Who Must Be Persuaded By Explanations
Today we're going to talk about who explanations need to persuade. And I
realized that when I first learned game theory, one definition that one of my
professors gave for game theory was the game theory are the theories of human
behavior where people don't change their behavior when they believe the
theories. They could be aware of the theories and aware that they applied to
them, and that they still behave that way. So in that sense, game theory is a
description of behavior that each person can fully embrace and they still
behave according to the theory. So now, that's not to say I guess that they
would be eager to believe the theory. Or that we would win many points by
trying to persuade them of a theory, they maybe wouldn't like the theory, and
wouldn't be willing to believe it. But that's at least one interesting framing
on who you need to persuade. So I guess the implication there is there are
other theories of behavior that are accurate theories, but only when people
don't believe them when they're about, say, people's delusions or
misunderstandings. And I guess if people have misunderstandings, then theories
of those misunderstandings can describe their behavior. But they wouldn't
behave the same if they believe the theory.
There's a phrase that I think it was Bernard, philosopher Bernard Williams
used about a certain kind of utilitarianism that he called Government House
utilitarianism, and what he meant was, like, the idea that a utilitarian who
wants people to be Kantians, because they think that overall, you'll get the
best results if people believe the false theory. So that would be that would
be an example of, you know, if they found out the truth, they would then adopt
the view that in fact, would lead them to act in the worst way.
So then, Government House utilitarian is not part of game theory, according to
Hmm. So that's puzzling to me, I mean, like, is the idea that the claim about
game theory, this is assuming– is this assuming that all the actors in the
game are playing as well as – they are behaving as well as rationally as
possible to start with?
Well, we can have variations on game theories where people have trouble
calculating things, and they can't exactly figure out the optimal thing. And
then we can have game theories describing that, where there's a stochastic
distribution over what they do, given the various mistakes they'll make.
That's still part of game theory. But you could see that theory and see that
it applied to you and still want to behave that way.
But like, wouldn't you if you say, understood, which sorts of mistakes you
typically made, couldn't you perhaps take measures to avoid those mistakes?
Perhaps, but it's actually pretty hard. Knowing in the abstract that you make
mistakes is different than knowing which particular mistakes you're likely to
make in any particular situation. So the abstract theory would just describe
the distribution of mistakes, and not necessarily which mistake you were about
But like, is it in prin– this is what I'm, what I'm not getting is like, is it
in principle impossible that there could be a situation where people are
inclined to make a particular sort of mistake, and they don't know that
they're making it? And then, you know, if they learn that they would change
Yes. And that's possible, but that's then excluded from game theory. That's
why this definition is that's not in part of game theory.
Yes. So that’s what I’m not getting, like, why is that intelligible line to
It's just interesting, that knowing about theories of them could change
people's behaviors. So that's – like, the physical world doesn't act different
when we have different theories of it. Mountains don't, the stars don't,
rivers don’t. And so you might think this is an interesting distinctive
feature of social science that you have these people who are the object of
your theories and what they believe about your theory will affect their
behavior. So then when you theorize you might change the world, because you
might convince people of your theories and then they would act differently.
And that adds an interesting complexity to Social and Human Sciences.
And that – it's at that point that we're no longer doing game theory, right?
That is, well, so if you were, say, ignorant of game theory, and then you were
behaving one way, and then I explained game theory to another way, and you
behave another way, it could still be true that knowing game theory, you know,
isn't changing your behavior once you know game theory. I don't tell you – if
I tell you the game theory in a particular case or something, I won't change
your behavior then but I might have changed it from some other contexts where
you weren't even thinking about what you're doing. But I guess it's somewhat
ambiguous there. What the reference case is for this not changing your
behavior, you know, certainly if you're not paying any attention whatsoever to
your behavior, then pretty much any reasonable theory of behavior would
probably change it.
Also, there are kinds of theories that wouldn't change it even though there'll
be nothing else interesting about the theory. Like the theory that included
you're fated to behave exactly as you have behaved, or something like the
fates are controlling you, well, then nothing else I say, you know, would
change your behavior because you think it's not under your control. Right?
Right. That would be a theory you couldn't compute very easily or you couldn't
figure out what the theorist implies specifically about your behavior. You
just know in the abstract, say that your behavior is consistent with physics.
Knowing that doesn't help you very much know what to do. But the reason why I
suggested this particular topic is that in the past, you and I have had
differences of stances or priorities in terms of when we have a theory about
some people and this theory may not be very flattering of them, or may not be
a theory they feel very inclined to embrace, my habit will be to continue to
elaborate the theory and maybe talk to other people about it, not them. And
your habit is more to think about how you could talk to them, and maybe focus
on the kinds of theories that you would be able to get them to buy into. And
this is a different stance, the two of us take toward our theorizing. And I
thought it might be interesting for us to explore that here.
Yeah. So, and like, I guess, I think there's the facet of my view that often
shows up in conversation with you is, you want to be able to say unflattering
things, give unflattering descriptions that people would reject. But for me,
this is part of a bigger claim, which is just that knowledge of the ethical
world is practical knowledge. So it's supposed to give you guidance as to how
to act. And so knowledge that doesn't give you any kind of guidance, where it
would be equally true that that speech that simply flattered someone, right,
so you say, “Oh, it's bad, because it's not flattering.” But I think it must
be bad if it were flattering. Like the point at which you think you've
understood it is the point at which you can use that to reorient yourself and
to improve. And that's just what it means to know stuff in the practical space
is to improve.
Sure. But there's a difference between a theory that you will be willing to
accept, and likely to accept, or find congenial to accept, and a theory that
if you accepted it would help you behave. So you just described that second
criteria. It should be something that if you believed it and act on it, would
actually help you. That's a different criterion than whether you're willing to
Well, if I'm imagining myself giving the theory, it's only going to help the
person if they're at least somewhat inclined to accept it. So getting them to
accept it is just part of the game. It just as much as it orienting them in an
improving manner. I haven't done something useful if I've produced something
that I know they can't – for instance, if it's spoken in a language that they
don't know, then I can be sure they won't accept it, and I haven't given them
But it could still be a valid theory.
I don't think it could. So I think validity in the practical realm is
practical. It is, does it have some kind of use? If it's useless, it's not
So many of us, including you and I, are intellectuals. We are professional
intellectuals. And we are part of a large intellectual project whereby we all
think about a lot of different topics and we try to pull together a unified
view of everything through our combined efforts. And, you know, any one person
might be able to use this combined view for practical purposes. And it would
be, you know, useful for someone to summarize all the stuff in order to help a
particular kind of person, make a particular kind of decision. But this
exercise of collecting a unified integrated view of everything, that exercise
seems to, you know, benefit from any relevant true things we can figure out
about it, even if particular people wouldn't want to hear that. That is, we
might just put together our best view of everything, and then look at who
would be willing to listen to which parts of it and then craft practical
advice for each person based on which parts they would be willing to accept.
And that might not be able to draw on the entire structure we have put
together but seems like we should first just try to put together the best
shared structure we can about what's true about all the important things. And
then as a separate matter focus on, for any one audience, what they would be
able to benefit from.
So you often, like conjure up for me this category of intellectuals, which is
supposed to include me and you and some other people but not everybody. And
then you give me a bunch of rules for what these people do. And I rarely
recognize myself in this category. And, like, I don't think I'm trying to
produce a unified integrated view of everything. That's not a project I see
myself as involved in, or it's too…
You’re contributing to it, even if you're not focused on it, or it's not your
highest priority. But we are all together doing that.
I think that I'm doing like more specific things in more specific contexts.
Like I'm trying to teach some students something about Socrates, or right now
I'm trying to figure something out, talking to you about whether, you know,
whether practical knowledge really does have to be practical, as I would put
it. And I think you're – I mean, I do agree that there is a kind of general
human endeavor of trying to know, but at that point I'm not sure that there's
like, some people who are – that anyone is excluded from it. That is, that
there are like experts, who are experts at that level of generality of being
intellectuals but not being, you know, specialists in something. That's just a
category that I don't use. It's not how I think.
I wonder if we could ask about some other class of things, whether we can
distinguish the practical from the non-practical. So if you imagine walking
into a department store. You know, maybe everything in there could be used by
somebody, but maybe most of it couldn't be used by you. You know, but we still
suggest that you go through the department store to find the things you want.
We don't make a whole separate store just for you for the things that will
only be useful for you.
We share a larger set of things that some of which may be useful for no one, I
guess, in a practical sense, they might enjoy them in some other aesthetic
sense. So, you know, when we have categories of objects in the world, we have
organized our sense of what kind of things there are out there. And we have
ways to categorize them, we also put them in shared spaces, like a section of
the department store or something. And seems to me like, we mainly first want
to have just a set of sensible categories for things and places to put them
and then not focus that so much on whether it's useful to you. We don't want
to organize all the things in the department store that you will happen to
find useful. So the question is whether this concept of practical is a sort of
stable fundamental category that we would organize things in terms of, or
whether it's more temporary and context dependent and local, such that you'll
have to wait until some context to figure out what happens to be practical
there. And so we shouldn't talk about whether knowledge in general is
practical, we should talk about whether it's applicable in some situation to
some particular person. The same piece of knowledge might be applicable by
somebody else, one person not so applicable by someone else, and could be used
for this project of putting together an integrated view of the world.
So like, I think a department store is very much organized in the manner of
practical knowledge. Namely, it's organized by use, not for one person, but
for people in general. And so there's stuff in the department store, there's
stuff that would never be in a department store exactly, because it wasn't
imaginably useful for anyone. And the categories, the way the department store
is divided up, it'll be like, all the cooking stuff is in one spot, instead of
having like everything that's red be in one spot. You know, or everything that
weighs over a certain amount being in one spot. That is there are a lot of
ways that we might see theoretically organize our knowledge of all the objects
in the department store that we would never think of doing in a department
store, because the department store is oriented towards practice. And when
you're walking around the department store, you're not doing a theoretical
investigation, you're not cataloging, what you're doing is looking for
something you can use and looking in a way that is liable to land you on
something you could use.
So our scenario here on discussion is, you're imagining a particular audience.
And I might have a claim about that particular audience. And it might happen
that that particular audience is not very open to that claim. They find it not
flattering, or not very congenial to their point of view. But other people
might find that claim useful. And we might find it useful for this purpose of
putting together an integrated view. So then whether it's practical depends on
who we're talking about, then it's not about the thing itself being practical
or not, it's about who will be able to use it, how.
And so, that's like the department store, right? So I'm saying, let's just
collect an integrated view of everything we can understand that seems useful
to anyone for any purpose. And then in any – in a particular context, let's
figure out what one person might want to hear. But let that be the secondary,
you know, afterthought to first figuring out a true integrated view of
So I guess I don't think those two projects are so separate. That is, there
are bits of reality that we only investigate. There's lots of reality we just
aren't interested and don't investigate at all. And then there's bits of it,
we investigate primarily because we're interested in change or improvement.
And I think that you know, it is important that we like, be guided in our
inquiries, to learn the things that could potentially be useful to us about
those things, because that's the only reason we're doing it is because we want
to find the useful things. And so, if we were making kitchen gadgets, and we
forgot to pay attention to like, what size are humans, and we ended up making
some like the size of like, you know, a grasshopper and the other the size of
like the Empire State Building, and there were all these kitchen implements
and we’ll be like, “Well, look, we're just going to collect them, and then
later, we'll decide like which ones we want.” That wouldn't be – that's not
the actual way to end up with useful knowledge.
Let's take an example of the sort of claim that I might have been trying to
investigate in this, you know, more abstract, general way that you would say
is less practical, because someone might find it hard to accept. So in my
book, The Elephant in the Brain, I guess one of the hardest chapters for many
to accept would be the chapter on medicine, wherein we say, the function of
medicine is less to make you healthy, and more to let you show that you care
about others and others show they care about you. And you might perhaps object
that that's a hard view to get a person to accept about themselves. And
therefore it's less practical. Am I correct in summarizing that?
OK. And then, I might say, but this view about people and their motives in
medicine could be useful for the purpose of figuring out whether we should
subsidize or tax medicine, for various aspects of public policy about
medicine, and, in fact, that's the primary use we recommend of the book. That
is, it's primarily recommended for policymakers and social scientists to have
a better view of how the world works and the motives people have for main
actions, and not so much for individual self-improvement. In fact, we warn
people that our book is not designed for self-improvement and that may not go
well, people may well, you know, find it harder to improve themselves than
they would have hoped. So is this other purpose OK?
I don't think your book really is oriented towards that group of people, even
if you say it is. That is, I'd be surprised if you just took all the responses
to the book and all the effect of the book, probably a tiny fraction of those
is policymakers and whatever. And like most of them is individual people
reading and reflecting about their lives. So it could just be like, you
pitched the book badly and you wrote it badly, if you wanted it to go to that
audience, because it didn't go to that audience. It went to a different
audience. And I'm evaluating it with reference to its actual audience, not its
with reference to its spoken audience. And maybe its actual audience was
really the audience that you had in mind, even if you said the other audience,
right? So like, because you really wrote it for the audience that it got, I
think. It's very naturally read. It's sort of like, it's sort of like make
sense that you had to say this book isn't for self-improvement, because
everything about it screams this book is for self-improvement.
Well, if there are multiple uses, potentially, of a fact or truth about the
world, there are different kinds of people who could get different practical
advantage from it. If you first just think about what's true, and then think
about who might get an advantage from it, you might well make a mistake of
making the format and style and marketing of it mistarget your group, but
isn't it still a valid thing to be doing to produce truths that are
practically useful by someone? And many readers seem to like this book. So…
So doesn't that suggest some of them are getting something out of it, even if
you say it's impractical?
So first of all, like, I don't think it really does give you guidance as a
public policy person. And the reason why it doesn't is because it's just not
at all clear. Like if people mostly are using medicine to show that they care,
it's not at all clear whether that means we should have less medicine, more
medicine, like, there's no clear implication of that. So like I don't actually
think it gives people guidance in those decisions. And the, I guess the thing
I wanted to say about the – let's say we could define – think of something
like trolling, right? So as we could define trolling someone as a certain kind
of trolling as saying something with sort of the aim of provoking someone
where, like, you might restrict yourself to truthful trolling, right? So you
might say, “I'm only going to tell you truths, but I'm going to tell you, the
ones that I know will sort of bother you, but in ways that are not productive
and won't help you improve, or really, there's nothing you can do anything
about. There's nothing you can do about them.” And like if we want to say, you
know, could that be in any way objectionable, truthful trolling? Like, it
seems, at least to me, in principle, like there could be some kind of
objection to that. And there could be an objection to it even if the person
wasn't doing that intentionally.
I would say that the vast majority of public policy and social science
analysis of human behavior is that a distance from drawing strong, confident
conclusions about public policies. That is, we are in a shared enterprise,
where we have broken the domain up into many specific topics and areas. And
one of those areas is final policy conclusions drawn from our consensus about
other things, but people often also just specialize in elaborating and making
more details on other things. So the, I think true claim is that most health
policy analysis has been based on the assumption of a different motive
regarding medicine. And if that motive is wrong, that suggests that most of
that analysis is wrong. And our replacing an incorrect assumption and pointing
it out with a more correct assumption about the basic nature of the entire
area of human behavior there seems like it's pretty presumptively useful for
producing policy. Unless you think we're never ever going to know about
anything about useful policy in this areas, it would seem like having a decent
grounding in the basic facts of this area would be substantially relevant for
policy. It doesn't require a direct connection between each fact and a
particular final policy for that to be true.
I'm not sure it's any more relevant at the group level than at the individual
level. That is individually if you know you're using medicine to show that you
care, but you – if there's no conclusion that's supposed to be drawn about
that from an individual about how they should behave, then, you know, a
policymaker now learns, “OK, here's this motive that nobody will admit, and
nobody to whom I'm proposing my policies like, we all know that the people who
are taking the medicine won't admit it. Here's this other motive. All of our
policies are based on a false assumption.” It's just still not clear like, to
me, it's not any more clear what follows from that, than what will follow at
the individual level. The individual also thinks has the false assumption that
they want medicine for more health.
Remember the distinction between willingness to believe and usefulness if you
did believe. So, I would say, I do believe the theory of my book. And so given
that I believe that I do think it has practical personal implications. And I'm
happy to elaborate those. That is, the key message is that, overall, medicine
is not very useful for health but it is useful to reassure people that you
care about them, and they care about you. And so when I'm in a situation where
health is at risk, I realize this fact, and so I'm less concerned with making
sure all possible medical treatment happens because I realized that that's
actually not very effective. But I realized that I should make sure that the
people involved feel cared for. That they’ve noticed that other people are
concerned about them, and that other people are willing to pay costs to make
sure that they are taken care of. I focus then on this motive of showing that
you care, as my more primary focus in my interaction with medicine. Now
sometimes you'll have cases where the evidence for effectiveness is relatively
strong. And then I can focus more on the effectiveness of medicine. But in
most cases, in fact, the evidence is relatively weak about effectiveness. But
still the desire to make sure you get the sort of care that cared-for people
get is strong. And I still acknowledge that and make sure that happens.
But so like, just take all of the medical decisions you've made, ever since
you've started having this view. Medical decisions about yourself and your
loved ones, like by what like percentile do they deviate from someone who
didn't have all those – that theorizing? Like what percent less medicine have
Well, the timing here would be about when I realized that medicine wasn't very
effective. That's a different timing than when I had an explanation for it.
OK. We can start with when you had the explanation.
No, but I'm saying is, having the explanation wouldn't change my behavior, it
would be knowing the fact that medicine wasn't very effective that would be
initially expected to change my behavior about how much medicine I consume.
I mean, you might be unwilling to change your behavior until you had an
explanation or something like that just, yeah, but in any case.
But basically, this has been over 20 years now.
So how – so what percent less medicine have you consumed over the past 20
years than your counterpart who did not realize any of this?
I think, you know, 30% to 50%?
Really? OK. OK, that's pretty substantial.
So again, it is practical advice to someone who believes it. It's harder to
persuade someone to believe it as the person themselves involved. But as
policymakers thinking about other people, I think it's more possible to
persuade them. And that's one of the reasons for focusing on social scientists
and policymakers as the audience because they're easier person to persuade.
But nevertheless, I do think many individuals are persuadable and have been
But the social scientists… well, so it's another claim if you think people are
persuadable, then I'm fine. Right? We were arguing what if they're not?
A fraction of them. So if you were writing an essay on this, say, for some
public venue, you might well guess that 90% of people wouldn't be persuaded by
this and therefore, choose other topics of your essay or other claims in your
essay. But if in fact 10% of people can be persuaded, that is a you know,
people who are helped by it. So it might depend on whether this is a center of
your writing or analysis, or it's a peripheral aspect.
Right. So, like, I mean with the policymakers, the policymakers know that most
people aren't going to be persuaded of this, right? So they have to make
policies that are going to be in accordance with most people's false beliefs
about why they want medicine.
Well, not quite. That is, most people, when they hear policymakers making
recommendations about policy, they don't actually know that much about the
rationale for those recommendations. There are some sort of obvious sort of
straightforward things that the policy might achieve or not achieve, that's
different than knowing the reasons that the policymakers have for making those
suggestions. So, one of the things I've said is that, traditionally, in social
science, especially in economics, we usually have the framing of our problem
as how to design institutions or mechanisms so that more people will get more
of what they want. And that problem is usually substantially hard, but we
usually think about markets or incentive contracts and others, voting and
other sorts of mechanisms then. And we analyze them in terms of whether people
will get more of the things that they usually say they want. And this new set
of analysis suggests that our institution or mechanism design problem is
harder than we thought. The new problem is that we need to find a new
mechanism or institutions such that people can continue to pretend to want the
things they've been pretending to want, but also give them more of what they
really want. So an example might be in education, most schools, schooling
pretends to be teaching you more material. And we have known for many decades,
some standard ways that people could learn more material in classes, but
people don't think that's quite as fun. And schools have just not adopted
those methods. So spaced repetition, for example, is one method that just very
consistently produces more learning. We can understand that if we say, “Well,
most people, at some level, realize school isn't about learning more material,
it's about showing off their conscientiousness and conformity and
intelligence.” And schools do achieve that. And so then we might say, “Well,
if we want to have a reform that actually people will be interested in in
education, we need to find a way to let them continue to pretend they want to
learn the material, there has to be some material they're learning. But we
need to give them a better ability to show off their intelligence or their
conscientiousness or their conformity. And then there'll be more actually
interested in that proposal.” And I think that's roughly right. But it's a
harder design problem. But now we have a better chance of getting people to
actually respond to our proposals. So we have a long history, to emphasize, of
assuming that people are right about what they say they want, designing
institutions that actually more effectively give them more of what they say
they want, and having people just not be interested.
So your thought is that like, like in the education case, look, people don't
actually want to learn. And so we may have to continue to have the facade of
learning material, but what we should be working with as policymakers is how
to better help them do signaling. And so in healthcare, people don't really
want to become healthy. And so we shouldn't improve our medical system, or get
more medical knowledge, or try to learn to live longer, because people don't
actually want those things, we just should just create the appearance of those
things, while giving them better and better opportunities to show that they
care. That's what a healthcare policy person should do?
Well, there's two constraints to keep in mind, both can stay relevant. One is
what do the customers here want? What will they respond to such that if you
offer them something new, they would be eager to get it? And then the other is
what are our priorities as the policymakers? So those are allowed…
We’re allowed to have our own priorities?
Yes, you're allowed to have your own priorities.
So wait, can they be actually having people learn stuff and real education?
They could. Sure, yes.
But in that case, you didn't give that as like a driver.
Well, the problem is, like, if you don't have some independent power to force
them to take your changes, if you can just offer changes, then you will need
to find ways to give them the things they want.
Which are to better show off and let’s say other abilities. So…
Right. But same with healthcare, people don't actually want health so you
can't give that to them because they won't accept it.
Well, depends on how many other options they have. So, that is, you could
offer a package that included both more actual health and more abilities to
show that they care and then they might go for that type of…
So you mean we do that with education and try to offer that package, you're
treating these two cases isometrically. So why do you…
No, I'm fine with that, too. I'm just saying, you won't be able to get people
to adopt to your, “they learn more” change unless you also package it with a,
“they can show off more” change. You have to combine those two things together
if you want to get both, they learn more and they approve and buy into the
Right. So how is it that you've been able in your own life to cut down 30% to
50% on health care and yet, still, like, for instance, presumably the people
around you have been horrified and have been, you know, bereft of these
opportunities to show that they care and encouraging you to get this health
care that you've been taking half or a third– or you know, only two-thirds of
what a normal person would take? And, like, how did they show that they cared
and then how do you show that you cared if you encourage your loved ones to
consume less healthcare? Because hold on, before you answer, I just want to,
like, in general, it's not that easy to show people that you care, like if you
just say to someone I care, that doesn't show that you care, right?
In general, it's kind of expansive to show that you care. And in particular,
it's important to show people that you care at certain times, like when
they're sick and stuff. And so like presumably like if it were easy to do so
doing something less costly than medicine, then we would do it. So anyway, I'm
wondering what are your tricks for showing that you care without health care?
Well, so there's my showing that I care about other people and then showing
they care about me. So right, I was talking in my 30% to 50% figure about my
own care, in which case that's about other people's showing they care about
So, first of all, I think that when people express concerns about your health,
you should listen, take it seriously, and seem appreciative of their concern.
Then if they suggest you do a particular thing, like maybe getting a doctor
checkup, or you know, appointment, fine, get a checkup, get an appointment,
have them somebody say something. But when you reach the point of a
potentially expensive treatment whose effectiveness is uncertain, that's the
point to cut back on the medical spending. I think if you have a heart to
heart discussion about it, and are engaging their concern, and you make sure
you listen to them about this, and you do other things that they might
suggest, you know, more exercise or change your diet or sleeping patterns. I
mean, most people will, in order to show that they care about you and care
about your health, they will offer many other things to do besides go to the
doctor and get some surgery, and be responsive to those other things. You
don't just have to go get the surgery some doctor recommends in order to let
people show they care about you.
OK, you may not want to answer this question on the podcast when I ask it and
you can just say pass. But what are the examples of expensive medical spending
that you have forgone? Like, are there concrete examples where you got to a
certain point and you're like, “That's too expensive? I won't take that
I'm not sure I've reached expensive things yet.
And you had to do a lot to get down 30% to 50%. And now you're telling me
that's your formula, you would still go to the doctor, but you would just stop
at the – so I'm asking where did that…
Well, so I mean, sometimes these things are like series of many visits about a
thing, right? You can just cut it off earlier and do less.
OK. But so you don't – you haven't tried to cut down your loved one’s medical
expenses, medical care by 30% to 50%. So you're still showing that you care
for them, you’re just giving them opportunities to do that for you.
Again, in order for me to show that I care about other people.
I can also not just focus on the medicine but focus on all the other aspects
of their health.
Like telling them to exercise more. That’s not going to go over well.
It could actually. That's the whole point, right? The whole point is that
other people want to know that you care about them. They want to know that
you're concerned about their health, and that you're willing to put in
substantial effort and pay costs to do so. And so, to the extent they say you
would be willing to accommodate their new exercise program by doing some other
chore while they were exercising, for example, or going out and shopping with
them for some exercise equipment, or reconsidering a diet perhaps you share
with them and what kind of places you go to lunch if they're your colleague,
or what kind of dinners you have at your home. The point is, there are many
ways you could pay costs in order to accommodate someone else's health efforts
that don't involve expensive medicine.
OK. But so like, all this has been a giant digression because if you say to
me, the things that are written in my book are practical. And I've only wrote
them in order to improve the lives of people, it's just that they're only
going to improve the lives of maybe 5% to 10% of my readers. And, you know,
that's the way it is, then I'm like, “OK, fine. You know, maybe it didn't work
for me, it can work for some other people.” There's a question about what
percent is high enough? But then you're thinking about it the way I'm thinking
about it, namely, it's designed not to teach people anything, not to add our
store of knowledge, but to improve lives, it just then improves like maybe
only a small percentage of the lives of the people who encounter it. That
would say that that's practical. It's just like not super effective as a
practical thing, but it's not totally ineffective either.
But I also am serious about this idea of a collective effort to accumulate a
more accurate view of the world. Each particular thing you learn may have some
relatively direct implications for practical action. But there's a synergy
effect of all the things we learned together, that fit together and give us a
new view of the world that just has more benefit in terms of learning a whole
new view of things.
So say you have a kid, right? And you know, when you're talking to your kid,
you often tell them the truths, I mean, you try to speak truly to your kid,
And, you know, what you're trying to do when you talk to your kid is teach
them a lot of things about the world that they can use in order both to know
stuff and to become a better person. But like, suppose that you relentlessly
were telling them like about their hidden motives, about all their defects,
about all their flaws. And in particular, you focused on ones where there was
nothing they could ever do about it, right? Someone might just say, “Well,
that's kind of mean, like to your kid.” And you're like, “Well, look, it's
part of their total package of information, and it's up to them to do what
they want with it.” You know, you still might think, “Yeah, but that's not –
it's just not actually likely to help them do anything. And I don't see why
humanity as a whole is that different from the kid.”
So imagine you have a friend or spouse even who insists on telling not only
your children, but you all the time, how to change a tire, how to change oil,
how to recalibrate the electricity in the house, how to readjust the air
conditioner, how to oil the legs on your chairs. They go on and on with all
sorts of advice that would be practical for someone else who might do that
stuff, but not you. You don't recalibrate air conditioning, or the stove, or
any of those other things. And so you might say, “Why should anybody learn
about any of those things?” Because it isn't useful to you, because you've
decided you don't want to do any of those things.
Right. That seems reasonable. But what I'm saying in that case is I don't want
to do any of those things. And I would be annoyed if someone kept trying to
tell me them.
Right. But why couldn't they tell somebody else? Why couldn't they write a
book on those things where somebody who wanted to learn about those things
could read that and learn all those things?
Sure. I feel like, I somehow feel like we're not getting to the bottom of the
issue here. Because it seems to me that, like what I, you know, what I felt in
reading your book was like, like, I thought it was a good book. Like, you
know, this makes me sound like I hate it, but I didn't hate it. But I felt
somehow that it just didn't get to the bottom of things. That there was a kind
of missed opportunity. That there was something like, like subtly too
superficial in saying, you know, oh, we just – we want to show – we want to
show that we care. We're not interested in health. Or, we're trying to show
off our – when we have conversations, we're trying to show off our toolkit of
skills. Where I feel like it's not that my– I have some requirement that the
person who gives me a theory of medicine or conversation tells me something
nice that will please me. I don't think that that's right. But I guess I think
that I want the critical thing that I hear to be kind of… so, you know how
sometimes like you make a mistake, and you're just – but you don't know where
you went wrong, right? And it drives you crazy. You're like, you make an
arithmetic mistake or whatever, right? And then at some point, you see exactly
where you went wrong and you're like, “Oh, OK, I see how I screwed that up.”
And there's like this click like, I get it. That's what I don't get in reading
your book. So it's like it feels like from a million miles away and it doesn't
click into place for me like, “Oh, this is what I'm doing when I'm having
conversations. This is how I'm screwing up when I'm having conversations.” I
mean, I think I'm capable of recognizing that I’m screwing up in doing things.
But your way of describing me screwing up is like me looking at someone else
screwing up. It never feels like me looking at me screwing up.
So when you're looking for a mistake, I think you typically have at least two
levels of representation of what you're thinking about, both of which you
consider valid, and it's about the mapping between those two levels. So
there's a more detailed level in which there are each line of the proof. And
then there may be there's the overall level of when the proof is correct. And
what you want when you see at the overall level that the proof is incorrect,
you want to go down to this detail level and find the particular place where
that is. And that is a perfectly reasonable thing to desire when you have two
levels of description, both of which are reliable, that you maintain together,
and you learn something at a more abstract level that you'd like to map it
down to a more specific level. That's a completely reasonable thing to want in
the case where you have those two levels. Unfortunately, we don't always have
those two levels. We are often in situations where we're struggling even to
have one level that works. And I'd say that the topic of our book is a bunch
of areas where we're just really kind of wrong at the high level about the
basics of it. And we just don't have a more fine-grained level that's reliable
that you could use to edit that because we're probably also just pretty wrong
at lower levels, too. So we just have to slowly find a way to build up our
representations to have a better abstract level and a better fine level. And
yes, there'll be future work to find out a more concrete level that we can
edit and rely on. But we aren't there yet. And the way to get there is to
slowly make progress. And the first place to make progress is at that abstract
level. If you asked me, you know how to get to Maine in First Street? Because
that's where the bank is. And I say, “No, you're in the wrong city. The bank
isn't at Maine in First Street.” Then that should be your first priority. Say,
“What? I'm in the wrong city. I'm not in the city, I think in?” Then you need
to like reconsider: how is it that you thought you were in a different city in
order to, you have some grip on reality. Then when you figure out what city
you're in, then you can work about how to get to Maine in First Street.
Like, here's like the spirit that your book infuses in me. Like I'm asking
myself, “Does Robin really spend 30% to 50% less on health care than other
people? I wonder if we really studied him and we compared him. Like, of
course, he'd tell himself that, right? You have all these reasons to believe
that he does that. When I tried to push him and ask him for examples. He's not
like doing a great job there.” And you know, I bet that like the people who
are convinced like the 5% are like, “Yeah, you're right.” I bet they all say
they spend less money in health care. But suppose we studied them, right? And
suppose we studied all you true believers, and we found you spend about as
much as the rest of us. And like, then should we be sort of like, extremely
disheartened that in effect, we are so bad at getting a grip on reality that
in a way is what your book is about. We cannot get a grip on reality in all
these different places, like why think we're going to get a grip on reality in
implementing the book? Why I think that the policymakers will get a grip on
reality? Like, why think we have any hope of a righting this totally toppled
It sounds like you'd be saying when I tell you you're in the wrong city, you
say, “I don't care. I give up on being able to figure out what's city I am
No, I don't think that's right because if you tell me I'm in the wrong city, I
still in the back of my head, have the idea of a map. If you tell me I'm in
the wrong city, and also, I'm on a different planet. And also, like, you know,
everyone that I've ever known has been an actor in a play and right, at a
certain point, I'm like, “OK, I give up, right? You tell me enough stuff. If
what you say is, “Look, you're in the wrong city.” And then I'll be like,
“Wait, show me how I went wrong. Why did I end up thinking… Oh, I get it,
there's another city of the same name and I flew to that city instead.” That's
the point at which will like click into place and I'll say, OK. That's
something your books not doing for me.
The key story here is that we are all just much more wrong than we realized
about a lot of big things in our lives. And that fact, if true, suggests we're
also wrong about a lot of little things. We're just big and wrong all over the
place. And the first step to dealing with that situation is to tell you that
you are wrong about big things. And so the most straightforward way to do that
is to try to name the ways that you're wrong in the most concise and general
way that we can and point you to evidence to convince you of that. It's not
the end of the story. It's the first stop that says, you know, “You're not
going to be able to fix all this until you have the foggiest idea that you
have a problem.” So here it is. So it's like the alcoholic with an
intervention, right? All the friends get together and sit around and say, “You
have an alcoholics problem, right?” And this is the point at which they need
to accept that, OK, at least all my friends seem to think I have a problem.
That's not the end of the story. You're not fixed at that point. Right? That's
supposedly the beginning of the story of trying to figure out how you could
deal with your alcoholic’s problem. So that's what we're making that first
move here. We're sitting you all down and saying, “You're wrong a lot about
these really big things.”
But like, when you say, the next step is figuring out how to deal with your
problem. I mean, this is – that was my point about, like, my skepticism about
whether you really spend 30% or 50% less, is that I think it's likely that any
attempt I make to deal with this problem will be infected by the fact that I
am susceptible to this problem and I'm susceptible to telling myself stories
that aren't the case. And so for instance, I'll tell myself, I'm dealing with
it and I'm telling myself, I'm spending 30% to 50% less when I'm not actually
doing it. That's how I'll deal with it. I'll deal with it by not dealing with
it. It should be my prediction about myself.
And if that's your response to your alcoholic’s intervention, then of course,
you're not going to fix your problem.
But that's not how alcoholic’s interventions work. That is, I don't think this
is right. I mean, I don't think just saying, you know, you… I think that at
least in many different kinds of interventions, what people try to do is make
that person's problem intelligible to them. They don't just say, “You have a
problem.” They try to get that person to see the problem, to shed a new light
on it, so that they have an aha moment, or it clicks into place and they say,
“Now, I get it.” And what I'm saying is like, you stop before that point. You
stop before the point where I get it. And so I don't even have the amount of
confidence where I'm confident that I'm getting it, like that is it can be
that my very feeling of like, “Oh, there seems to be a problem here. Why don't
– how do I know that, that that my…” like this, I felt very much when reading
your book. There's a strong temptation to flatter yourself by thinking that
you can see through things that other people can't see through. And I think,
“Ah, that's one of the dangers of this book, is that it encourages you to do
that. But it fans the fire of The Elephant in the Brain by inducing in the
reader some kind of feeling of superiority.” Like why not think, we're
actually making the problem worse by introducing a new form of self-deception,
where you think you're better than other people and more writing and where in
fact, you're not. It could actually be making it worse.
That critique that you just expressed could be applied to anytime anyone ever
points out any problems.
OK. I could say, “There's a pothole on Main Street, and it needs to be
filled.” And you might tell me, “Ah, but once we feel like we know there's a
pothole there, we will feel overconfident that the problem will be solved, so
we won't do anything. And so there's no point in even telling people there's a
pothole on Main Street.” I mean, we could just make up a story about any
problem and the fact that telling someone about the problem will just make it
ineffective that we will ever solve the problem.
No, but I'm not making a general claim that it's – this is always true. I'm
making a claim that this worry is a worry for the kind of problem that you are
But I don't see why this problem is more susceptible to that than any other
problem. Why is this the kind of problem that knowing about it would have the
opposite effect of making the problem worse and all the other problems are not
I think that maybe it's because of how wrong you're saying we are. Like if you
say there's a pothole, that's just one pothole, the rest of the street is OK.
What if it's just nothing but pothole?
OK. What if we told people that all of Main Street was a pothole after – say
there was a big rain, an unusually large flood and in fact, Main Street is now
one big long pothole, do you really think telling people about that is going
to be counterproductive?
I mean, no, I don't think so about that case. Because I think well, OK, but
there's other places I can go, there's other places I can walk. And what I
wonder about with your book is, is there anywhere I can safely walk? Is there
anything that pothole left, including in my thinking about this very book,
including in my following the reasoning of the book. And so, when other people
like, you know, in the past when I read about self-deception and you know,
problems of reasoning and thinking. Like, if I can get to the point, there's
something destabilizing about knowing you did something wrong, but not knowing
exactly how you went wrong, like knowing I'm, you know, I ended up in the
wrong city, I'm in Detroit and I shouldn't be there or whatever, right?
There's something destabilizing about that, where I don't know what to do
next. And I don't know how to solve the problem. And the first thing I want to
do is make sure I can trust my mind that I'm not going crazy by tracing out
how this could have happened. In the absence of that, in fact, in the absence
of any hope of doing that, it's not that I think you know that you've done
something worse. It's just like, I don't, I don't think you know you haven't.
So, for comparison, my dear colleague, Bryan Caplan, wrote a book on The Case
I know I did an event with him about it.
And his whole book is focused on what is in effect, one of our chapters, the
one of 10 areas on education.
Now, you could be complaining that our book has too broad a coverage. That is,
since we show 10 different areas of life where we question your motives, then
you don't feel like you have any place you can stand. But what about Bryan's
book? Was that the right level of scope showing you just education and that
you were not very honest about your motives in education? Is that a small
enough thing such that you think you have another place to stand and now you
can improve yourself there? Is it the fact that we put 10 things in one book,
that was our problem instead of 10 separate books?
I think that I have… I'm not saying don't have any problem with Bryan's book.
We have different – we have different discussion of that. I think Bryan's
book, there's like the conceit that there's some special people, those people
actually care about learning. They're like how you think everybody is, for
instance, Bryan is one of those special people. And so education is wasted on
most of the riffraff because they don't have the pure motives, and they can't
learn or care about it. But there is this kind of retreat to like, there's the
possible person who could be well-intentioned and well-motivated and could
actually care, they're going to teach themselves so they're not going to need
school. But the point is that kind of retreat to there's like, a place of
sanity somewhere in there, even if it's not going to apply to most people,
does actually avoid this criticism. So I would say Bryan avoids this criticism
in a different way. It's implausible to me that there are these special pure
of heart people. But it does get you around this problem. I don't think your
book has that feature.
So maybe, then we failed to communicate our book well enough, but I would say
that any complicated area of life has many relevant motives. And that's true
about school and medicine and politics. And so we can't be talking about what
is the one motive in any one area, we have to at best be talking about what is
the most common strongest motive. And so we do say, part of your motive to be
involved in medicine is health. And part of your motive to go to school is to
learn the material. That is, you know, when you deceive people, like when a
child says that “the dog ate my homework”, they use an excuse like that,
because in fact, sometimes dogs do eat homework. The child doesn't say “the
dragon ate my homework”, nobody's going to believe that excuse. So in general,
when we make excuses for things we usually hide behind things that do happen
sometimes, they just aren't as common as we are pretending in any one case. So
that's what we mean by hidden motives. We don't mean that the stated motives
never apply and have no relevance. What we mean is we are pretending they're
more important than they actually are. And so, our coverage of medicine and
politics and conversation, et cetera is intended in that sense to allow for a
fraction of the time in people where their motives are, as they describe, just
to point out that they are over emphasizing those for a reason.
Right. And my – I guess, I think that the discrepancies that you're pointing
out, really are interesting and important. Like the, you know, if you really
thought conversation was an information exchange, why would people prefer to
speak rather than listen? And, you know, a thousand things like that. That
this thing that you tweeted about recently about why is it that people would
prefer not to do more medical research, and think about ways they could avoid
dying, rather than dying? Right? That is like, it seems like they, you know,
don't put as much effort into preventing themselves from dying, as you might
expect. And I guess I think those puzzles really are, it's maybe it's like I –
this is why I'm saying I feel like it's a missed opportunity. Those puzzles
indicate that we are in some way deeply confused about what we want. That is,
we think we want something that we want something quite different. But the
place where you land, in terms of what we actually want is like, not something
we could affirm. And so I find it like, that can…
That's not something that you want to affirm. That's different than it being
something you could affirm. That was the key distinction I was trying to make
at the beginning of the talk.
Right. So, I think that the, like, the problem with the, you know, want to
versus could affirm is that it needs to have the – in order to be able to
affirm it, I would… sorry, in order to be able to take the correction I would
You’d have to admit to yourself that you mainly care about showing that you
care, say in medicine and other people showing they care about you, that's a
much larger motive than you had realized. And you'd have to accept that.
No, but it's not just that, I think it's like the thing that will get me to
accept it is just this aha feeling of recognition. And nothing but that will
ever get me to really accept it. I can say that I accept it under the pressure
of your book so as to not look like a fool. But I don't think that that thing,
I don't believe that it's really accepting it. That's what I'm saying. I think
really accepting it, just like it – I even think if somebody told you, you're
in the wrong city, it would only be when you could trace out how you could
possibly have ended up in the wrong city that you would really accept it. And
so, I think you would only really accept this explanation if we could supply
you with a reorientation.
So we're almost out of time. But I started out in physics and there are areas
in physics where we can use the math to figure out what the answers are. But
we just find it hard to have intuitions about. Say, basic quantum mechanics,
so we have some intuitions about how billiard balls and a world of billiard
balls works. And then in quantum mechanics, we just have these math equations
that are quite accurate and powerful. But we can't have a vivid mental picture
of them so much, we have to sort of create a mental picture of the math. And
that's the best mental picture we can get. And that's what it's like to sort
of learn about something strange that your mind wasn't designed to intuit. And
that may be true of you as well. Your mind may not have been designed to
intuit and have a vivid mental picture of these things that you were designed
not to see. But like with quantum mechanics, if you trusted the abstraction
enough, if you trusted enough the abstract math that had told you that it's
really there, you could use that abstraction to figure out things. And in some
sense, that's what I'm asking you to do here. I'm giving you the abstract
reasons to believe these things. And I'm asking you to rely on that
abstraction to make choices, even if you can't vividly intuitively see it. And
maybe that's not possible for many people, just like many people just refuse
to deal with quantum mechanics.
I mean, so one possibility that you seem to be assuming is the case, is just
that the thing is, you're persistently assuming that the thing that I'm asking
for doesn't exist. And…
Or asking you to consider that possibility.
Sure. Right. And I guess I think that that is possible. And you know, that may
be the situation that we're in with quantum mechanics is that we have a bunch
of equations and we don't really understand anything by means of those
equations so that we can manipulate things and even maybe produce technology
and stuff by continuing to manipulate symbols. And that may be good enough for
some people. For me, I wouldn't feel like I knew anything as a result of that,
until I had an interpretation of it. But in any case, but maybe there just is
nothing. And I mean, that's possible. And so like, in a way, what I'm saying
is like, there's this much better thing that I can imagine that you could have
given us that you didn't give us in your book. And it's conceptually possible
that you could have given it to us. But it may be that actually, there is no
such thing there, that is there is no deeper explanation to be found. I am
sort of open to that possibility. I guess I'm just…
I'm open too it possibly existing too. But it seems to me, the way to find it,
if it does exist, is to practice using the abstraction, practice applying the
abstraction to concrete cases, and see if you can fill in a more vivid,
detailed picture of it that way. That the way to create an intuitive vivid
image of some alternative reality is to inhabit it as best you can. And
initially, that may be via abstractions. That, in fact, is the thesis of your
book I believe, about aspiration. You have a thing that you grasp abstractly
and not very vividly. And you struggle to find a more vivid internalization of
it. And what you have to do is repeatedly expose yourself to the abstract
versions, in search of ways to make it more concrete for yourself.
That's true. But another feature of my book is that you have to be kind of
relentless in focusing on the good of it, and seeing the good of it, and
insisting to yourself to see the good of it, even when you can't quite see the
good of it. You have to try harder to see the good of it. And so I feel like
what your book is…
You could do that here, too, yes, you could do that thing.
…giving up on trying to see the good of it. You're not trying to see the good
of these things. You're just like, let's – well look, turns out, we care about
the wrong stuff. So let's you know…
No, but remember, when I said, “How do I change my attitude toward medicine?”
I say, “I focus on showing that I care and let other people show that they
care about me.” Those are good things. They are different good thing.
So when you change your attitude towards education by like focusing more on
signaling and like…
Yes, yes. Focusing more on the value you get by being able to show what kind
of a person you are through education.
I don't think you’ve focused that much on that.
Seeing its value, to understanding its value, intuiting its value, and helping
to think about other ways you could achieve that. So for example, my simple
suggestion for how we can have a better kind of school is that school is
mostly about words on paper. And some kind of people are good in conversation.
And our schools at the moment don't show those people as better. They aren't
very good at distinguishing those people, because most of our teaching and
evaluation of students is on paper, or, you know, outside of the context of an
actual active back and forth conversation. So Socrates might not be rated very
well in modern schools, being good in conversation but not maybe so good at
writing an essay explaining something. So if we could create a kind of school,
wherein the focus was on conversational interaction, and pretending to learn
the material about such thing, but actually just doing it, and showing that
you're good at it, then some people who were in fact good at that would like
to go to that school. And they would find meaning in that school. And that
would be a new addition and improvement to school. I can see the value of that
by seeing that that is a valuable thing. I see that as valuable. And I think
it's a shame that people who are good at that can't emphasize that and then
show it to the world that they are good at that.
Yeah. But when I said you can't see it, I wasn't meaning you can't, when you
look over at someone else, see, that would be a good thing. What I'm saying is
that I've interacted with you in a bunch of intellectual contexts, including
watching you teach a class and you, you know, completely as far as I can tell,
avoid this thing of pretending to learn, showing your status, you just don't
do any of that stuff. You're completely failing in the means of education. You
don't do it in conversation. You don't try to do it, like you don’t try to do
it more and more. You just ignore all this stuff, which is like the main value
I don’t talk about it explicitly, but I think I am achieving, that is I
achieve it in myself and in my students. I help them show off through often
conversational interactions in class. I show them how to do that by example.
And I elicit that from them and then I tried to correct and adjust their
attempts in order to produce impressive, admirable inquiries.
I guess that's not what it seems to me. Maybe we just have to disagree on that
In any case, it's been nice talking.
All right. For now, goodbye.