Good night, Agnes.
Tonight, we're going to talk about free speech, which is something other
people have been talking about and you've given a talk on it recently and
written on it. And it's something I've written on in the past. So, let me
start out with what I think are relatively minor claims about it. And then I
think we'll get to your bigger claims.
So that to review the usual concept of free speech, the usual argument for it
is to say that people should be allowed to have an open space where they can
make whatever arguments and claims they want without much regulation,
restricting the sort of arguments and claims that could make because of in
competition, of listening to these different arguments and complaint, claims
that we will better sort it out that way, than some sort of central regulation
that limits which arguments and claims we can make. So that's an argument for
both legal free speech in terms of government regulation should be low, and
also for norms of inclusiveness and tolerance of speech.
So, in these terms, on my first claim, is to say that what we really then mean
is free hearing, not free speech. That is, the argument suggests that people
should be allowed to listen to whatever arguments and claims they want. Not so
much, who's allowed to say things but what you're allowed to listen to.
So for example, in the US, free speech legal interpretations, we might say
that convicts aren't allowed to speak or that foreigners aren't allowed to
speak, but that citizens in good standing are allowed to speak. But that seems
the wrong way to do it, you should instead say, well, citizens in good
standing should be allowed to listen to whoever they want. And if that
includes convicts or foreigners, then they should be allowed to listen.
So, that would be my first correction. And it would be a substantial
correction, at least to the current legal interpretations of free speech. That
is according to the argument, people should be allowed to listen to what they
Is freedom consistent with the existence of incentives like, so, suppose that
you can freely speak in a certain space but there's going to be a cost to you
for doing so. Maybe a slight reputational cost, or monetary cost to enter the
marketplace of ideas. It like, would it still count as free if there are costs
associated with participating? And, we can ask on the hearing side, right?
Like, are you are you freely hearing if the speakers have to make that kind of
calculation or to decide whether to speak?
Well, like we say, a free market in apples or something, one isn't
recommending that everything literally be free, but are recommending that we
don't add artificial costs to whatever more natural costs already exists. So,
if it's expensive to make apples, we're not requiring that people sell apples
below the actual cost of making apples. But we don't want an artificial cost
to be added to the cost of making apples or the cost of consuming apples.
Because like a lot of what I see as like motivating the free speech wars at
the moment, is that there's people who don't like the natural costs. At least
if we think of social outrage, it's pretty natural.
We should discuss that. But, you know, but that's a separable issue, I think,
from the free hearing versus speech issue, right, which is basically just
about what exactly should the liberty apply to? Not talking, listening.
And my second claim is also the same sort of framing, which is to say, well,
what we really want is people to be able to sort of offer whatever sort of
persuasive evidence and information they have, through whatever channels they
have. So we wouldn't, for example, only want to require them to speak in
English, for example, we would allow them to speak in whatever languages they
might want to. We wouldn't say, limit them only to speaking and not to
writing. So we would want an encompassing world of all the different channels
And this is, for example, say the argument for protests. We might say,
protests can convey information of a form that mere letters to the editor
can't because they're showing that you're willing to pay a cost to be out
there protesting. And that conveys information that isn't just in the words of
the protesters. And that's commonly accepted as a free speech sort of an
argument. So it's not just the ability to say words, but also the ability to
convey information in other forms, at least if there's not other bad side
And so, on that basis, I have said, well, then we should be able to freely bet
on important topics, because our bets convey information that our mere words
cannot as well. Our willingness to bet the amount of bet, the actual amount
that other people bet against us. That's all information that we get about who
thinks what and for what reasons that we can't get as easily out of words.
So, a similar sort of thing, be open about channels. Now, of course, if you–
you might be able to communicate something that you couldn't otherwise
communicate by carving some words into my flesh, but that's going to have
costs for me so we're not going to allow all possible channels, no matter what
the other, you know, harms that result. But we still might have a broad
liberal concept of the various kinds of channels of information we might
I'm having a little bit of trouble sort of seeing what kind of purchase the
free hearing as opposed to free speech model is going to give you. So like, is
it like, could you give me like a case where on the free hearing model, we
would protect it, but on the free speech model, we wouldn't or vice versa?
Like, would it be like repetition? Like if someone's already said something
then the public has heard that idea, and so then we're allowed to suppress
So, in US presidential elections, foreign governments are not allowed to
speak. And in fact, if you somehow facilitate a foreign government speaking,
you are guilty of treason?
So that's a very vivid and clear example, where you might say, well, foreign
governments should be allowed to speak. If people want to hear them, people
should be able to hear foreign governments speak about domestic elections.
So the idea is that it becomes rhetorically easier to argue for the free
speech rights of convicts or foreign governments, if we posit those is
actually the hearing rights of citizens? That's the point?
Yeah, it's the framing, right? When you think of it as free speech, you say,
"Who's allowed to speak?" If you think of it as free hearing, you say, "Who's
allowed to hear?" And we might say whoever is allowed to hear should be
allowed to hear whatever they want to hear.
Or even on children we might say, children might not have free speech rights
perhaps, but adults might have the free hearing rights to hear children.
I mean, suppose that I don't want to tell you a bunch of stuff. But you have
hearing rights to know what I think, then it seems like you have the right,
you're hearing right is the right to compel me to tell you stuff?
That would be going farther than I intended. So, usually, we're thinking of
regulations of a world where some people choose to speak and other people
choose to hear. And they are matched by, you know, the existing media or
communication structure. And I know you have some opinions on that. But under
the usual conception of people, you know, so if I call you on the phone, and
you pick up the phone, and you hear that it's me, and you continue to talk to
me, then we've agreed to talk. And it's a mutual agreement, either of us can
end that call at any time. But if a third party was to prevent us from talking
to each other on the phone, we might want to know why.
But it seems like, you know, you really, you don't have free hearing, what you
have – or the right to hear, what you have is the right to hear whatever
anybody happened to want to say.
Yes. So, in this country...
You have a right to hear subset of things, right? So it looks like still
conceptually, there's this prior idea of people wanting to say stuff, right?
That's where the concept start. People who want to say stuff, right? And, if
their freedom to say the stuff, right, is also your freedom to hear that
stuff, namely, exactly the stuff they wanted to say, right? But it still seems
like primarily the choices with them, like the fundamental choices with them
as to what to say.
Both parties need to agree for it to happen. But again, the usual wording
focuses on one side, and it can be helpful to sort of realize that you could
look at the other side either. I'm not going to push too much on the name, I
agree, it's both speech and hearing together, but it's still, the legal
interpretation does at the moment limit is about who can speak not about what
you can hear. So the law is focused on the right to speak.
Yeah, but I mean, I guess I think like, suppose you really saw it as a right
to hear. That is not just – it's not just a branding thing so we can get in
the convict speech and the foreign government speech, but actually, you
thought there was a right to hear or right that people have to know stuff.
Right? Then you're going to think like, well, how much pressure you're allowed
to put on people to say stuff is going to depend on the circumstances.
But fundamentally, these people have or people have a right to information.
Right? And that means you have to satisfy that right by giving it to them.
Right? That is you have a duty to speak. Like our rights and duties are
correlates, right? So, if you really thought there was a right to here, then
you think there were duties to speak, which you might think. But it seems to
me fundamentally, if you don't think that then what you believe is not that
there's a right to hear, but just that the right to free speech has like
benefits and harms that fall on hearers as well, and we should like take that
I think we agree that the usual conception is about when one side agrees to
speak and the other side agrees to listen, and whether they should be allowed
to meet up in that way. If I call you on the phone, should we be allowed to
talk or, should the government or someone else be allowed to prevent us from
talking? Or, if I write an article in newspapers, should you be allowed to buy
the newspaper and read it? That's the usual concept.
And there are lots of interesting questions to go beyond that, in terms of,
you know, what information should people be compelled to release. For example,
most courts actually think that pertinent to pretty much any court case,
pretty much anybody can be forced to disclose pretty much anything that the
court thinks is relevant to that case. So we have a very inclusive concept of
obligations to reveal information relevant to court cases.
Right. And like, if you really believed the right to hearing, you would see
that as very much of a piece with, you know, this territory, you'd be like,
"Yeah, that's just the right to hearing over again." It's just the right to
I'm not going there.
Right, right. So what I'm saying, I kind of think fundamentally, you're not
OK, fine. But still, I'm going in two directions about hearing and about bets
basically but that's another mechanism by which we can communicate
information. And at the moment, we greatly limit bets in terms of political
speech and other related speech. Whereas, we are much more open about words,
although, as you may know, the history of free speech in the US is that, you
know, we have the precedent related, say newspapers and books, but then when
radio and TV came along, they did not really apply free speech to TV and
radio, it was a new thing.
And so, the default was that the government could greatly control radio and
TV, and they did greatly for a long time. And so that was a case where they
did not broaden the channels idea to say, whatever channels you could use to
communicate, you should be able to use those channels to spread the
information you have. It was only the particular kinds of public meetings and
books, the newspapers that were protected as free speech for a long time.
So, that's like two different directions I'm tempted to get but maybe just to
start with. So, it seems to me that this question of like, should the
government be involved in regulating speech? That's not the live question
right now. Like in terms of what's, you know, New York Times Op Eds, and
whatever, it's not about that. It's not about government regulating. It's sort
of both sides agree, the government shouldn't regulate speech too much. That's
The question in a way, it's really the free speech people who want more
regulation, as I see it. That is they want their speech to be shielded from
social consequences, right? So they want to be able to say stuff that like
super offends people, and then not experience negative consequences from the
offense. And that's because like, we're living in a very offended time, and
so, it's like pretty hard to– it's pretty easy to say something that offends a
lot of people. And then that's now part of free speech, right? And so then on
the other side is like, well, you just don't want to deal with consequences.
And that's why I was asking you about this like marketplace. right? And, like,
you know, do we have free speech if there are costs? And here, it seems to me
somehow the idea of a marketplace is not doing us that much good. It's not
helping us really, because if we really imagined it as a market, then of
course, there are going to be costs, right? And a free market isn't free in
that sense. But I think that the idea, the defenders of free speech, their
idea is really that there shouldn't be costs, that is that people should be
shielded from the potential like, you know, reverberations of their speech in
some way, like not fired, for example.
So, you know, I think you and I will get into difficulties trying to sort out
say exactly what other people are saying, because there are so many of them
saying so many different things. So the simplest thing would be for us to say
what we think.
So, I would make an analogy to say religious tolerance. So, you can imagine a
society in which there's a dominant religion, and then people who have an
alternative religion are shunned. That is the people refuse to rent rooms to
them, serve food to them, do business with them whatsoever, because they're of
the wrong religion, and they're just going to be shunned out of society. So
that's a world that clearly feels that that religion is very important, so
important they're willing to, you know, not interact with a set of other
So that's a social mechanism that's available, that's not necessarily a
government mechanism. But it's a powerful mechanism and something that people
have often thought should be used sparingly. Regardless of where it comes
from, if you allow that mechanism to go wild, then, you know, there are in
some sense no limits to what sorts of behaviors it can enforce. So, of course,
people have enforced sexual preferences and racism, and all sorts of things
through this social shaming of people who deviate from some approved behavior.
So, the argument I would give is that, you know, we have mostly accepted a
norm that in sort of a public sphere of conversation, we are relatively open
to a wide range of arguments and claims, and we reserve sort of shock and sort
of refusing to deal with something for a pretty restricted range of things.
That's what we were in the past, but recently, people have just greatly
expanded the range of things for which they're willing and able to use this
method, weaponized it, you might say.
And now, people are basically fired and canceled for a wide range of pretty
minor offenses, which is basically putting a lot of people afraid of saying
much. And that's the sort of thing people have complained about that they're
now in a world where they do fear saying even pretty mild things, because
there's a machinery set up that will pounce on even very mild things and have
OK. So I mean, this is part of why this whole conversation confuses me,
because like, we started with, like, free speech, and as you said, it's
timely, and it's in the news and whatever. And then you have this little
speech about it, like what it is and whatever. And like, and then I'm like,
OK, but let's talk about the actual issue. And you shifted over to like a,
basically a totally different topic, namely tolerance and the enforcement of
social norms. And like, you know, the...
That is the current topic, isn't it?
But what I'm saying is, what does have to free speech? It has nothing to do
with free speech, as you define free speech, as you gave an analysis of free
speech, which is, well, there's a marketplace of ideas, right, and people– an
idea that there will be costs associated with participating is at least very
friendly to that of a marketplace, right? That's why I was asking you about
that. And then, and now you're like, "Well, you know, we shouldn't have– we
should have a moratorium on social shaming when it comes to free speech."
So, I mean, for example, you know, if we had a restaurant in the area, and it
was run by people with the wrong religion, then people might be happy to go
there, and use that restaurant and even work there until somebody tries to
create a coordinated social shaming ban where people are, basically have
agreed to watch out for each other and to go, you know, to tell everybody,
you're not allowed to go to that restaurant, anybody who goes to that
restaurant will be, you know, suffer the biggest consequences.
You can imagine a world in which we, you know, make everybody not go to the
restaurant, and they go out of business. And they, you know, people learn not
to do whatever it is it takes to create those mobs and get them mad. That's– I
mean, it's not government, but it's also not just people going to whatever
restaurants they like. That's another mechanism that's been added. And we can
ask whether that's a good or bad mechanism and to what extent we want to
encourage or discourage?
Sure, I mean, but like so– so it's sort of like, there are always going to be
social norms. And those social norms are going to be enforced through social
mechanisms, right? That's just a fact of every society ever, as far as we
Now, then, we can also step back, we, who can separate ourselves from these
norms, somehow, magically, I guess, and say, "Well, which ones of these do we
And so now you're saying you don't like so much the ones that are being
enforced around certain kinds of public speech. But fine, like and we can have
that conversation and we do have it about, you know, all kinds of different
activities like religion and sexuality, whenever, none of that has much to do
with speech, right? It has to do with tolerance really, as I see it.
I guess why– this is actually really gets it when I was saying I was torn
between two directions. Like, I now see that my being torn was like, are we
going to stay on topic or not? And almost all conversations of free speech
almost immediately move off topic, right? And that's what I feel we did we
start talking about tolerance and religion and whatever, because basically,
tolerance is just our capacity for ignoring stuff. Right? It's just sort of
like what can we ignore? And we learned to ignore religion, mostly.
I was rereading On Liberty to prepare for this conversation, John Stuart
Mill's On Liberty because in a lot of ways, it's the foundational text for
free speech, right, marketplace of ideas, which actually doesn't show up there
famously. But people always cite Mill on marketplace of ideas even though
nobody's mentioned it.
So and, you know, one thing Mill says about religious tolerance is that he's
says, "You didn't really get religious tolerance until you got people who are
skeptical of religion and kind of indifferent to it." And like, that's what
the tolerance really looks like. It looks like indifference, like when people
stop thinking that religion is really important, that's when you get religious
tolerance, right? When people stop moralizing sexuality, that's when you get
sexual tolerance. And so, suppose from him, suppose if that's right, that
tolerance is really a measure of all the stuff we can learn to stop caring
about. Right? Then to me...
I just don't think that's true. So, I feel like by this point in my life, I've
lived in many different intellectual communities that cared about a lot of
different things. But it was extremely rare for anybody to try to create a
cancellation campaign where somebody– where everybody was told they needed to
stay away from someone because of something they said. That almost never
happened in my many decades of many different intellectual communities. And
they were all communities where people cared a lot about the things they're
So this idea of, you know, coordinating to sort of cancel someone on the basis
of something they said, is previously a pretty rare thing. It's become much
more common lately. But even lately, it's a rare thing in the sense that most
of the time when people are talking about most things that they care about,
it's not raised as an issue. So, we're talking about pretty rare thing.
So, I think that social media has created new mechanisms for coordination and
punishment. But I think people have always been punished for saying things
other people don't like. I've watched it happen in faculty meetings. There's a
constant punishment going on for somebody saying something other people don't
like. It's reflexive, it's immediate, and it happens in every social context.
That's what people do. If I care about something, and I'm talking about that
thing, and you seem to me to be in some way threatening it, I will start
The difference between merely not liking something and reacting to it
negatively, and norm enforcement is that norm enforcement is about
coordination across a wide scope of people. So that's the thing that's much
Great. And in a faculty meeting, I might get two or three people on my side
against you, but it's probably consequences aren't going to be bad. I agree
with you that social media has allowed for very large-scale coordination, but
that just means that the size of the punishment has grown. Right? And...
And that's the thing to be talking about. Yes, that's the topic.
I'm just going to read this line from the from John Stuart Mill, I looked it
up about tolerance.
"So natural to mankind is intolerance in whatever they really care about that
religious freedom has hardly anywhere been practically realized, except where
religious indifference which dislikes to have its peace disturbed by
theological quarrels, has added its weight to the scale." I just want to read
it because I thought it's a great line.
So, I mean, when we were– earlier, when we're talking about speech, as I see
it, actually talking about speech. The thing that we were– you were talking
about hearing and how people sort of neglect the hearing side. And I think
there's something right about that, which is that there's this weird way that
speech involves a coordination. Right? It involves a coordination between what
I want to say and what you want to hear.
And for me, the deepest question about speech is like, how do we affect that
coordination? How do I how do I magically say the things that you want to
hear? Right? And I can't, I can't check with you in advance. And be like, "If
I say this, is it going to be OK, do you agree?" Right? Because then I've said
it. And so, speech is our mechanism for securing consent? And that creates a
problem for securing consent to speech because we can't use speech to secure
consent to speech?
And so, one way to be very free is just to say, it's fine. Anyone can say
anything to me or something like that, like in effect, you take consent off
the table. Right? But if you think, no, look, I mean, of course, I don't want
you to say anything at all. Like, you know, there's plenty– like don't tell me
the spoilers of the movie I'm about to see and don't like, give me an
eight-hour lecture. And...
So this, this comes down to what the essence of consent is. So, early in the
pandemic, I was interested in the idea of challenge trials where people might
try out the virus voluntarily and see how it went under controlled conditions.
And looking up a literature on consent, I found that many people argued, even
the majority argued, that you couldn't consent to an experiment about a
disease that people didn't know what would happen.
So the concept is there sort of a full information consent, that is if you
ever have any sort of uncertainty, you couldn't be consenting. And that– by
that standard, of course, you can't consent to get married. You can't consent
to a job, you can't consent to a move to a new city, because none of us can
assure you about the details of how that'll play out. So I think if consent
has to be meaningful, we have to be able to talk about it with respect to a
situation of uncertainty where you might regret giving the consent, expose
that as you discover what actually happens. But we can still say you consented
to it because you had a position of uncertainty. And given that uncertainty
you made the choice you made.
So it seems to me, we do that all the time for talking and listening. We don't
know exactly what someone will say. We might well say something we didn't want
to hear. And we might regret having listened to them. But we can certainly
consent to listening even if we are uncertain.
I think the worry isn't that we're going to regret. The worry is that we won't
regret it because we'll be persuaded by it. Right? So, like suppose...
We can regret now that we would later become changed. Sure.
But if we would regret it in advance, we wouldn't consent, right? And it's the
whole idea of regret. So suppose I have, I say, "Look, I'm going to say
something, if I say this thing to you, it's going to like really change your
mind about something. So the thing that you will believe after I say this, you
would now say that's completely wrong."
And, like, on the one hand, it seems to me you should say, "Don't tell me
that." Because that's something you don't want to believe, you know you don't
want to believe. If you wanted to believe it, you'd go ahead and believe it,
right? But you don't want to believe because you think it's wrong. And now,
I'm going to bring it about that you believe this thing. Why would you ever
This is a case where I think standard decision theory and statistics count for
the usual intuitions just fine. That is, you might think you're pretty
confident of something right now, but you're not sure. And if you get more
information, you will update your beliefs, and it might turn out, contrary to
your expectations that you change your mind a lot on the topic to the thing
that you initially thought was unlikely. But you can, in the standard decision
theory terms, want to do that, that can be what you want. You want to become
better informed, including learning surprising things that you would have not
expected to believe.
But its also true that maybe setting aside, sort of rational agent decision
theory framework, I think it's also completely reasonable to expect that you
aren't completely rational. And in fact, people can just persuade you, not
just inform you, in ways that are not accounted for by say, our decision
theory rational agent framework. And you could regret that like it's any, you
might be persuaded to just become somebody different and maybe you don't want
that to happen. And that's the chance you're taking when you listen to
someone. It's also just the chance you're taking when you interact with people
in a wide range of other ways. So it's not unique to speech. You know,
marrying someone, for example, they change you.
But you've been married.
Well, but still it's not necessarily the speech that will change you. There's
a lot of other things that will happen that will change you.
I think it's– marriage does changes you. But... so let's...
But I do think you can consent to things that, you know, are risky in the
sense that there's a chance that, from your point of view now, it will go
badly. And there's a chance to go well, and you make your bet, and you can
So, do you think that that is what we're doing when we're having
conversations? Is it you're kind of rolling the dice? That is that each
conversation is in a way a gamble like – that it's like the challenge trial,
I think maybe we don't, but we certainly should realize there's lots of things
that can go very wrong as a result of conversations. They aren't harmless.
They aren't riskless. And all interactions with other people have a similar
character that we have large things that can go wrong. We could fall in love,
for example. And if that happened, and fails, there's going to be great pain
to suffer. And maybe we don't want that to happen. But we take that chance
anytime we interact with people.
But it... so, I mean, one thing you might... I do see how– if you see another
person just as a source of like, an independent source of information, right?
So like, then you would say OK, of course I want more information, because
then I can base my beliefs on like, better information. Right?
And so that it's almost like, "Well, would you like look around the room for
something?" Because that's also gathering information. Right?
But, you know, we talked about this earlier that it doesn't seem actually like
people operate that way in conversations as like, "Here's my information. Now
you give me your information, then we can say goodbye, and we will have
converged to a similar place." Right? So like, it looks like what people often
tend to do is to try to say like, bring one another over to their side of some
So I think what I was saying was it looks like people don't actually view
conversations as information exchanges.
Well, they don't view them only as information exchanges. Certainly, they are,
in part information exchanges. They are also other things.
Right. But so let me ask like, can you consent to the other part? That is
suppose you could believe that that conversation will be 10% information
exchange, OK? So...
But then 90%, this other thing, which is like the person attempting to
persuade you of something that you don't currently believe.
I think, yes, you can consent to that. But you consent to all the other many
things that happen in human interactions, many of which are risky, and many of
which are unknown. But I certainly think the argument, the usual argument for
free speech is vulnerable to this counter argument that says, well,
information exchange isn't the only thing going on. If you allow people to
freely speak and freely listen, then other more illicit processes may result.
And you might think, shouldn't we be regulating and limiting those other more
illicit processes, even as we allow information to be exchanged. And in
principle, that makes sense, I just think the sort of usual point of view is
it's just really hard to do that. And so, maybe it's best off letting all the
other things go wrong as long as the information comes along too.
I don't think any of that, that is I don't think those processes are illicit.
And I don't think the problem will be solved by regulation. So, I just think
that the information is a relatively small part of the story. And it's a
relatively small part of the benefits too. I think that... Yeah.
So, I mean, having heard really early some description of your writings on the
subject recently, I will read you as saying that there are many sort of less
licit processes of persuasion. And that you don't want to give the label "free
speech" to those illicit processes. You would rather reserve it for your most
reliable, idealized version of the conversation, which for you is a Socratic
And that other kinds of bullying that is allowed by free speech you disapprove
of. And so, you would like us to focus on the most idealized, you know, best
version of conversation, and think about how we can move toward it.
I mean, I think just first, first of all, what I wanted to say was that, you
know, as I see it, it's not obvious how we consent to... it's not even obvious
that one consents to the acquisition of information, that that's like even the
right model. Like, insofar as you see, like, when I look around the room, I
gather information, but I don't think consent makes sense there. It's just
something, I'm just gathering information.
And so far as you view the conversation is also involved in that information
gathering process, can the concept of consent is sort of out of place. It
starts to be in place, once we are doing– talking about more than gathering
information, which I think most of conversation is more than that. And not
just the bad stuff, the good stuff, too. This is something other than that.
But I think that the problem is like if we think about consent, even with the
medical case, right? So like the consenting to those COVID trials, challenge
trials. The idea would be that the person has some goals, some preferences,
right, that can be satisfied by being in this trial. They may want to
contribute to science, they may just think it'd be fun, whatever, right?
And so, and they realize that some of their preferences may be thwarted by–
like to stay healthy or whatever. But they have to do the math there, right?
And your thought is, let's trust them to do the math, and take a risk of
having these preferences thwarted, in exchange for the promise of having these
preferences satisfied. So that's like the general model of like, you know,
allowing people to freely make choices.
But I don't think that applies so well to conversation. Because I think that
it's like, if you asked me, what are my preferences with respect to my
beliefs? I would say I have just one preference for them to be true. That's my
preference, right? Like, I basically have no other preferences. I mean, I want
them to be true and I want them to stay true. I want to not lose any of the
true ones either, right? So I want them to be sticky when they're true. I want
to be very, very dogmatic about my true beliefs and be unshakable on them. So
to really hold on to the truth, right, that's what I want.
And, I think I've already got it at least like, you know, to think, I mean, I
wouldn't like I wouldn't think anything I do think unless I thought was true,
right? And so it's hard for me to see how I could better satisfy my
preferences by potentially being moved off of the things that currently seemed
to be to be true, through your say, persuasive efforts.
So, the basic concept of free speech isn't expressed in terms of the concept
of consent. Consent isn’t central to that concept. You know, consent is a
concept, we do apply in many circumstances, but it's somewhat of an
idealization. That is, you know, the main fact is we are just interacting a
lot, and we affect each other a lot. And then you might criticize some
situation as somebody, you know, doing something illicitly. And then one way
you might describe that is there was a lack of consent.
And then, you know, for that criticism to go through, we have to ask, "Well,
what counts is consent?" And because it's widely accepted that we shouldn't do
things without people's consent, people have tried to sort of layer on a lot
of requirements to consent, so that it becomes very, very hard bar to meet, in
which case, they can complain about a lot.
But another attitude is to have a very weak concept of consent, which is
relatively easy to meet, and then a lot of things that happen can pass that
bar. So I think the usual legal concept of consent historically has been a
relatively weak version of you were explicitly offered a choice, and you made
it. But again, more recently, people have added on more requirements to
consent in terms of how much you need to understand about the situation and
how many things they might have told you and how many other options et cetera.
And of course, the more you layer onto this concept, the more requirements you
add, the harder it is to meet these requirements. And then the fewer things we
ever do that ever have consent, but in which case, the questions, but what do
we do? Because we have to go about the world doing things. And it seems like
this criticism that says we hardly ever consent is basically an excuse for
other people to come in and tell us what to do, because they're protecting us
from the lack of consent.
Right. And I get that that's like a scary prospect there. But that's not where
I'm headed at all. So that is, I think there's a real problem about consenting
to speech. But I don't think that the solution to that problem is regulation.
Because I think the regulation would also be bullying, so... But I do think
that there's a problem. And I think that free speech and consensual speech is
basically synonyms. I do think we're talking about consent.
I don't think you're right that we don't usually frame it that way. The reason
we don't usually frame it that way, is exactly because of the bias that you're
correcting with the idea of free hearing. Namely, we're just imagining people
spewing stuff out there, right? And then we're like, what restriction should
we have on what they can spew out there? Rather than understanding that speech
is fundamentally communicative, so that if there's someone doing the speaking
are not really speaking, unless someone else is doing the hearing. Right?
And now, we have to ask the question, what is it for that process to happen in
a way that's free, i.e. in some sense, voluntary or willing on the part of the
participants rather than forced, right? If somebody is compelled to speak
If you pick up my book, and you heard good things about it, and you read my
book, you didn't know what would happen, the results of it. You might have had
things happen to you, you didn't want to have happen, and maybe some things
happen that you regret, but it still seems like the usual concept was you made
your bet and you chose to read the book voluntarily. Isn't that consent?
So that the weak concept of consent would say that's enough. So it seems like
you must be arguing for stronger concept of the consent, more requirements are
added other than that you chose it. What are the requirements, I guess, what
more requirements do we add other than you chose to read the book?
Yeah, I mean, so I think that this problem absolutely arises for people
reading people's books. And there's a real consent issue there too. So I
don't– I think you might say like, '"Well, we have to get by, and we need
communication to happen." And so we're just going to sort of adopt this kludge
of like, people mostly get to say whatever they want, but there's a few rules
in the background. And then occasionally, people get really angry, and then
they like, cancel someone or whatever, they get punished for what they say.
But mostly, we're just going to kind of work with the conceit that we all sort
of consent to being spoken to, and all sorts of different ways.
And that might be like the sort of the best we can do. If there's no other way
to make good on the idea of consensual communication. But I guess my thought
is, there is actually a way, that is there is a kind of speech that doesn't
raise this problem, so that you couldn't be bullying the person. And there
couldn't be a problem of why would they buy into this in advance.
And so, that's – to set the stage for that that I want to say there is a there
is a general problem of consent consenting to speech. And it's in fact, that
is the problem that creates all these free speech, worries and issues, right,
and calls for everyone agrees that there should be some regulation, right? And
the question is just how much? And so there's a kind of escape patch. There's
a kind of speech that doesn't raise this problem. And that is a refutation,
which is explaining to someone why they haven't actually answered the question
that they said that they've answered. I think that that is totally...
Now, refutation is a pairwise interaction.
Yes. All speech is.
Well, when you write an article for the newspaper and thousands of people read
it, it's not pairwise there.
It's many pairwise interactions. I interact with each of them.
I mean, it's not plausible to say that we should only ever speak in a
refutation format, right? We have... Right. So well then, what are you saying?
I was saying that, like...
If we only talked about the weather, we probably also wouldn't have much
problem either. So suddenly, also, the problem could be laid at the kind of
topics we talk about. There are many things that contribute to some people
being upset about what we say.
I think people could get upset about the weather. No question. People can get
upset about anything, including refutation, right? In a way, the whole problem
is that the standard sort of defense of free speech is going to be a function
of what makes people upset. And that kind of speech is never going to be
totally free. Right? So like, but refutation whether it counts as free has
nothing to do with whether the other person gets upset or not. That's just
irrelevant. How they feel about it, whether they opted in, none of that is
But if you point out there's a particular kind of conversation style, that's
especially good. I don't see what that does to speak to the larger question of
our norms and laws regarding all the other kinds of things we could be saying
and how, right? We have to judge for all the other things that people say
whether or not to allow norms or laws to restrain them and how far.
Yeah. And OK, that's a fair criticism of what I'm saying. I don't think it
immediately gives you answers to those questions. But I do think that if you
were to grant to me, which is a big, big grant, OK, that none of what we
typically call free speech is really free, because it's not consented to, but
the basic model is still consent. If there's a consent that can never be
secured, and we have something instead of that, which is like a kind of
regulated, minimally regulated space of ideas or something like that. But then
there's this other speech that is actually free.
Then if we want to think about, OK, how should we regulate the space? Right?
How much should we crack down on, you know, social consequences of speech? If
we want to know the answer to those questions. It seems to me at least useful
to be like, OK, well, our paradigm is going to be this thing over here.
Yeah. A paradigm seems too strong a claim for that. That is a paradigm would
be sort of the thing you were trying to model everything else on. And that's
too strong a claim here. You could say there is nice features of this one
conversational style but that's not enough to make it the paradigm. Paradigm
is a strong claim to make about something.
I'll go further and I'll say that, like, if you're saying that the mark of
problems is that people get upset, and you admit that people get upset when
you refute them, then by that definition, refutation doesn't avoid the problem
of people getting upset. It seems like what you want to say is that people
shouldn't get upset about refutation.
No. I don't think that. I think they should get upset. It's fine. It's fine
for them to be upset. No, I think that in the case of refutation, you are
achieving someone's ends by refuting them. You're doing something that's good
for them, regardless of whether they think, whether they like it, or whether
they agree to it, whether they think it's good. It's in fact infinite.
But that's not the topic here. I mean the topic here is people getting mad and
upset when you say things. The topic isn't whether you're achieving your ends.
No, the topic is whether that madness and upsetness is relevant or
significant. And I think it's significant in the case where you basically are
still using a consent model, which is most speech and it's not– it just
doesn't matter from the case of refutation, it doesn't matter that the person
gets upset at you.
You're deciding it shouldn't matter but they could think it matters.
They could think it matters. That's fine.
Well, but from the point of view of free speech norms, and it's about the
norms we all want to adopt, which is about what we will all sort of agree to,
or enough of us will agree to, when we anticipate, you know, the consequences
of our choice. And so, if we anticipate getting mad, then we might be
reluctant. Your philosophical argument that we shouldn't get mad will persuade
us not to get mad?
I don't think we shouldn't get mad. You keep saying I think that but I don't.
I think you should.
Norms and laws are something that we come to agree to. You're somehow saying
we should agree about something about refutation. I'm not even sure what
exactly you're proposing. But if the question is, what norms of the law should
we agree to? And you say reputation is great. I'm going, "Yeah, but why is
that relevant to the question we're asking?"
So, I think that– so if you look at like, you know, Mill's On Liberty, and you
look at the argument that he makes for freedom of thought and expression. It's
not an empirical argument for the most part. It's not– he doesn't study like
places where there has been more free speech and less and be like, "OK, you
get more progress over here." He does think it's going to lead to more
progress, that's the argument.
But his basic idea is that he thinks that the idea is that people are going to
want to suppress are false ideas, what they think about as false. In
particular, he's thinking about heresies, religious heresies. And he thinks
that if an idea is false, then you know, well, first of all, it might not be
false so you shouldn't suppress it, because maybe it'll turn out to be true.
But then even if it just is false, it's going– you're going to understand the
truth better by knowing of this thing as false, it's going to like, kind of
shine a reverse light on the truth. Right?
And so, his basic idea is that all ideas are somehow in some sense, a positive
sum, a positive addition, that is even the bad ideas are good. That's his sort
of fundamental thesis or argument for free speech. And I think you need some
argument, right? You need some reason for thinking that there should be free
speech. That's his argument.
And so, I think that it's no longer the case that people are only wanting to
suppress ideas they see as false. I think people often want to suppress ideas
they see as true. And so actually, Mill's argument doesn't address the current
predicament. He didn't anticipate that people would want to suppress ideas
they saw as true, as far– I mean, I reread it kind of quickly, so maybe I
missed some part.
And so you might say, well, what... what is going to be the argument? And so
it seems to me that the idea of an inquiry into the truth that happens by way
of showing someone that they didn't– that they haven't thought something
through in the way that they thought they did, that that's a potential... it's
a potential core around which the value of open expression could be built. So
that it's like, it's for the sake of this form of thinking. And it had
I think we have to get into– we have to get into what it is that people want
to have happen when they want a true idea to be suppressed? Because there are
a number of different interpretations there. One could be that they privately
know it's true, but they don't want other people to hear that it's true. And
they think there'll be bad social consequences for other people hearing and
believing that it's true. Or they could think that, even though we all know
it's true, it would be bad if we said so. But it seems refutation is in the
service of coming to believe what's true, if it's not good to believe what's
true, then refutation is also vulnerable then, not just a marketplace of ideas
that might produce true beliefs, but refutation is doing the same thing.
Right. So, I think that I'm not trying to come down on sort of like, how free
should the market of ideas be, right? That is, I think, in a way, there's just
a debate over regulation there. But I think the question is like, what should
the guiding idea be guiding that debate of like, sort of both how much
regulation from the point of view of the government? But that's not really the
The real interesting question is, to what degree can we cut down on social
sanctions for speech? Right? Which is almost like asking for more regulation,
really the free speech rule, but more regulation. Now, it's ironic. And I
think – I guess my thought is that we can think about that question but I
think you're like, what's the ideal speech scenario?
And now, you want to say, OK, but people will sort of speak censor the ideal
speech scenario, or cut you down in the ideal speech scenario. And so I think
that it's not relevant. Like, if you're refuting someone, part of what you're
doing is you have to keep them in the conversation, which means you have to
keep them not censoring you. And you have to keep them not shutting you up.
And you have to use whatever interpersonal resources that are at your disposal
to do that.
But if we take an example of someone who has been canceled lately, it sounds
like you're making the claim that had they properly followed refutation
processes, they would not have been canceled. That is there was some process
they could have followed to make their argument and prevent cancellation if
they had been refuting, because, by your account, refutation, won't allow the
cancellation process to happen, because they will have been including people.
But of course, you know, almost all these people are speaking to wide
audiences, they can't ensure that everybody in the audience is in the frame of
mind, they are trying to get them into. It seems like a very high bar to
demand that everybody who speaks in public successfully get everybody to be in
the proper refutation frame of mind so that no one will ever object. That just
They might well object later. I mean, Socrates was epically canceled, right?
It's not like, no, I'm not saying refutation prevents you from like, losing
your job or anything like that. Right?
You know, what Socrates said was like, this isn't really bad. Being put to
death, that's not. Living false things, that's really bad. I looked to it and
have a bad thing happened to me. So, Socrates didn't feel that his free speech
was much constrained. I mean, he got to speak for most of his life.
Eventually, he got killed for it, but he got to have a lot of conversations.
like he would say. But... so I think it's certainly... I mean, I think the
question is speaking to a large audience. If you're speaking to a large
audience, then you're taking a risk, right? And you might think, yeah, you're
thinking about the, you know, you've consequences, right? You've rolled the
dice. And the larger the audience, the, you know, the more of a risk you're
taking, because there could be like, a small group in there that then can
coordinate over like, how much they hate what you said.
And so, what we should maybe still give people the free choice to like, take
those risks and go on Twitter or whatever, right? Because, you know, people
can make their own choices or whatever. I mean, someone can make that argument
too there. But the kind of thing I'm talking about that's going to be our
paradigm for free speech is going to be a conversation basically between two
people. And so yeah, I think issues of cancellation just aren't going to arise
there. Except when you compound it, like Socrates did, right? He just– there's
a lot of times, a lot of enough times that eventually, the effect was big.
So I'm still not seeing the relationship between the main issues of free
speech and a choice of a paradigm conversation structure.
Yeah, I think...
I don't see the need for a paradigm conversation structure. We accept there's
lots of conversation structures, and we want an approach to free speech that's
robust to a wide range of conversation structures. We certainly don't, you
know, it'd be awkward to have different rules for different conversation
structures. We're trying to think about social laws and norms.
I think it's a fair objection. I don't think I have clearly worked up that bit
in my head. But I guess I do – so I think that, you know, I think in a way,
what you're saying is like, you're sort of going off topic. And I think that
that's right. But I also think that the regular free speech conversation goes
off topic, when we start talking about tolerance, and it's very hard to
actually talk about free speech.
I want to push back on that. So I want to say, law and norms aren't as
different from each other as people often think, or as you're talking. That
is, laws are formalized norms. Laws are improvement in some ways over common
norm enforcement. But before law, we had norms, and that's all we had. And
once we've added laws, then we can have more reliable enforcement of what once
And now, if we have some norms that a lot of people are trying to enforce, we
should ask the same sort of questions we would ask if they were laws, because
it's the same considerations. And we might even ask, as I've suggested that if
these are important enough to enforce, maybe we should enforce them as laws
and not as norms. Because norm enforcement is inefficient, in many ways
relative to legal enforcement.
So, and again, you know, we're talking about something cases that are
relatively rare in the sense that a tiny percent of conversations, although it
has way out of proportional consequences for people, where somebody says
something, and then some crowd gets riled up about it, and passes it on and
creates a big discussion about it and tries to get people fired, or their
speech canceled, or, you know, all sorts of other things happen to them.
And that's a lot like a legal process, which is, you know, most of the time,
we aren't invoking laws on most people, most of the time. Laws are there in
the background to constrain and encourage behavior. But we have a rare law
enforcement process in the sense that most of the time, most people aren't
accused of a crime.
And similarly, with the sort of norms, it's rare to have someone people try to
cancel somebody. And that's to be in contrast with the many kinds of norms
that are constantly appearing in our ordinary interactions with each other,
which we are constantly enforcing in relatively weak ways, right? In terms of
just who looks at who in a conversation, or who seems to act tired or, you
And so, I would want to separate out ordinary norm enforcement from these
sorts of cancellation scenarios, which are really quite different in a
character. I mean, they are both enforcing of norms, but this sort of thing is
much more like law than like most norm enforcement.
It seems to me, the difference is just that the level of coordination allows
for bigger punishment.
But this level of coordination is still rarely used. That's the key point. And
just like law is rarely used, that is, we have a limited set of laws, and we
rarely accuse people of violating the laws. And when we do, we have a big,
expensive, complicated process we go through that is a matter of coordination
to enforce laws. And that's at the opposite end of the spectrum from all the
little ways that we enforce sort of very mild norms in ordinary human
See, it seems to me that there's sort of two things one could do with the
thought that you just had. One of them is start whatever process is required
for a kind of judicial transition of this norm. Right? And I don't know what
that is, you know, maybe you'd have to talk to the right people, you have to
lobby certain people or whatever, maybe you'd have build up some political
That's one road you could take. Another road you could take is like, kind of
use this idea to refute other ideas and like battle it out in the space of
ideas, which is mostly a matter of refutation. Right? And that's the thing you
do, right? Which, it still seems to me like fundamentally, in some sense,
inside of your own activity, the fundamental model of what you're doing is
refutation. And that's what it is for the speech and talk to be free is for it
to be in a kind of combat, right, where there's a theory or an approach that's
given. And then you sort of undermine it on its own terms. Like if this is
what you wanted, here would be a better way of doing that, or something like
So, I think you know that in the ancient world, there were a lot of theorists
who offered grand theories of the world. And they were often of the forms like
one core thing is everything, supported thing, right? Everything is fire,
everything is circles. Everything is love. And, you know, the world gets tired
of those sorts of things. And today, we are more interested in like, theories
that add or change an item to a menu of items, or add or change a process to a
menu of processes.
And so, that's how I'm reacting to your refutation is the one core model I'm
going, you know, there's lots of different kinds of interaction, lots of
different kinds of conversation, lots of different conversational purposes.
I'm not against refutation. I'm just not willing to sort of take it as the one
model that we should use for all issues in conversation. I think we should
accept that there's a wide range of purposes and styles of conversation and
look for accounts and regulatory accounts and approaches that are robust to
the wide range.
That would be, the only reason you're saying that making that point to me is
that you think I do want to be the one central one and you're trying to refute
me. So you like it that much. You like it enough to employ it pretty much
exclusively in every conversation you have or that I've had with you. That's a
pretty big endorsement.
OK, I am not against refutation. Let me go on the record saying I'm not
It's like you're not against not against breathing.
Exactly. I just don't think we should focus entirely on breathing when we're
focused on many other things that are interesting.
As long as we keep doing it constantly.
Yes. That's OK, too. Let us continue to breathe in the background and refute
in the background. All right, we've probably done enough for tonight.
All right. Until we meet again.