a potential thing that could go wrong with democracy is that 51% of the people would vote to sort of take everything away from the other 49.

Agnes:

Or even 90-10.

Robin:

Right. For example, right. But that's a sort of failure motive. democracy, and so that's a reason people have been wary of democracy. Now, it doesn't seem to have happened nearly that bad, and plausibly it's because we have these norms that make us wary of going that far. People have proposed, even in the recent election, like, hey, those billionaires, let's just take the money, there's a lot fewer of them than there are of us, and other people go, that doesn't seem right. But that's not freedom of speech. No, no, it's about coordination, so that's what I was, but I was saying the problematic issue in Cancel culture and freedom of speech is the coordination, that is. If it were just individuals deciding they didn't like something and then not going to a lecture, not buying a book, not reading a Substack, that seems much less problematic. If it happens that almost everybody doesn't like something, they can hardly get any viewers. Like if somebody makes a movie and nobody likes it, I think that's fine. I don't think They don't even like it because it offends their morals. I think that's OK. But it's the coordination that's more worrisome because then even if many people did want to listen to something, it can be shut down through the sort of shared institutions like a university or platform like Twitter, etc.

Agnes:

So let's say that There's a relatively small group of people who believe that the earth is flat. And then there's more people who believe that the earth is round. And let's say the people who believe that the earth is round coordinate to shut down the people who believe that the earth is flat. They coordinate in the following way. They try to produce really good arguments and books full of explanations. And when the flat-earthers are willing to speak to them, what they try to do is persuade them out of their view. And so they're taking out a project. We're trying to eliminate flat-earthism. And they coordinate to try to eliminate that view. And it's the many against the few. I think all that is fine.

Robin:

Right. It's a certain kind of coordination. Right. It's a certain kind of coordination that I'm worried about, not all possible coordination. Right.

Agnes:

So that's really the point. It's not about coordination. It's about a kind of coordination. And that's why it's relevantly a question about free speech. We have to specify the kind of coordination. Um, so what, for instance, if I use your fear of losing your job to get you to say the words, the earth is round, then the thing that is pushing you, the kind of pressure on you, isn't relevant to the question, is the earth round? That is, whether or not you will keep your job is not a consideration in favor or against believing that the earth is round or flat. So that seems to be the, to me, that's the issue, is that there are pressures that are not rational pressures that can be brought to bear on belief.

Robin:

That is indeed an issue. But I fear if we try to define the problem that way, we're making it way bigger and less tractable. So I think there is this shared concept of free speech and a norm favoring it. And I think it's more narrow than wanting to promote all sort of good thinking and discourage all thinking that isn't based well on evidence or something. That just seems way bigger than this issue. I think people have a sense that people should be allowed to speak And even if they aren't giving very good arguments, even if they're being rhetorical, even if they're being demagogues, even if they're being orators, we should still let them speak.

Agnes:

So, like, you said that earlier, and then you're like, people have this norm, and then I gave you some arguments that we don't have any such norm. And then you said I misunderstood you, and what you really meant was that people who control an area shouldn't be interfered with by people who don't control it, which I said was be analytic part of the definition of control. And I still, I'm just, I'm not sure you're saying anything when you say the words people should be allowed to speak. That is, I just don't know what you're talking about. Whereas that's why I substituted a different claim, which is that there's such a thing as believing things for the right reasons. And that there's such a thing as trying to, like, encourage changes of belief that are not driven by the right reasons. And that cancel culture has to somehow be involved with that second thing. Because the thing, just let people talk, that's just not a thing. There is no such thing.

Robin:

I was trying to make analogies to other ways that we are wary of certain kinds of coordination. And I was claiming that the issue, for me at least, is being wary of certain kinds of coordination with respect to cancel culture.

Agnes:

I gave a characterization of what the certain kind is. You didn't. It's fine if you want to give a different one, but you need to give some characterization. My characterization is coordination to induce non-rational changes in belief. If that's not the evil one, then what is the evil one?

Robin:

So think of the example of a cartel or a union. you know, these are familiar examples to economists, they are the kind of, you know, problematic coordination that works. So if you have a bunch of employees, say, that coordinate, say, to create some educational classes where they can get training, or maybe coordinate to sort of share information about what job opportunities there are, or maybe share rumors about which places are good to work for, those kinds of coordination don't seem particularly problematic. Similarly, if firms coordinate to train their workers together, if they coordinate to have people know about their kind of product, if they coordinate to be able to share information about problems that might occur, those are also not problematic. What if you list the ones that are not problematic? You're not telling me what is ... I'm just trying to point to the other ones.

Agnes:

The problematic ... What is it that the other one's having in common? That's what I want to know. I don't want examples. I want to know what is the problematic thing.

Robin:

Again, we have to start from examples to try to think of ways to characterize them in common, but... Okay.

Agnes:

You've given a bunch of examples, so now tell me whether the cartel, the union... Okay.

Robin:

The cartel and the union are basically saying, people in this group You may not associate with other people without the approval of the group. That's the key thing. The group puts limits on who can associate with who, how. In turn, that's the key coordination. So a union says, you can't go to a firm and take a job they offer you. because we need to all agree together on what jobs we're going to take for them. None of us are going to take any job unless we have a deal collectively with them. That's a coordination to raise the wages of the workers. And similarly, a cartel would have a coordination like that to raise the prices the customers would pay by the cartel saying, you can't just go sell whatever you want to anybody you want. You have to abide by our central rules about who you're allowed to sell to. And the analogy for speech is saying, you know, you have to follow our central rules about who you talk to, who you listen to.

Agnes:

It seems to me that most groups do the thing you've just described, namely place limits on who can associate with who, how. Country clubs, obviously, like- I'm talking about outsiders.

Robin:

What outsiders you miss are, not the insiders. So a country club is a group of people who will agree to associate with each other in the club. But if you said you can't- You can't bring outsiders in here. No, that's not- That's what outsiders use it. But that's not what we're talking about. Say you had a country club that says, we won't allow anybody in here who associates with Jews, and there haven't been country clubs that could be like that. They said, your association with those people makes you illicit in our group, because we're a group where they're trying to coordinate on repressing that, opposing that other group, and therefore we're going to place limits on what other associations you can have outside of our group.

Agnes:

So it's something like the group that tries to be your only group. Like what you like is people can form whatever groups they want, but those groups have to not, they have to not make the claim to be the whole background culture or something like that. Because a culture wouldn't do that, right? A nation does, like we're, you know.

Robin:

I mean, we don't let people travel to other countries, even if they are from our country, and some countries allow dual citizenship.

Agnes:

We might not, for instance, let... Okay, sorry.

Robin:

So, if we... So, for example, the US wanted to punish Cuba after the Cuban Missile Crisis, and so it said Americans were not allowed to go visit Cuba. I would also disapprove of that, that I would say that's somewhat of an illicit way the United States is trying to punish what it sees as an opponent, a rival, by preventing U.S. citizens from traveling to Cuba.

Agnes:

But say that I say to my children, You know, you can't, any of your friends who play video games, they can't come to our house. Say I say that. From your point of view, that's like the country club case where they just can't come to my house and you can associate with them in school. But my kids might say, yeah, but that's not really a very relevant distinction because I can't really be friends with anyone that I can't bring to my house. And similarly, the people who oppose the country clubs, people who say, you can't join our country club if you ever talk to a Jew or something, they might say, for the same reason, there's something problematic about not inviting the Jews, not allowing the Jews to be part of the country club. And so I'm not sure that I see such a principal distinction between whether what you do is prevent some people from coming in versus you try to legislate the associations that happen outside.

Robin:

So let me highlight a factor that's relevant in these problematic cases, but not in many of the other cases you might be thinking of. And that's sort of the importance of scope of the coordination. So say a union negotiating with the firm, it's important that the union contain a large fraction of the workers who might work for the firm. If it was 2%, it just doesn't work. So there's this attempt to get a large fraction of people to all be on the group saying, you know, we're not going to accept a job unless we all do. Similarly for a cartel, you need all of this, you know, a large fraction, even most of the suppliers of an industry, it's like OPEC, you know, back in the day, restricting oil prices, they needed most oil producers to be in. And when new oil producers showed up who weren't in, that was the fall of OPEC. And so that's also the cancel culture thing as well. You need a large enough group of people to be not willing to associate with this person for it to have a substantial effect. If it was just 1% of employers, it wouldn't take you because of your problematic features, that wouldn't be much of a problem. It's when 75%, 90% of them won't hire you, then it has strong force, these threats. And that's the coordination that we're talking about, a group of people try to become a large fraction of all the people someone could associate with, and then using that strategic power to impose terms on them. And that's what a boycott is, and that's perhaps what even a petition is trying to create the impression of. That there's not just a few people have this opinion, but a large enough fraction that they will be able to have real pressure.

Agnes:

So like monogamy in your view is kind of like cancel culture or whatever, or this thing, whatever we may name, a bad kind of coordination because say we're married and I'm like, no, you can't sleep with other women, not only not in our house, it's not only that you can't bring them into our house, I don't care where you go, you just can't. I'm trying to limit the number of interaction partners you can have and to legislate that it can only be me. So you would object on the same grounds then to that?

Robin:

So the concepts I'm trying to apply here are the familiar to economists concept of market power. So that, you know, is relevant as what's the relevant market. So if oil producers you see have a, you know, conspiracy together to limit oil prices, that will you know, have power to the extent that people need oil and they just don't have other substitutes. So if the solar, you know, wind, wind power isn't part of the conspiracy, that's not a problem. That's just not much of a substitute for the rest. If we have, you know, a pretty competitive market for who you marry, who you have as a partner, then we don't have the distortions of market power happening in choosing a particular marriage partner. Now, after a while, once you have become attached to your partner, then there is a lot of market power. And so that's actually an issue in, um, And the same way for an employer and employee, once they had a long-term relationship or friends, then there is more market power. And then we tend to think of, as economists, that that's well enough disciplined by the earlier market, when you had lots of choices, that you choose where you get locked in.

Agnes:

You might think that like, In a culture with social monogamy norms, actually the earlier part was also subject to a certain kind of coordination. That's what norms are, is coordination, right? And so you never really had the choice. Because all the women that you would approach would have said, do you know? And they had the whole force of culture behind them and social norms.

Robin:

Right. So, if basically very few people were interested in polyamory, say. And then early on in the mating market, you were looking for partners and there just were very few who satisfied that criteria that would limit your options in that direction. There just wouldn't be very many people who wanted that. That would be different than if lots of people were attracted to it, but there was a effort to repress that. And people weren't allowed, say, on dating apps, say, to specify that that's what they wanted. If they said that's who they were, they'd get fired from their jobs. They wouldn't get a security clearance. Those are the kinds of things that people complained about. So for example, with respect to, say, homosexuality, there was a time when Not only did it seem like not very many people were interested, but there was active repression of anybody even trying to say, hey, I'm interested, who out there might be interested in this? There was a coordination to make that very hard. People wouldn't get jobs, they wouldn't get security clearances, all sorts of things they wouldn't get if they publicly announced that they were looking for that sort of thing. So that was a way that society did actively coordinate to suppress that subgroup of people.

Agnes:

before that they coordinated to suppress them in a much more effective way, which was such stringent social norms that it would never even occur to anyone to really even classify themselves that way. It's like to, you know, suppress like women's desire for freedom Once the women have a little bit of a taste for freedom, you have to engage in active suppression, but the social norms are a way to do that without any active suppression. It just seems like the idea of limits on association is just very deeply built into culture. That's part of what culture is.

Robin:

I mean, part of the origin of, say, free speech was a few centuries ago when there were some people called liberals who wanted to talk about, say, birth control or women's voting or other sorts of options, and the culture you know, most people didn't even notice there were options, then some people wanted to call those options to people's attention and to advocate for them, and then they faced repression. That is, and that's the origin of much of modern free speech. Discussion is, in fact, those efforts to call attention to the possibility of changing norms, to question norms, and do that via speech. And then there was active repression of that speech as a coordination to prevent those options from being things people are aware of.

Agnes:

Right. But it seems to me that if we're trying to defend it, defending it by saying there shouldn't be any limits on association is that seems to me to just be a much broader category, which is very, very close to the idea of culture itself. And so it seems like what we ought to be defending is not anything about association, but something about the process of trying to arrive at the truth. Like you said early on, right, you said something like, well, if people are allowed to say what they think, then we'll all learn more. Somehow that got alighted from the discussion and now it seems like it's a discussion about people being able to have as many trade partners as possible or something like that. And that seems to me to be a slightly different issue. from the issue of how do we enshrine the value of the pursuit of truth? Because at least it's conceptually possible that a society that is in some ways very repressive, in some ways very repressive, could still be oriented towards the pursuit of the truth. It would just be repressive about stuff that's not relevant to the pursuit of the truth.

Robin:

So to stand back for a moment, this is a common thing that happens in the world, and we're struggling with it. The scenario is that we have some rough norms and patterns of behavior that we like and some that we don't, and that we give some names to, and that in history, people have had examples that were categorized as, you know, in fit with his name or not. So free speech is an example of that. You know, when people grow up, they hear many things and those are called examples of free speech and other things are said not to be. And we, in some areas, have sort of the perception that many people support that, maybe in other areas less. And then we want to think about it more clearly. And we realized that the examples we've been given, you know, inherited, don't actually very clearly fall across, you know, a nice clean line. And so we're looking for lines we could draw that would help make sense, not only of the examples we've seen, but maybe of the rationales that people have offered for why there should be such a distinction. So I was pursuing the idea of this trade partner analogy as the way to characterize this, but I'm not that confident in it. And if you have a different proposed way to characterize, I'm happy to switch over for a while.

Agnes:

They gave it to you earlier.

Robin:

But it's not very actionable. So that is, we need a concept of free speech such that we could agree together on what were or were not examples of free speech. If you say pursue truth, that's free speech. We, as you well know, we don't know how to coordinate on which particular examples are or are not us pursuing truth.

Agnes:

Well, I mean, so as I told you, I wrote this op-ed in which I explained how to apply this standard to a particular case, namely petitions. And what I said was, in the case of a petition, the petition is more forceful if it has more names on it, but having more names on a document doesn't make it more true. And so that would constitute a non-rational form of pressure. Now, there are alternatives to petitions that are maybe not so far from them, like you could do a poll or something like that, right? You could have good polling, And then the fact that the majority of people believe something is at least a little bit of a reason to move in that direction. So there are things in the vicinity of petitions, but as I say, if you wanna say, why is a petition bad but a poll good, I can use the resources of my let's pursue the truth theory and explain that some things are considerations in favor of believing something and some things are not considerations in favor of believing that thing. And that's the relevant distinction. And when you try to get people to believe something on the basis of something that is not a consideration in favor of believing it, then you're employing non-rational pressures. That's what I think cancel culture is.

Robin:

If you think about oration, just the fact that some people are better at putting words together to make things sound nice and eloquent, that's also not directly truth revealing. I agree. So bad. I don't see how our disapproval of that helps us decide when people have crossed a line for some sort of... So we're trying to draw a line for, say, cancer culture, and we're going to allow oration. We're not going to ban oration from the world.

Agnes:

I would look to ban anything. So this gets to the second part of it, which is like, suppose you care about rational argument. What's your method? Which things are you going to ban? You're going to ban zero things. What you're going to do is make rational arguments in favor of conclusions and try to persuade people. That's what it is to believe in it. So for instance, if somebody, if I thought you were using oratorical tricks on me, which I think you pretty much never do, But if I thought you were doing that, as a fan of rationality, what do I do? Do I shut you up? Do I ban you? Do I lock? No, I give you arguments and I say, look, here's the thing you said, but that's not actually a reason for believing the claim that you're making. So please don't talk like that, because I want to believe it for the right reasons. That's my strategy. That's the only tool in your toolkit.

Robin:

So first, I do think signatures on a petition contain information, and it is an information analogous to a poll. It's farther up the percentile distribution. So in many topics, we have this idea that There's the median person and how much they care, but then there are the people who care a lot, and often they have disproportionate influence. Many political changes over the years, say, have happened because a small minority had very strong feelings about it, not because the median person cared that much. And so it's valuable, in order to guess which way these things will go, to get a sense of opinion of extremists, not just opinion of the median person, and petitions are a way to collect information on extremist opinions, because You know, it's somewhat of an extreme act to go sign a petition. So they are informative in that way. But yes, they also sometimes have the implicit threat that they're not just trying to inform you about their opinions. They are perhaps getting ready to coordinate to pressure people to following their point of view. Of course, a poll could do that as well, depending on the context.

Agnes:

Right. So so why are they? objection that people made that I'm interested in to the petition piece is that people need to know what to pay attention to. And what the petition does is draw attention to a set of claims and the fact that many people believe something might be reason to pay attention to it. I'm not sure what to say to that argument.

Robin:

But I mean, I think we're closely in agreement here in the sense that in this context, we can see two possible social functions that the petition might serve. One is the positive one of giving these sorts of information. Then there's the negative one of being a focal point for the sort of collusion that I was concerned about. But if not for that second kind, you shouldn't have much concern for it, right? If it could only do the good things, it couldn't do a bad thing.

Agnes:

Absolutely. I'm just doubting that what's bad about that collusion is best cashed out in the way that you had of cashing it out, namely that there's some limit on the number of trades people can make with other people. I think the collusion is bad insofar as it attempts non-rational influence on belief formation and for no other reason. It's irrelevant how many trades people can make. So there's a question, why is the collusion bad? But I think you're right that I could imagine a world where your point about, like, petitions give us information. I mean, this is why I said a poll, because it's not just that the petition attempts to, you know, the petition about Scott Alexander was attempting to pressure The New York Times, right, to be like, think of all these people who are gonna Not just cancel their subscriptions, but try to get other people to cancel their subscriptions. That was clearly behind the ethos. That's Ben Levitt on Trape.

Robin:

That is the collusion of sort I was talking about, right?

Agnes:

Exactly. But the thought was, but the reason that puts pressure is you're going to lose money.

Robin:

Right.

Agnes:

And you don't want that. And so you shouldn't do this thing. Now, the loss of money isn't the reason not to do this thing. The reason to do the thing is it would be wrong to reveal his name or whatever. That is the actual, notional consideration is that. But suppose that the New York Times could just get, you know, was just unsure. Is it wrong or right to reveal Scott Alexander's name? And then they wanted to get information on that. And if the petition were presenting itself as a way of allowing the New York Times to get information on how many people think this is wrong or how many people, you know, really think it's wrong or something, Yeah, I agree that might be okay, I just think that that's not fundamentally what it is. It's not just that they're trying to put that pressure, it's also that... people put a lot of pressure on one another to sign the petition. That is, it's not only extremists. It's people who want to signal that they're part of a certain group. I think that was very true of the people who signed the Scott Alexander thing. It's like, if you're, even in Scott's response to me, like his thing was, oh, she said she supports me, but she didn't sign the petition, right? So how much can she really, she signaled that she doesn't support me. In fact, they never said I supported him. I just said that I believed that it was true that they shouldn't reveal his name, right? It's not the same as supporting him. But the question, are you a supporter or not, then becomes the question for signing the petition. And so it's collusive or pressuring, non-rational pressuring at both ends, both in who signs and in the effect on the person who receives the set of signatures.

Robin:

So norms and even laws, such as crime laws, tend to work much better when they're trying to suppress something that's rare anyway. Something that's ubiquitous is much harder to deal with. So, for example, most crime laws are trying to suppress things that don't happen very often, even jaywalking, or stealing, or fraud. Because they're rare, we try to draw a line around them, and then we make rules that try to suppress them, and then we can be rather successful with that. But obviously there are lots of things that go wrong in the world. People are often not respectful of each other. They, they insult each other, things like that, but they're, they're so ubiquitous and it's really hard to draw a line that we just don't try to use law or even norms that much to try to suppress things that are just ubiquitous. So that's why I would recommend the cartel view here, because in fact, cartels are rare. That is. You know, most workers can't be bothered to make a union. Most industries can't manage to make a cartel. You know, it's just takes a lot of work and trouble to set up these coordinations. Most people don't get canceled. And so the key canceling problem seems like it's a rare phenomena that is somebody has to bother to set up some sort of a, you know, discussion or outrage about something and gets the momentum going and then something happens. But it doesn't usually happen to most people, even when they say mildly offensive things. And so that's why it seems there's the potential for having a norm that would sort of draw a line around it and try to suppress it because it's rare. Whereas you're talking about just everything people does that isn't fully truth oriented. And honestly, most conversation and most thinking isn't very truth oriented. And if we're trying to fix all of that, we're just not going to have much luck, at least with a you know, a particular norm with a boundary that trying to go for a particular problem.

Agnes:

I guess we're just asking different questions. It's like, you want to know, what should we ban? And I'm happy with the thought, well, that requires a lot of practical expertise. Like, neither of us has really enough of it. And there's people probably who specialize in that. And I'm pretty happy to leave it to them. But maybe they want a little philosophical advice every once in a while. But I'm not that worried about the question, what should we ban?

Robin:

It's not banning, it's having norms against. Right.

Agnes:

What is wrong? What is the thing that is wrong? What is the thing that is going wrong? What is the thing that is going wrong in a petition? And I don't think limits on who can associate with who is a plausible answer to that question. And when you say, oh, cartels are not common, I just think every family is a kind of cartel. And if I put restrictions, even just saying that my kids, say my kids can't leave the house past a certain time at night, I mean, that limits their set of associations. We're just doing this all the time. Limiting who can associate with who is the basic backbone of any kind of

Robin:

But I tried to point out how I'm not talking about all possible limits on associations. I was talking about a particular kind that's rare, this kind like a union.

Agnes:

But when you say a kind like a union, and then I want to know what kind is that? And it kind of seems like maybe it's just because there's a lot of people involved?

Robin:

It's one that needs to have a large fraction of the relevant people as part of the association. Group, that's what it is.

Agnes:

Family, it's also true that 100% of your family members tend to have to be involved in enforcing the family rules, or at least the parents. So that's a large fraction, a large fraction of your parents are going to be involved.

Robin:

I'm invoking a standard set of economics theory here by which we understand when market power is a problem and when it's not. And market power isn't usually seen to be so much of a problem in the family, so that's why we can not be so worried about it.

Agnes:

I mean, I can't tell, you know, maybe then there's some secret reason that economists have that you haven't been able to explain to me that would explain the difference between these two things. Or maybe the economists themselves just don't really know why market power is the problem and that's why they can't tell me why it's not a problem in the family. But, as I said, the thing that I, the consideration that I brought to bear actually does say what the problem is, right? So that to me is progress, is picking something out that is, I'm not sure why limits to association are a problem. As I see it, limits to association are a condition on culture. non-national pressures on belief formation, I can explain why that's a problem. Because if your beliefs are formed by non-national mechanisms, you're going to have false beliefs. False beliefs is bad because truth is a constitutive name of belief. There you go. That's an explanation.

Robin:

It's also an explanation why 90% of what everybody's doing is wrong. Sure.

Agnes:

I'm happy to oppose more things. As I say, I'm still trying to figure out what to ban. I'm happy to try to correct all that stuff too. A lot of the stuff we do wrong is wrong for similar reasons. That's okay.

Robin:

It seems like you're just trying to recruit this particular discussion of free speech to another agenda that you have. And that seems somewhat unfair in the sense that these people who are trying to talk about free speech, they aren't against more generally trying to be rational about things. Many of them are rationalists and have that very specific label, but they are saying, look, we are in a culture that has a tradition of promoting a certain kind of thing that's been called free speech. They think that's worth reserving and protecting. And in service of that, they would like to, as Scott's post indicated, be more clear about where to draw the line around this concept free speech that they want to coordinate on protecting. But you're basically throwing that away. You're saying, well, that's not interesting compared to all the other ways that we also are not very logical or rational about things, and therefore it seems like you're just throwing away the whole phenomena that they're trying to deal with. It's a particular norm in a particular context that they are trying to be clear about, not just generically doing well in thinking.

Agnes:

Well, when we say they, the people we're talking about are me. That is, I wrote this piece. And Scott is, in part, responding to me. So I'm kind of the they here, right? So what questions I was trying to raise, that's part of the issue here. And I made a very concrete argument about a very concrete case and where I think we should draw the line. And I'm happy to talk through Scott's cases. I think I roughly agree with you. I think I might draw the line a little bit earlier. I think you were at like A10 or A11, and I was at like A7. Um, um, but, um, but I can explain why I would draw the line where I would using my criterion, right? So, um, you know, um, um, um, um, he has a list of things where you're, like, offended by the fact that there's a podcast that promotes, uh, pedophilia, and you could just not subscribe to it. You could, um, uh, uh, not subscribe because it, um, There's one episode of that kind, etc. And, you know, he has a bunch and I would draw the line where you post on social media. That was gross. I can't believe they would do this. Where, to my mind, what you're doing there, the way I interpret that speech act, This could be a misinterpretation, but the way I interpret it is you're trying to put non-rational pressure on people to be opposed to this thing, to this podcast for not the right reasons, that is through disgust or whatever, instead of just making the case against it. That would be the way to try to influence the beliefs of other people.

Robin:

I draw the line. you're well aware that you're a very good writer in the public sphere and that your writing is quite different than a mere straightforward listing of art premises and conclusions of arguments. So you are accommodating at the very least all of these irrational habits and styles of people in your writings. I mean, why wouldn't your norm here also want to disapprove of that?

Agnes:

I think that's a good question, and I think that talking in such a way that people find it engaging to listen to you is not necessarily using oratorical methods of persuasion, though it can be. So here's a thing that people say to me a lot, and I think they say it more to me than most people. They say, I really like reading the stuff you write, even though I disagree with most of it, or even though I disagree with a lot of it, or even though I usually disagree with you. Actually, my niece recently reported sort of shame, almost ashamedly to me about somebody else. She said, she told me she likes reading you, but she usually doesn't agree with you. Now for me, that's a really good sign, right? It's a sign that I'm able to engage people without putting my finger on the button of like trying to convince them of my views through those methods I have of engagement. And so that's an art and I partly rely on people reading my writing and giving me feedback to make sure to steer me away from that danger zone and editors above all. But I think it's a real consideration and one I take seriously and maybe sometimes I stray over onto the side of using my powers for the forces of evil and that's a bad thing and I shouldn't do it.

Robin:

Even more extremely, you might argue that much of fiction or even music illicitly persuades people of things through its arts, its, you know, the beauty and compellingness of the art form of the movie or the song. And therefore we should be suspicious of all of that compared to simple, straightforward, logical presentation of, you know, arguments.

Agnes:

I agree and I think we should. To say that we should be suspicious of something is not to say we should never engage in it. Those are two different claims, right? So, and I think music and fiction are different. I've written about fiction and exactly why we should be suspicious of it. Namely, I think fiction has a negativity bias. It is designed to show us all the bad stuff in life that we don't otherwise want to look at. Um, for that very reason, I mean, I think life has a bit of a positivity bias. So there's something maybe, you know, there's some kind of informational value in fiction in that way. Um, but it also, it is itself distorted. So that's the case of fiction. I think that is, I think these suspicions exist and they're worth taking seriously. In the case of music, I think it mostly affects your mood and it's fine to say, well, to affect someone's mood is to affect their beliefs, but it's really short term. So your mood just goes back later. Like the music just kind of washes. It reminds you anyway, kind of washes.

Robin:

There would be associations of the mood to the topics discussed in the song. So, for example, if it was a patriotic song, then you might have a positive mood toward the nation induced by the song talking about the nation.

Agnes:

I was not even thinking of music with lyrics. But right. So if you're thinking about music with lyrics... Most music does have lyrics. No, that doesn't seem true. Classical music is a very large fraction of music and most of it doesn't have lyrics. So lots of music doesn't have lyrics, lots of music does have lyrics. They're just both kinds.

Robin:

So let's talk about rationality norms then. So maybe that's the key thing. So we might agree we would like people to be more rational. And then some communities have developed norms to promote rationality. But norms are enforced via social pressure.

Agnes:

And so... I think you can enforce norms by explaining why the norms are correct. Like in my family, we have norms. And when my children try to flout those norms to as much degree as possible, I try to explain to them why the norms are correct rather than using forms of pressure.

Robin:

Most academic fields, for example, have norms of what a proper research paper looks like, say. And most of the time, you know, people aren't explaining why those norms are correct. In fact, people rarely do. They just have the norms and then they enforce those norms on, say, submissions to journals. They reject them if they don't seem to satisfy the norms. And certainly even media like newspapers, they have norms for what a good news article is supposed to look like. And they enforce those norms on junior and senior journalists. And they have some rationale sometimes in terms of truth orientation, but they just seem to also have some local inertia. Different groups all claiming to promote truth has very different norms about what proper methods are.

Agnes:

But I think there are different ways that you can respond. I think it's fine to take the norms as like a starting point, but if someone has like some good reason for opposing them or for wanting a justification of them, then one thing you can do is just use, you know, social pressures to shut them up. And another thing you can do is explain why the norm is a good norm or take into consideration their reasons for thinking that we should have a different one.

Robin:

Do you know how long we've been talking?

Agnes:

Not quite an hour, but we can stop.

Robin:

Well, one more observation is that courts have rules of evidence. That is, they have rules about the kinds of things you're allowed to say in court to persuade a jury or judge. And I guess you would be against those rules. That is, they don't... In each case... I mean, obviously, someone could say something on one side of the case, violating those rules, and we could say, then the other side can explain why that was a bad thing to do, but that's not how we do it. We just don't allow certain kinds of things to be said in court that don't satisfy the standard rules of evidence.

Agnes:

So the thing that I think would be bad is if we tried to get someone to believe, say, that the standard rules of evidence are correct for reasons other than reasons that would make them correct. For instance, by saying you're an idiot if you don't follow those or whatever. But no, I'm not generally opposed to the existence of rules. And in general, people, in order to coordinate, have to follow rules.

Robin:

But I was emphasizing, it's the same as the idea of limits of... So I was talking about groups of people having a norm about certain kinds of things being cancel culture, and then we'll try to... have a norm that you just don't do that and we look askance if you do that. But you say, no, that's pressure. That's not persuasion.

Agnes:

But instead of looking askance, there's another thing you can do.

Robin:

But in courtrooms, again, if there's the rule, you can't there's evidence you can't present and you're not allowed to present and you can't argue against the rules of evidence in a particular trial. You just can't present the evidence.

Agnes:

Right. So I guess I think in the in the context of a courtroom, The forum for arguing against the rules of evidence isn't the courtroom, it's, you know, some very far removed thing from the courtroom. But if our method of dealing with whether those should be the rules of evidence is just shaming anyone who ever questions them, that's a bad method.

Robin:

But then why couldn't we have rules against cancel culture? And then this is the sort of forum where we argue about that. But then when we have cancel culture show up, we just impose the rules that we don't discuss in that context for reasons for it.

Agnes:

Yeah, I think that actually that would be fine if we had the other thing. That is, if we had any sense that we had arrived at some set of rules using a good procedure. But as far as I can tell, the- It's true for law as well.

Robin:

I mean, the procedure we use to come up with rules of evidence is not very inspiring to me.

Agnes:

Well, then I'm not happy with that either. I mean, I'm fine with being unhappy with more things, but I think that if we had the sense of ... So maybe the point is this. You say, okay, we want to have a system which draws lines. And I guess my thought is we're at a very early stage here of this conversation where we're not, I don't think we've even settled what is the basis on which we should be drawing these lines. And a lot of people have strong intuitions about exactly where the lines go. And they're sure that they know what this free speech thing is. I just think we're way further back. We don't really know what we're doing. And if we were to think this, come and think this thing through, then we could institute some kind of a system and then we could enforce the system. And when the people, on whom we were enforcing, and we're like, wait, why are these the rules of evidence? Why are we doing this? Then we'd have something to tell them, instead of just, aren't you in favor of freedom, or aren't you friends of Scott Alexander, or whatever, right? Right. Well, right now, that's what we've got, and the thing we've got does not qualify as free speech.

Robin:

If we've had this for several centuries go out, then plausibly it would take several centuries more to get the sort of consensus you're expecting, and we just can't wait several centuries to take action. Why not? well, we have to make choices now about canceling or things like that.

Agnes:

Oh, sure. But look, we've always had to make choices on the basis of poor theories and poor intuitions. And maybe then my thought is, yeah, I'm not that interested in weighing in on, should we make this bad decision or this other bad decision? The much more interesting question is, let's try to make progress on the real issue so that we can move towards making the good decisions. I'll let you have the last word because we should stop.

Robin:

Well, thanks for discussing this.

Agnes:

Okay.

Robin:

All right. Till next time.

Agnes:

Yeah.