Bullshit
Robin:
Hello, Agnes.
Agnes:
Hi, Robin.
Robin:
We both just attended an event in Florida.
Agnes:
The event was a conference, just, you know, not to leave the art hanging.
Robin:
Right. And on the demand for bullshit was the topic.
Agnes:
Right.
Robin:
Which wasn't formally defined for us. And there was a pretty wide range of
talks. So I guess, in effect, there was a wide range of effective definitions
of the word, given that nobody really complained about anybody else's usage or
the appropriateness of their talk for the event.
Agnes:
I think the conference wasn't really exactly about bullshit. It was like, why
is public discourse so messed up? It was a little more about that. And so the
question of defining bullshit was a little bit in the But it was not in the
background of my head as I was listening to the talks. I was thinking to
myself, huh, what is bullshit? And over the course of listening to some of the
talks and sort of disagreeing with some of them, I came up with the definition
of bullshit. So the first thing that I, first realization that I had is that
when we say, when we describe something that someone else says as bullshit,
We're generally insulting it and we rarely apply that to ourselves. And that
use of bullshit is sort of parasitic on another more central case where we
would apply it to ourselves, namely the practice of bullshitting. That is,
sometimes people get together to shoot the shit or just like bullshit together
And I think we can sort of understand what that means, what that practice is.
And then the kind of bullshit at stake in the conference, which is always
something someone else is doing, is never something you're doing in the kind
of public discourse sense. It's always the public discourse that other people
are creating that's bullshit. In that sense, it's parasitic on the
bullshitting activity where you would self-ascribe that you're bullshitting.
So what is the bullshitting? The practice. I think it is a tacit agreement to
relaxed epistemic standards in the context of the conversation. So it's like
we implicitly contract that we're not going to challenge each other very hard.
We're not going to ask for evidence. We're not going to bring up conflicting
evidence. We're not going to ask for explanations. We're not going to ask for
a lot of details because we're letting each other bullshit. Okay, so that's my
understanding of what bullshitting is. It's a form of conversation with
relaxed epistemic standards, where those are mutually agreed to in an implicit
way by the participants. Now, I think when we call things bullshit, like in
public discourse, we're basically doing that by saying, what we mean is that
the other person the person who we say is bullshitting is doing that thing,
that thing that's part of the practice, except in a context where it's not
welcome, where we haven't agreed to those relaxed epistemic standards.
Robin:
Right. So the definition I had taken from the previous literature and used in
the introduction of my talk doesn't seem greatly at odds with yours, but you
could decide how much it is. I focused on the idea that in communication, we
have many channels of communication, body language, facial expressions, tone
of voice, pauses, metaphors, et cetera. And that the literal meaning of words
is typically only a small fraction of that overall bandwidth of communication.
And that sometimes we are in a state where the literal meaning of the words is
not expected or treated as a large fraction of the communication. And
therefore, if it's not exactly literally right, it doesn't bother us that
much, perhaps, if we are accepting it. That is, there are just modes of
communication where the literal words are not front-centered and maybe
somewhat at odds with the other channels and that can be okay or not okay. And
then the interesting dispute is when people want to accuse other people of
illicitly being in that mode or being in a mode way that they don't approve,
which then raises standards. Well, when, how much should the literal meanings
of the words typically, you know, how much of the communication should that
carry? How precisely should that be consistent with the things said through
the other channels? What is the appropriate place for precisely attending to
the literal meaning of words and imposing strong epistemic standards on those
literal words?
Agnes:
You just smooshed two things together there at the end that are just not the
same. So I think our definitions are not the same. That is, attacking the
literal meanings of words and imposing strong epistemic standards just totally
cut across each other. So I think your definition is wrong. That is, I don't
think bullshit has much to do with the literal meanings of words, except
insofar as occasionally there's some correlation between being not literal and
relaxing epistemic standards. But I think you can bullshit and insist on
literal meanings of words and be only talking literally. In fact, I think that
you can bullshit someone by insisting on the literal meanings of their words,
right? When it's like clear from the context that they don't mean it literally
and that they're making a good point, but you're pushing on the literal thing
to bullshit.
Robin:
I meant that most of our epistemic standards will be applying to the literal
meanings of words. That is, we don't typically use epistemic standards with
respect to body language or tone of voice.
Agnes:
Oh, not body language, but I think there's just a lot in context that's
relevant that's not given by the literal meanings of words. I mean, sorry,
maybe I don't know what you mean by literal meanings of words, but much of the
meaning of a given utterance comes from understanding its role in a context.
Robin:
We find it hard to have critiques of that. I mean, so most of our epistemic
critiques are going to be taking things relatively literally because we just
don't really have ways to not do that. That is, you know, if you tell me two
probability numbers that, you know, are compliments that don't add up to one,
I can critique that if you actually say those things literally. But if you
sort of imply them with your tone of voice or something, it's really hard to
critique.
Agnes:
Right, so okay, so let's separate two things, maybe that so one of them is.
Um, so something like, um, what you explicitly communicate, which is not
restricted to the literal meanings of your words. Um, um, because you can
explicitly communicate things. Um, no, I, I, I don't think the distinction is
like, even that's not going to work. Um, I'm not really sure how we
distinguish stuff about tone and body language from. uh what what you say um
and i mean and i'm not sure that distinction is relevant to bullshit but i but
i i don't i can't at the moment i'm i'm sort of struggling to come up with a
clear way to distinguish maybe you're looking for something like the semantics
pragmatics distinction No, probably not. Probably not that.
Robin:
I mean, I guess if we take your epistemic standards point, I mean, I think is
to say epistemic standards are only practically applicable to certain aspects
of communication.
Agnes:
I don't think that's true. I think that when people know each other well, they
could say, Wait a minute, you raised your voice at the end of that. Did you
mean to be questioning it? And if so, why? That is, I can apply epistemic
standards to your raising of your voice.
Robin:
When it's clear, perhaps. But a lot of communication just isn't clear enough
to be so precisely agreed on. And that's a feature of a lot of communication.
Agnes:
One thing I can do is ask for more clarity if there isn't more clarity. And
that's applying epistemic standards.
Robin:
Right, but I think part of the point is that what happens when people accuse
other people of bullshit and use that word is that they typically accuse
people of not meeting epistemic standards or perhaps not sort of being
consistent in the literal meanings of the words and then we often find it hard
to judge those accusations. That is, accusations of bullshit are often
rebuffed by defenses that, no, I wasn't bullshitting. And we often, I mean, a
key part of the phenomena is it's hard to really lay the accusation clearly
because there are many ways in which people can pretend not to be
bullshitting, even if they are, or to evade the accusation. So that's part of
what I was looking for in the, when is it feasible to actually verify that
somebody is or isn't bullshitting?
Agnes:
That seems to me an accidental property. The fact that when somebody is
bullshitting, we might find it hard to discern that they're bullshitting
shouldn't be part of our definition of bullshitting. It's just something that
happens to be true of bullshitting. We first have to agree on a definition,
the essential property of something, before we get to the accidental part of
it. So let's go back to just the definition part, and then we can talk about
applying this standard of definition. So you said, oh, they're accusing them
of not meeting epistemic standards. I don't think that's right. You're not
being precise enough with my definition. My definition says it's not that
they're failing to meet a set of epistemic standards. That's not enough to
count as bullshitting. It's that they are engaging in a practice. They're
making a bid for lowered epistemic standards and you're rejecting that bid.
Robin:
Someone could just fail to meet epistemic standards of yours and you're right
I agree but unjustified in order to accuse them of failing to me right but in
order to successfully accuse them of not trying to meet epistemic standards,
the straightforward way is to show that some particular thing they say doesn't
meet the standards that is, if they consistently met the standards. it will be
much harder to show that they weren't even trying to meet the standards. That
is typically the way that you, the sign of not trying to meet the standards is
that you fail to meet them. So that would be the typical way that you would
successfully accuse someone of, you know, you would have to add more to your
case, but part of the case for saying they're not even trying to meet
standards is they aren't in fact not meeting the standards.
Agnes:
In my experience, the label of bullshit is typically not applied in a way
where there is any desire to make the case to the person who is bullshitting
that they are actually bullshitting. And it's typically applied in a kind of
dismissive way. And people do not tend to offer a lot of proof. And so I don't
have the sense that this problem of how to determine whether they're really
bullshitting is one that sort of occupies people in a significant way.
Robin:
Well, for example, if we take political conflicts where people are divided up
politically, often they want to accuse the other side in their political talk
of bullshitting.
Agnes:
Right, but they don't do that by way of seeking the other side's recognition
that they really are bullshitting.
Robin:
I agree, but the question is, what evidence would you point to?
Agnes:
Why would you point to any evidence?
Robin:
Because sometimes it's not obviously accepted that people are bullshitting, so
they want to.
Agnes:
Fair, fair. So I think that the evidence that people point to is not
necessarily that they don't meet epistemic standards. Typically, the evidence
that people point to is simply if the statement is false. So if I'm just
imagining a tweet about, hey, something Elon said is bullshit, what's then the
tweet going to say? It's going to be like, because, and then it's going to
cite some fact that is a negation of something that he asserted. So I think
even though it's true that what we're accusing someone of in accusing them of
bullshitting is making a bid for lowered epistemic standards that we do not
want to accept, the proof of it lies just in saying that what they said is
false.
Robin:
I think in addition to pointing out something that's false, I think they try
to point out that it should have been obvious to them that it's false. That
is, it's such an obvious falsehood that they couldn't have been saying that
without lowered standards.
Agnes:
Good, good, good, good. I think that's right, right. Because you want to show
that, I mean, sorry, I think often, often they're not doing that. Even
typically, they're not doing that. Typically, they're just showing that it's
false. But there's a background presumption Elon should have known or could
have easily found out And then the point is, if we assume that background
presumption, then the only reason he would have said something which he could
have easily found out was false is that he was making a bid for lower
depressive standards.
Robin:
So if we back off and say, why would anybody bother to accuse people of
bullshit, et cetera, I think we have to come back to the phenomenon of
disagreement. We had a podcast on that long ago. Basically, people should be
and are a bit puzzled by the fact of disagreement. And, you know, most
obviously the fact of disagreement means they could be wrong but maybe I could
be wrong. And so the most reasonable response to generic disagreement would be
to doubt your own view as well as doubt them, but we often want to find a way
to blame them instead of ourselves and then bullshitting is a explanation for
why they might be wrong and why we therefore might not be wrong because we
weren't trying to lower our standards. So if we say they were trying to lower
their standards, then that's a way we can explain our reasoning as solid and
not in need of updating in the face of their disagreement. It's basically a
way of explaining disagreement by blaming it on their motives and their
careless sloppiness and stupidity. Those are our favorite explanations for
disagreement. And so that's why I would say the concept of bullshit is so
attractive. If people ever lower their intellectual standards or their
epistemic standards, well, that's a great thing to invoke to explain why I'm
right and they're wrong.
Agnes:
Right, which is why my definition is so good. But let me take a step back. I
think that's right. But I think that it's worth something we should have done
is just sort of contrast this definition that I'm offering with the standard
definition of bullshit, which is the one that Harry Frankfurt offers in his
book on bullshit. And that is a definition that was brought up in the
conference. And I'm sort of looking it up to get it exactly, but it's
something like speaking.
Robin:
Here's how I summarized it in my talk.
Agnes:
Yeah.
Robin:
Statements produced and heard with less concern for truth or clarity. Okay.
Sounds a lot like less of lower epistemic standards, really. It's not that
different. Less concern for truth or clarity is, I mean, the epistemic
standards about concern for truth and clarity. So that's what they mean.
Agnes:
Right. So. So, I mean, I think that the one, like my recollection of the way
that Frankfurt put it is that it's something like speech that doesn't have any
regard for the truth. Your version of what Frankfurt is already an
improvement. Because I just don't think that there is any speech that has no
regard for the truth. I think even if you're lying to people, you have regard
for the truth. You always have regard for the truth when we talk. I'm not even
sure. So I think less regard for the truth is better. But I think that you
could believe something, and you could believe that it's true, without, say,
having much of a justification for it's being true, or an explanation of why
it's true, or an ability to rebut counterexamples. But you could still, it's
very much a fact of life that people have beliefs.
Robin:
Our epistemic norms are that in that situation, you should be less confident.
That is, you should know that you don't have a justification, and therefore,
you should know that you don't have a very good basis. And you couldn't
reasonably believe very strongly in the situation you just described. So if
you do believe strongly in the situation, you have less regard for our
epistemic standards, i.e., and also for the truth.
Agnes:
I mean, I guess I think those, I'm not inclined to collapse those two things.
Just because I think that, at least for us personally, a person might although
they can't respond to counter-arguments. feel just very, very confident in the
truth of what they're saying, and they think they do have regard for the truth
because they think this thing is true, even though they have no evidence for
it. They strongly think it's true, so they care about the truth.
Robin:
I'm happy to accept your definition if we can then use that as a basis to
discuss the phenomena instead of discussing definitions, because it seems like
the interesting questions are, why does this phenomena exist? Why do people so
often use it as an accusation? using it hypocritically as an accusation, that
is, are the people accusing other people of bullshit just as guilty of it
themselves? Is there a problem with too much bullshit in general, or is it
just an opportunistic accusation that's made? These seem to me the more
interesting questions about bullshit.
Agnes:
So the reason why I think it's important to establish that it's not just
something like speaking without regard for the truth, but it is something like
making a bid that isn't getting accepted is that I actually think it helps
address your puzzle about disagreement, to think about it my way. That is,
your puzzle about disagreement is, that we feel like our disagreements with
other people sort of should have implications for what we think, but that they
don't have. That is, we're not drawing a set of conclusions from the
disagreements. And what calling someone a bullshitter allows you to do is to
say they're opting out of the rational discourse game. And so I'm being
perfectly justified in, you know, ignoring them. That is, I'm epistemically
kosher even though I'm ignoring a rational interlocutor because they're not a
rational interlocutor because they've opted out. And they've opted out by
asking me to opt in to a lowered epistemic game.
Robin:
So for this to work, it has to be true that people kind of know when they're
bullshitting. Because if you didn't know when you were bullshitting, you might
in fact also be bullshitting and you wouldn't, you know, it wouldn't be true
that you're more likely to be bullshitting than they are. That is, we need
some way to be telling that they're bullshitting and I'm not, or some degree
of version of that, so that I could be justified. It wouldn't actually matter
that I wasn't.
Agnes:
So suppose the point here is only to rule them out as someone with whom I have
to engage. If they're bullshitting, I don't have to engage with them. Even if
I'm a bullshitter, I still don't have to engage with them. I'm just as bad as
they are.
Robin:
So think of it as a continuum of you could be bullshitting more or less, in
the same way that you could be drunk more or less, or any other way that you
could be less capacitated. So if I disagree with you when you're drunk, I
might think I'm justified in holding my opinions and rejecting your
disagreement. But if I could be just as likely to be just as drunk as you but
not know it, well, now it's less clear I should be rejecting your possibly
drunk views against my possibly drunk views.
Agnes:
I guess the way that I was understanding the concept of bullshit is to accuse
someone of bullshit is to accuse them of having fallen below the epistemic
threshold where you're obligated to take their ideas seriously, which could
happen with drunkenness, right? So like someone could be drunk enough that you
shouldn't take what they say seriously. We could have such a threshold. I
mean, we have such a threshold with driving, right? And the point is, it
doesn't matter how drunk you are. If you think that they pass that, if you
think they pass that threshold, they're like, they're not someone you should
take seriously. You should find someone who isn't drunk and take them
seriously. It doesn't matter whether you're drunk or not.
Robin:
The careful analysis of the situations does not reveal a threshold above or
below which you should suddenly switch. It's a continuum. They are just less
reliable the more drunk they are, the more bullshitting they are. But you're
also less reliable in the same way. So it's about the relative incapacitation
or capacity in comparison.
Agnes:
Well, it might not be a label. when I apply the labels. So I can describe
someone as incapacitated by alcohol. And what I mean is that they are above a
certain percentage of blood alcohol level or whatever, where I determined that
to be incapacitation, right? And it could, even though there's an underlying
continuum, our label doesn't have to respect that continuum. Our label can
just pick a cutoff point and say, anyone who's above that point can't drive or
can't testify or can't do a bunch of other things. And it could well be that
when we say of someone that they're bullshitting, what we're saying is they've
asked for epistemic standards to be lowered to a point where whatever they say
can be dismissed.
Robin:
But you can't dismiss them just because they pass an arbitrary threshold.
That's incoherent. That is, if we're talking about epistemic standards or
evidential standards, then, for example, if we're both considering which of us
should drive, even if you are below some official standard for being, you
know, sober enough to drive, if I'm even lower than that standard, it still
might be better to have you drive than me drive. So, and similarly in
discussing the actual evidence and relative weight you should put on us depend
continuously on these parameters, not, and they don't reflect some arbitrary
cutoff.
Agnes:
But it might be that we are, I guess I'm just inclined to think that it is a
cutoff point. And the cutoff doesn't have to actually mark some specific same
point. So here's a point that Harvey Sachs makes about language, the way
language relates to measurement. We have lots of linguistic concepts that we
use, that we talk to each other, like old and young. or I didn't get a lot of
sleep last night, or smart or whatever, where there's actually quite a bit of
content in those words, but people are going to have different cutoff points
and we don't try to coordinate. In a group of 80-year-olds, a 60-year-old is
young, and everything that goes along with young is going to go along with
that person. But if he's around 40-year-olds, he's going to be old. And that's
part of how we use language. And so language pulls apart in a certain way from
the scientific reality. And with many concepts, what we do is we're like,
they're above the threshold. And we don't say what the threshold is, but we
say something like, look, they're above what a reasonable person should think
the threshold is. They're at the point where we should dismiss them. And the
point isn't about taking what they say as being a basis for updating. The
point is, do I have to converse with them or communicate with them or be in
dialogue with them? And like, should you get into a car with me if I'm
incapacitated? Should I drive you? No, that's the question. Not which of the
two of us should drive, but should I drive you? No, you should just not get in
a car with me. Forget about getting in a car with me because I'm too drunk to
drive. You're even drunker, right? You shouldn't drive either. Gotta find
somebody else to drive. If anyone's gonna do any driving, it's not gonna be
either of us. That would be the idea.
Robin:
It seems to me like you're focused again on defining the word and how people
use the word. And I want to talk about the phenomena. So the phenomena is we
disagree and we wonder how much to worry about our disagreements. And that
phenomena is continuous with the degree of bullshitting or the degree of
alcohol. And then the interesting thing to me is how to respond to that
phenomena of, many people often changing their epistemic standards and maybe
not even be fully aware of how they're lowering their standards. I think quite
often we switch, say, in politics or other moral or other personal things, and
we don't quite notice that we're changing our standards. And that's a problem.
Agnes:
I would describe the phenomenon differently for you. I think the phenomenon is
we don't know who we should talk to. and who we're obligated to listen to.
That is, who should we be in conversation with, is the question that we're
asking ourselves. About whom are we allowed to say, I'm not going to talk to
them. It's not about updating on the basis of what they think, it's about who
are we supposed to be in dialogue with. I think we can dismiss people from the
dialogue on the basis of incapacitation.
Robin:
a degree of incapacitation. But you think there's a sharp threshold in that
degree where you can just dismiss people?
Agnes:
Because I think the question, should I be in dialogue with them, is a yes-no
question. It's not a degree question. If you were updating, you could update
more or less. You could take this thing, right? But conversation isn't like
that.
Robin:
Your prototypical examples of accusing, say, the other political side of
bullshit is a continuous case of how much weight to put on their opinion.
You're not actually talking to them at all.
Agnes:
I don't think that is what you're saying. That's what I'm saying. I think what
you're doing, the conversational move you're making, is dismissing something.
You're not saying, put less weight on this. That's not what it is to call it
bullshit. To call it bullshit is to say, chuck it in the trash. Don't update
on it at all. Treat it as zero information. And I get that from the point of
view of a certain rational theory, that's not how people should act, but that
is how they act. That's the social space that they're in. And in my view,
that's the interesting phenomena, is the space that they're in, not a model
that we've come up with that doesn't fit the way people actually think and
act.
Robin:
Well, now you need to explain this phenomena of dismissal. You're claiming it
doesn't track evidence or informativeness. So what is the basis for dismissal
if not evidence or informativeness? You're going to claim bullshit is all
about the righteous dismissal of people, which is not directly related to
their evidence or informativeness. What is the basis for dismissing people
then?
Agnes:
It's that they've made a bid for lowered epistemic standards, and you're
rejecting that bid because it's too low.
Robin:
But that bid comes on a continuum. Why have that threshold for when you reject
them?
Agnes:
I guess I think that's fair, but we could say, why have the threshold for when
we call someone tall? We do call someone who have a certain threshold tall.
Robin:
Calling someone by name is different than dismissing them. Dismissing them is
an action. It's not just an arbitrary label.
Agnes:
But we can perform actions on the basis of judgments of someone's being tall I
mean, but then you'd have to have a reason for a cut off if you have a cut off
and how you treat people as tall. I don't think I think often there's a cut
off but we're not able to point to it or specify it very precisely but we can
be pretty clear. when it's been passed. So I would find it, I would be hard
pressed to tell you when the cutoff is between day and night, but I might be
able to say to my kids, look, it's too dark for you to go out. And they're
like, well, exactly how dark does it have to be for me to be able to go out?
And I might not be able to answer that question, but I'm saying, no, it's it,
it's too dark. It's not dark by this much. It's just too dark. You can't go
out. That, what I'm saying is the claim that something is bullshit. If you
just think intuitively about the character of the claim, the feel, the vibe of
calling something bullshit is not,
Robin:
um calling something something to which you give less weight than other things
can we get past your definition of the word and talk about the phenomena uh so
people you know are trying to you know insult other people and not themselves
and people have the difficulty of knowing you know when how much weight to put
on other people and what to believe right um The highest level question do we
have, is there too much or too little bullshit in the world, would we would we
even know that our most accusations of bullshit fair or unfair. These are more
actionable questions rather than just what the definition of word is or does
it have a threshold.
Agnes:
Definition is pretty important and pretty actionable. That is, I think we're
acting on it all the time. And if we wanna understand what we're actually
doing, we need a definition. And it also suggests what the issue is at stake.
So it suggests that how much bullshit you see is gonna be a function of how
cooperative you're feeling with the people around you. Because if it's your
friends, right? And your friends are in some sense inviting you to bullshit
and you might be happy to like participate. And then in that space, you're not
going to call what each of you are doing bullshitting because it's like in
the, it's, you're not, you're not going to, you're not going to be volleying
any accusations of bullshitting, right? So you're going to see more
bullshitting in a world with less trust and less friendship because anything
that even could be heard as an invitation to let's bullshit together is going
to get rejected.
Robin:
So let's take an example from the conference, say politician speeches often
are accused of being a bit bullshitty. And politicians would have the
opportunity, as with most news articles, of asserting that they are not
bullshitting if they would post bonds payable if somebody were to show them to
be false. And news organizations and politicians don't do this, so some take
this as an evidence that they are in fact somewhat bullshitting because they
don't take up this opportunity, relatively straightforward, of showing that
they are trying to meet high evidentiary standards. Okay, so is this correct
that politicians are bullshitting? Is it correct that the customers of
politicians want them to bullshit? How can we tell where in our lives we want
how much bullshit? And some people seem to be thinking that other people are
allowing too much bullshit, but what would the standard be for deciding what
was too much bullshit for other people to be engaging in?
Agnes:
Yeah, I mean, I guess I think that the idea of playing along in a game of
bullshitting fits pretty well with the politician case. Because if you look to
who is going to permit or give the politicians license to talk without posting
the bonds, It's going to be their constituents and the people who agree with
them, right? The people who want them to start posting bonds are going to be
their opponents. And so if you want to know who is going to accept your bid to
speak with lowered epistemic standards, the answer is going to be your
friends. Your enemies are not going to accept that bid. Now, the question is,
what would a world look like in which we always maxed out our epistemic
standards for every communication? I think maybe you've described it in terms
of a world where everything you said you had to post a bond. And nobody wants
that world and it seems like an incredibly unfriendly world. And so what that
reveals is that in any interaction, in any interchange in which people have
even the slightest bit of friendliness towards each other, they are in some
sense tolerant of a little bit of bullshit. um they're tolerant of a little
bit of like we're gonna be we're not gonna be so precise we're not gonna have
such high standards right and like i was a bit salty towards you like a little
while ways back when you said of my definition um oh you're you're just saying
that they're not meeting epistemic standards i'm like no you're not being
perfectly precise my actual definition said um that that you're you're just
not accepting their bid for lower epistemic So in insisting on that level of
precision, one thing, you could say I was being pedantic, or you could say you
were bullshitting. We could put it in either of those ways. If you're like,
look, that difference doesn't matter that much. And I'm like, no, it does
matter. Every difference matters. And so it seems to me that the world of zero
bullshitting is just the world where there's not enough trust and love and
fellow feeling for us to tolerate any relaxing of any epistemic standard at
any time.
Robin:
So I could give some more examples with respect to the question of, do we have
too much or too little? So for example, in the law about fraud, we're actually
quite strict. So a favorite example is an advertisement in the back of a
magazine that says, surefire bug killer. And if you send your money to them,
they send you back two blocks of wood. It says, instructions, place bug
between blocks, squish. And it is therefore a surefire bug theory that they
were successfully sued for fraud, because we've decided that that's bullshit
in the context of magazine ads. On the other hand, there was a set of cases
where states passed laws to make it illegal to claim that you had military
medals when you don't have military medals, and then the Supreme Court struck
that down, saying you have a free speech right to lie about military medals.
So we actually have quite a complicated variation in these standards. And
there's defamation law, which basically acts as if you have been required to
post a bond for some sort of claims you make with each other. And defamation
law has a lot of variety in when it's strong or weak. So we clearly do vary
our standards. for clarity and epistemic standards, but then we often apply
high standards to our opponents and rivals. and to businesses and to rich
people and to, you know, the foreigners, and we'll apply low standards to our
friends and lovers and family and et cetera. And so, you know, do we think
that these standards are actually tracking something important in terms of
when they're high or low, or are we just opportunistically using them to dump
on people we don't like?
Agnes:
Well, sort of more generally, it's not the question of applying which
standards where it's the question of who are we willing to relax our standards
for. And so my definition predicts that you're going to find more bullshit in
a world with more hostility because there's going to be more of a refusal to
relax the standards. And then that refusal lends to what the other person
saying being called bullshit. Right. So a world in which two worlds where all
of the beliefs are equally well evidenced. But in one world, there's more
hostility. In that world, there's going to be more bullshit. So bullshit
doesn't just track epistemic equality, it tracks- Yes, according to your
definition of the word.
Robin:
But if we set aside the word and just ask- But I think it's true.
Agnes:
I think that that picks out a truth about the world. That's why it's useful to
get definitions, because I think this was a concept worth defining, and now we
learned something. about bullshit. That is, we learn that there's something
that looks like an epistemic problem that could be a social problem. So that's
like an insight that we get by way of this definition, and only if we actually
try to define the word carefully.
Robin:
So you've talked about relaxing standards, but I think there's a sense in
which the default standards are already pretty loose, and what we often do is
invoke situations where we have unusually strong standards rather than
Agnes:
I agree. And so, so wait, can I add a caveat? So I think there's a problem
with my definition, which is that I think sometimes you can bullshit by
raising the standards too high. You know, where like, um uh uh like where like
somebody makes an objection to what you're saying and they haven't perfectly
worded it or something or they you know and you kind of know what they're
getting at and you could just respond but instead you're like pedantic about
it and i think that being pedantic can be a way of bullshitting it can be a
way of in some sense evading a certain set of epistemic standards by insisting
on too high a set of epistemic standards that's a It's a complicated case, I'm
not sure how to think about it, but I do think actually it is possible to vote
by raising the standards.
Robin:
So we could just say people are selective with their standards, they are
context-dependent with their standards, and the standards they invoke may not
be optimally truth-conducive. So certainly when people talk about, say,
academic studies or especially statistical studies, most everybody who knows
about statistics and statistical studies knows a number of ways faults you can
point to any study. Almost none of them can fail to have some of those faults.
And so a common tactic is to find the faults that that study embodies and then
dismiss it on the basis of you couldn't trust a study with those faults.
Agnes:
Exactly. And that would be a way of bullshitting by way of high epistemic
standards.
Robin:
Because in some sense, reasonable epistemic standards would have you put some
weight on these studies, even though they have faults. And so you are, in some
sense.
Agnes:
I agree. And so maybe the definition should be modified to something like
that. So I still think it's right that bullshitting is people agreeing to
relaxed epistemic standards. That is, when people are bullshitting, they're
never raising it above a certain point. They're doing the relaxing thing. Is
that even true? Is that true? Maybe sometimes actually like intellectuals or
academics or something can bullshit by being like hyper precise in a jokey
way. Maybe that is a thing. So maybe it's something like wrong epistemic
standards. That is ones that we know that are wrong, but we're going to go
with it or something. And then when you call someone a bullshitter, you think
they're asking you to buy into this wrong epistemic standard and you're
refusing to buy in.
Robin:
So I do think often intellectuals or academics will take a sort of natural
environment when which people are talking and offering evidence and vibes, and
then their consensus or opinion will move but it won't have been justified
fully by the literal evidence that was offered. often it will move in part by
the vibes they express to each other and just the consensus they perceive and
you know the question is an academic might call that bullshit because it's not
meeting academic standards of they didn't show sufficient evidence to justify
the change in opinion but we might say but in fact people's vibes and their
opinions are evidence and this was rational for them to move in this way And I
think that's a common conflict academics have when they enter the larger world
and people making decisions and drawing conclusions is we often draw, almost
all non-academics draw conclusions that do not, you know, on evidence that
don't meet academic standards. That is most of the evidence that people use to
draw conclusions in the world would just not make the standards of peer review
for a paper. You couldn't write a paper without evidence and have it accepted
for peer review people and say, no, no, you know, you haven't shown this you
haven't done this variation and you haven't precise about this.
Agnes:
And so, so, so, sorry, one thing this is a slight like digression but like
your talk at the conference was about how the concept of bullshit. is really a
concept that sort of like academics or intellectuals use because for us truth
is sacred and so we're offended when someone doesn't care as much about the
truth as we do and we call that bullshit. And it occurs to me that one bit of
evidence against that analysis of bullshit is that I think academics don't use
the word bullshit very much and they don't use it as much as the most routine
context in which the word shows up is going to be in politics. It's not going
to be by academics. It's going to be by political people towards other
political people, people who are not especially academic and don't have
especially high academic standards. And so that suggests it's less about
seeing the truth as sacred, and it's more about some kind of group schism or
something, like as I'm presenting it to be.
Robin:
So as we've discussed before, I've been pursuing this concept of prediction
markets for many decades. And there's good arguments that prediction markets
do meet high epistemic standards. And there's also a puzzling, to some degree,
lack of interest. And so then part of it can be attributed to the fact that it
is applying overly strong standards compared to the standards that some
communities prefer. And then that makes us pause and wonder, like you could
easily understand that maybe high epistemic standards were just a lot of
trouble and extra work. And that you might therefore relax them in some
context for the purpose of saving that trouble. But this suggests there's more
than just that effect going on. There's gonna be often positive reasons why
people would like to lower the standards. And then to puzzle what those might
be.
Agnes:
Okay, so I just came up with like a really good, my definition, which you hate
so much, the fact that I'm producing a definition, but it's giving you a
potential explanation for resistance to prediction markets. So first thing to
notice is that the explanation that you gave in your talk, which is like
intellectuals love the sacred, would predict that intellectuals would love
prediction markets and they don't. So that's already a problem. That is
intellectuals, if they see truth as sacred, they would be like, this is so
great, we're getting more of the truth, but they're not doing that. And then
this other part is, like, it looks like we value, we have a positive value on
lower specific standards, on being able to lower our standards. And the
question is, why would we value that? And my theory gives you an answer, which
is that it is a social signal of bonding, right? So it's like, if you and I
like each other, then we're gonna let each other bullshit. That's what it
means for us, not you and I. You and I are an exception. But people in
general, that's what it is to be a friend to someone, is you're not gonna call
them on their bullshit all the time. You're gonna, you know, every once in a
while you will when it's really gonna help them, but mostly you're gonna let
each other bullshit. And if we have prediction markets, you know, we could
never be friends.
Robin:
So we could test this against common business applications of prediction
markets and their obstacles that we've seen through more, you know, experience
there. For example, when projects have deadlines, A prediction market can give
a much more reliable, accurate estimate of whether a project will make a
deadline, but typically people running projects don't like that. They want to
be, what's they say, in control of the narrative. They want to say whether the
project's going to make the deadline or not and have people believe that. And
they even orchestrate project status meetings where people will repeat the
suggestion they give to them about that they're on track to make the deadline.
And so this is a concrete example of where people don't want to meet higher
standards because they want to be able to control and shape the narrative of
that they're going to make the deadline. This isn't just so much about people
bonding and liking each other. There seems to be other reasons there.
Agnes:
Because look, if it were just like some one guy wants to control the
narrative, All the other guys would be like, no, we don't want you to do that.
We want to make money or whatever. But in fact, the reason why the rest of the
company or whatever isn't overthrowing that guy who's like, I want to be able
to just control all of you and make less money because I want to be in charge,
is that there are factions and there are groups and there's a group that
supports that guy. And that group wants to be able to be the group that
supports that guy. They want to be able to stick together as a group and
prediction markets would destroy that. That's the problem. It's not because
that one guy wants to control things, it's because the group that that guy's
part of wants to stay.
Robin:
Right, but it's not just that that group happens to want to bond together,
that is groups often- All groups want to bond together. But they don't just
want to bond together, that is a faction in a corporation that is pushing a
narrative is doing that in substantial part because they see that narrative as
advancing their agendas.
Agnes:
I see that narrative as advancing their identity as a group. I mean, that
might be what their agenda is. If we're taking a Durkheim view, going back to
the sacred, the groups might be the bottom of the story. That is, the
fundamental thing is that we want our group to be in charge so that our group
can be the number one group. At least that should be on the table as a
possibility, since you were using that framework in your explanation.
Robin:
We read this book together a while back about corporate politics in the 80s. I
can't remember what his name was.
Agnes:
We did a podcast on it.
Robin:
Right. Anyway.
Agnes:
Moral mazes?
Robin:
Yes. Moral mazes. I think, but basically that and many other corporate
politics stuff I've analyzed over the years definitely gives me the strong
impression that it's not just merely that a group wants to bind around some
arbitrary shared thing, that they have shared agendas and interests that they
pursue via shared narratives that they push. And that's part of, and
prediction markers would often interfere with their construction and
maintenance of a shared narrative. For example, a shared narrative will have
that their project should continue, their people should stay in their
positions of power, etc. their designs are good designs, and often rival
groups are saying that their project should be ended, their designs are bad,
should be replaced, their people should be displaced and replaced, right? And
that groups have these conflicts over concrete choices that would advantage
some or the other in organizations, and their relaxed epistemic standards
often facilitate those narratives and their maintenance.
Agnes:
Yeah, I mean, I guess it's gonna be a combination of two things, right? So
it's gonna be a combination of the self-interest of the members of the group
as they conceive it. So maybe they have peculiar desires for certain end
results or just for making money. And then it's gonna be about their,
membership in that group and their ability to pursue an agenda by means of a
group. That is, you couldn't really pursue almost any agenda without a group.
The fact that the prediction market poses a threat to the stability of the
group is going to stand in the way of almost anyone trying to ever do
anything, given their conception of what it is to do something is to do it
with a group.
Robin:
Much of our history is wrapped up in stories of often what we've decided are
villainous groups maintaining a narrative that didn't meet epistemic standards
and a heroic group that fought for better standards and then better practices.
You think of the Salem witch trials or, you know, using hand washing hands for
doctors or, you know, also, or thinking that there were weapons of mass
destruction in Iraq. Right. We often have these stories of conflicting groups
with conflicting narratives and then the
Agnes:
a winning group, winning in part by its attention to higher standards, and
that we celebrate that, looking at least back in history, making us- If we
went to any one of those winning groups, and we examined all of their
propositions, we would find many, many flaws, and we'd find many false things
that they believed, going back, I'm sure, to Ignaz Semmelweis and
hand-washing, right? I'm sure there's a lot of stuff he got wrong about germs,
right? you got one important thing right and that's when we remember and
that's we sort of just treat that as that's what he thought but maybe we don't
want like all of our claims to be examined so carefully because our group
coheres on the basis of like a vague unity in a sense that we got something
right and a more accurate group conception is not going to be worth bonding
Robin:
But surely as a philosopher, you might be able to comprehend this larger
stance by which humanity might say these groups have been pushing their
narratives are often hindering humanity's advance and that the raising of
epistemic standards tends to be correlated with us all doing better by being
able to defy mistaken group consensus.
Agnes:
might be that raising the standards a little bit about one specific thing in
one context does help, but it may just be that the ability to lower standards
is also a necessary condition and the prediction markets would destroy that.
That is, in order for all of humanity's efforts have been basically by way of
groups, right? One person doesn't achieve anything. Groups achieve things, and
they achieve things because they feel bonded to each other in part, because
they feel bonded to each other, because they feel like they can work together.
They feel like they can be as one, right? They have a sacred thing that they
have in common, and that's dark time. And if I'm right about what bullshit is,
and it's somehow just this important thing to feeling yourself as a group that
you are bullshitting together, then it may be that we just can't raise them
all. We can't raise it all the way up. We can only raise it like a little bit
about something some of the time.
Robin:
Can't I read your new book, Open Socrates, which is for sale everywhere. You
guys should try it, which we've talked about before. Can't you read that as a
call for stronger epistemic standards in certain contexts? Isn't that
substantially what it's a call for?
Agnes:
Actually, no, I don't think that's quite what it's a call for. I think what
it's a call for is like accepting people's bids. And it's going back a little
bit to the thing you said About aren't you drunk to basically the moral of
that story is you're also drunk and you're also bullshitting. And so when
someone is bullshitting, like, what you have to do is just figure out how you
can learn from them. given the imperfect situation you're in, because you're
always in an imperfect situation. And this thing that you would really like,
which is to have an excuse on the basis of which to dismiss a large portion of
the population, ideally all those people who disagree with you, you should
just not go looking for that. But that's not the same thing as saying we want
to max out all the epistemic standards. I didn't say max out.
Robin:
I said it's a bid for a higher standard. So for example, you're saying you
could dismiss them, but don't. That seems to me an epistemic standard. I don't
think so, because I think that, like, you're saying find somebody, for
example, who doesn't share your load-bearing beliefs in order to challenge
yourself. That seems like a standard. You're saying meet that higher standard,
and you will do better, you might be tempted not to meet that standard, not to
engage with them, not to try to feud with them, not to find such people to
talk to.
Agnes:
I don't think that's what I mean by meeting standards. So I'm here. I'm not
talking about what standards we have or how high they are. I'm talking about
bids. People are making bids, and you have to decide whether or not to accept
them. That's what bullshit is about. It's not about how high the standards
are. It's about whether you accept the bid or not. And whether you accept the
bid or not is not necessarily a function of how high or low your standards
are. You might accept it because they're your friend.
Robin:
um okay but again i'm not trying to talk about the word bullshit people like
your friend i'm not trying to talk about the word bullshit and its official
definition i'm trying to talk about actual behavior i'm saying you are
recommending higher epistemic standards in your book that's what i'm saying
you are doing not i'm not saying you're referring to the word bullshit in your
book i'm trying to not talk about bullshit i'm talking about
Agnes:
I think that's not the most enlightening way of tying my book to the
conversation that we've had. The way that I would tie it instead is to say
that, like, is to say, actually, maybe your epistemic standards should vary in
different conversations. And in fact, there's a point and I talk about this in
one of the chapters where Socrates is like, let's lower epistemic standards,
that can be fine. but to put the emphasis on, not on the standards, but on the
conversational bids, and to say there are bids out there and you should accept
them because you're more desperate than you think you are.
Robin:
And I call that a standard. Yes.
Agnes:
You can call it whatever you want, but I can also say that's not a good way to
express the point of my book.
Robin:
Basically, there's different behaviors and some of them can be described as
better epistemic behaviors and you're recommending the better ones. That's
what I would mean by a higher epistemic standard.
Agnes:
My book is trying to tell you that your life should be about the pursuit of
knowledge. Yes.
Robin:
And if you do that, instead of the other things you were likely to do, you are
behaving in a way that I would say meets a higher epistemic standard because
that's more likely to produce insight than the alternative behavior. So you
are making a bid for higher standards.
Agnes:
Okay. Maybe that's fair.
Robin:
Well then, you know, you aren't quite so agnostic as you were saying a moment
ago about we don't really know anywhere about, you know, whether our standards
are too high or too low. Clearly, you're making a big bid of saying in a big
way, our standards are way too low.
Agnes:
I think that's right. Maybe a way to put the point of my book is that we can
turn this process of making bids for epistemic standards, we can make them
really high. And that can still be bonding. We can bond over that. Whereas
what prediction markets look like they're doing is just like taking the
bonding out of the story. I can put the bonding back in the story and raise
the epistemic.
Robin:
I'd say prediction market participants have ways they bond with each other,
but they are not quite the same kind you're thinking of.
Agnes:
Yeah, I mean, the point is, from the point of view of people who are bonding
in the normal way, the prediction market looks like a threat. There's always
some alternative way you can bond. You're never interested in that when you're
bonded already to a certain group of people, because that's what bonds are.
You're not willing to give them up for some other bonds, right? Oh, I'll take
your family away, but I'll give you a new family. That doesn't work. So your
move to introduce prediction markets looks like it threatens the bonds we
already have. And what I'm trying to argue with my book is there's a way to
raise epistemic standards that strengthen the bonds we already have.
Robin:
Although I might say prediction markets could also strengthen bonds, but
that's a claim about the actual effect of the behavior.
Agnes:
I think not the ones we already have, like not the ones in that company that
are based on not looking at certain things. They're not going to strengthen
those bonds. you might get new bonds and those might be stronger. But again,
that's like offering someone a new family. It might be a better family. They
might still be like, I'm gonna stick to my old family.
Robin:
Okay, so in summary, we both see a world of varying epistemic standards. And
then sometimes people complain about somebody's standards often through the
label of bullshit. And then we know that standards vary a lot in by context,
but we both have differing ways in which we would like to raise standards. We
would like to recommend higher standards in different ways, but not obviously
in conflict ways. And we are still struggling to understand how we can explain
sort of why the levels of standards are the levels they are and therefore
inadequate in the context of trying to help people raise the standards. That
is part of the puzzle is why aren't the standards already high if we're trying
to recommend raising them.
Agnes:
Yeah, I mean, I guess, I think that doesn't have much to do with the topic of
bullshit. That is, I think that, so that's why, like, I thought we were going
to talk about bullshit, and bullshit is about accepting bids for lowered
standards, but then there's the question, absolutely, how high they are and
how high should they be?
Robin:
That seems to me to be a different- Well, bullshit is also about the phenomena
of accusing people of having overly low standards.
Agnes:
Sure, but you could be doing that even if all the standards were really pretty
high. And a lot of period bullshit, you're going to be doing that in those
circumstances.
Robin:
Right, but it does suggest there's this widespread sympathy for higher
standards. That is, if we're willing to accuse people of having standards too
low and using it to label bullshit, it does suggest there's a demand of a bit
out there for higher standards.
Agnes:
I agree. And I think that when we're bullshitting and we agree implicitly to
the lowered standards, it's always only implicitly. We're never going to be
explicit about that because it's a little bit shameful to be lowering our
standards. And so it'll be like as a joke or an ironic or whatever that we
might say that.
Robin:
Right. So I think we're about out of time. So thanks for talking.