Meta
Agnes:
Hi, Robin. Today, we've agreed to talk about a topic that is not a topic. And
so this conversation is definitely going to be terrible. It's probably going
to be our worst podcast that we've ever done. So listener, you should probably
turn this off right now. Our topic is meta.
Robin:
The fact that we post this should be some information that maybe we've changed
our mind, because in the worst case, we just won't post it.
Agnes:
That's right, I'm speaking from the point of view now. I'm making the
prediction in fact that we won't even post it. So if you're listening to this,
then you shouldn't be listening to me. We're already doing great, we're
talking about meta.
Robin:
Exactly, exactly. So as a context, I started a blog in 2006 and had two
candidate titles for my blog. The one that I actually used was called
Overcoming Bias. And the other one, I even had a whole separate image for it
was called more meta than thou as the candidate topic for my blog. So I was
thinking of sort of distinctive descriptions of how I think that might catch
people for a blog. And I thought one of the distinctive descriptions of me was
that I try to go meta a lot. And in fact, I was going to have this combative
version of like, haha, I'm more meta than you proved me wrong. That was more
meta than that.
Agnes:
Right. I mean, it would also say a distinctive thing about you is you're
trying to lord things over other people that that would be another thing
contained or that I enjoy that the fun of it.
Robin:
Right. So that question is, is that am I really trying to do that? Is that
just sort of a fun tease?
Agnes:
Right, well the problem with tease is that it's like it can come across in a
variety of ways, including not as a tease. Okay, and so you decided that
actually your brand was more that of overcoming bias?
Robin:
That's the one I chose, but maybe in part out of fears of people misreacting
to the
Agnes:
But is that your brand? How much bias have you overcome in your years blogging
and overcoming bias?
Robin:
Well, that's not the name of the blog. How much bias I've overcome.
Agnes:
Okay, but overcoming it suggests a project, did I miss?
Robin:
Or it suggests a sort of stance, a position. A stance. Yeah, it's an
aspiration.
Agnes:
You're on the other side. Everyone else is on the bias team and you're on the
antagonist.
Robin:
No, but you might be focused on the topic of how can I overcome bias and how
might that work and what would be the consequences? So that's at least an
orientation to the topic and I think Early on in the blog, certainly we talked
about that more often than most blogs would have. So it wasn't a false label.
Yes. It was thinking about how to overcome bias.
Agnes:
But do you think that thinking about it makes you overcome it? That is, have
you overcome it as a result of thinking about it?
Robin:
I might think that, on average, counterfactually, had I not focused on
overcoming bias, I would have less overcome it, and that would be the sign of
the effect, but the question is how big is the effect?
Agnes:
What biases do you think you would now have, had you not had the bias as a
particular bias?
Robin:
Well, it's not about cutting them out one by one. Presumably, it's more just
about reducing each one by a modest amount, rather than eliminating some from
the set, and it's very unlikely you would just eliminate biases. reduce them
to some degree.
Agnes:
I see. So we're using all of them by just like a tiny amount. And so.
Robin:
So, you know, out of my blog became the Rationalist Community because my
co-blogger Eliezer Youkowsky split off and created LessWrong. And that created
LessWrong is a name in some sense for trying to overcome bias. That was the
aspiration of that.
Agnes:
Suggest that he didn't think your blog was sufficient in the dimension of
overcoming bias.
Robin:
He had a concept for how a blog would be organized with allowing many people
to contribute and having a karma system for letting them vote on that, which
my blog didn't support. So it's not so much saying my blog was bad that he
wanted to make this new community, which he did. And it worked for a while. So
rationalism in general. has been motivated by this idea of, if you try harder
to be rational, you might succeed. So overcoming bias is an initial point in
that direction of saying, let's think about that. So the question is, do
rationalists succeed to some degree? I would say that just consciously sort of
enumerating how you could be more rational and maybe practicing a little isn't
really that effective. It's somewhat effective. It's better to change your
incentives or your context, which will more reward you for avoiding bias.
That's, I think, the stronger effect. But in order to do that, you would have
to think at the meta level. Let's see, we're going meta. If you think about
the meta level of rationality, you ask, well, what Would having a blog help me
be more rational? That's itself asking about rationality. That's the sort of
thing you could ask on that blog, right? Right. So that question, if it's
useful to think about, is then the thing that you would ask there. And again,
I would suggest that the most useful thing is to change your incentives.
Agnes:
Right, so, I mean, in a way it is a question about meta. It's like, how useful
is it to go meta? To say the best thing is to change your incentive, or, I
mean, the best thing might even be for somebody else to change your
incentives. So you're not even, just to have different incentives.
Robin:
Right. So then you write about it and you inspire somebody else to do it by
writing and having them read it.
Agnes:
But they've got to change your incentives.
Robin:
Right. Still. But still, it could still be a benefit of writing about it. Not
so much that you read it yourself as that somebody else reads it, takes
action.
Agnes:
Right. But like how, I mean to try to, I wanna try to stay on the topic of
meta because the challenge here will be whether this can be a topic at all.
How productive is it to reflect on the very thing that you're doing?
Robin:
Says a philosopher.
Agnes:
Yes, only a philosopher.
Robin:
Isn't that your main line of business? Is that what you, is that the main
brand? You know, if philosophers are nothing else but the brand of we reflect
on things that other people do.
Agnes:
I don't think as a philosopher you're usually reflecting on the very thing
that you're doing at that moment, which is what the meta would be, right?
Robin:
No, not true.
Agnes:
I mean, so I guess there's just abstraction.
Robin:
Yes, that's meta. Yes.
Agnes:
Okay. Maybe we should just actually, maybe we should start with defining what
meta is. Cause I think we're, we're, we're, we have. Okay.
Robin:
Well, for any topic X meta X is a more abstract discussion about X. And so you
can go meta, meta X. You could have a discussion about as we're doing now,
what happens when you go better on X. So. It's a kind of abstraction. It's not
the only kind of abstraction. There are many other kinds of abstraction, I
suppose.
Agnes:
Yeah, I guess that's it, is that there are many kinds of abstraction that
don't strike me as meta.
Robin:
If people in the world are loving and then they start to talk about love, then
their discussion about love is meta to the love they have. If philosophers
then have a discussion among philosophers about what happens when ordinary
people talk about love, that's meta to a next level. And you have done that
sort of thing in the past. You have thought about love itself or even about
what happens when people talk about it. And again, I think, I mean, certainly
one level of meta you wouldn't have any objection to. I think more the
question is, how fast does it run out? Meta.
Agnes:
I'm not sure I have any objection to any of it. I mean, I'm not sure it's a
thing. So say that we are talking about like politics on Twitter or something
and how people talk about politics on Twitter. I guess that's meta.
Robin:
You might do it on Twitter.
Agnes:
Well, right, so, but what I'm saying is, from your point of view, the fact
that we're talking about people talking about something, it like, Is that what
makes it meta? Like, it can't just be that it uses concepts or something,
because that's just going to be everything. And I don't know, if I say love is
blind or something, that doesn't seem very meta to me. Yeah, I'm talking about
love. I've used the concept, but I feel like The concept of meta is somewhere
in the family of the concept of recursive or something like that, where
there's an element of self-reference or, I don't know, something like that to
it, that it's not just being abstract.
Robin:
Okay, so I just looked up the word meta and Google Dictionaries does say...
Started with that. Of a creative work referring to itself. or the conventions
of its genre, self-referential. So now that referred, that seems to suggest
it's not just an abstract discussion of X, but maybe it's a discussion that
includes a discussion of itself.
Agnes:
Okay. That's a win for me.
Robin:
Okay.
Agnes:
Mark it down. So yeah, don't worry. It's smart.
Robin:
So... So then you might ask, when is it ever useful to reflect on yourself?
Agnes:
That's what I meant to be asking. Yes.
Robin:
Okay. But certainly that's a thing philosophers do a fair bit of, right?
Agnes:
Philosophers not only, I mean, they do it, they also describe it a lot. They
think people are doing it all the time. In fact, there's a thing, and I've
sort of criticized it, I call it the stepping back move, where it's like,
okay, you want to do something, but wait, step back. Step back and evaluate
whether you really want to do this. And there's sort of this idea that there's
somewhere there, there's some place there that you could be stepping back to,
where then you're stepping into the shoes of some other person who's not
really you, but who's objective or impartial and who's going to assess the you
that you've stepped back from. But it seems like you actually are just the
same person when you've stepped back. And so you could pause, you could spend
more time thinking about a decision, but we're inclined not to say those
things, but to say you can somehow step back. And to adopt this meta point of
view, where you're sort of looking at yourself and assessing yourself from the
point of view of not yourself. And that I guess I am skeptical of that there
is any such a thing.
Robin:
Seems like you're jumping to the conclusion that that would be the only or the
main reason why one might go meta.
Agnes:
No, I'm just giving one example.
Robin:
but you'd be skeptical of one reason. I mean, I might say, I'm skeptical of
going meta when you're hungry. I don't know that meta is particularly useful
for people who are hungry, but that doesn't mean it's not useful for
something. It just means there's a vast number of purposes you might have in
the world, and this particular one thing isn't useful for most of them.
Agnes:
Might be useful when you're hungry. If you're hungry and you can't get any
food, And what are you gonna do? You can sort of reflect on the fact that
like, well, you're not as hungry as you could possibly be and, right?
Robin:
Right, that seems counterproductive to me. So the usual fix is to try to
ignore it, to try to like set it out of your mind so that you can distract
yourself.
Agnes:
Maybe you do that by going meta. I mean, for some people, that is a way to
distract themselves at least from a certain feeling by occupying their mind
with the thought of it.
Robin:
Well, I mean, we can have a concrete example here. We're here doing a podcast.
Agnes:
Yeah.
Robin:
And we could talk about the fact that we're doing a podcast.
Agnes:
Oh, yeah. We actually probably should do that to be consistent with our theme.
But OK. Yeah.
Robin:
So we could talk about sort of what tends to happen in podcasts and how our
podcast might be different than other people's podcasts. That's a way of being
meta. Okay, let's try it. It's not for the purview of stepping outside
ourselves. You don't have to want to step outside of yourselves in order to
have that discussion. Right. So just, you might be perfectly skeptical that
it's possible to step outside yourself, but it certainly could be possible to
compare yourself to other things. So we could say, how is our podcast
different from other podcasts?
Agnes:
Right. That would be it. OK, so how's it going now? Here's the fundamental
problem with answering that question. Neither of us listens to other podcasts.
Robin:
We've been on other podcasts, so I can certainly compare this to other
podcasts I've been on.
Agnes:
OK, so how does our podcast compare to other podcasts you've been on?
Robin:
Well, the typical podcast has an interviewer who has dutifully compiled a list
of questions and they go through them one by one, ticking off the question and
asking it, politely listening for an answer for a while. And when you've done
pausing in an answer, they go to the next question and they read it. I'm not
very happy with this format, which is why I'm glad that we have a different
format here.
Agnes:
But in our format, there's still somebody who asked the questions.
Robin:
Well, either of us can ask questions at any time.
Agnes:
I think I ask more questions. I don't think you ask many questions.
Robin:
Okay, well, then we can compare.
Agnes:
So I'm still the, I'm the interviewer. I'm just a very weird kind of
interviewer that doesn't prepare my questions in advance.
Robin:
Okay, would you think in this particular podcast, Going Meta, that you've
asked more questions? See, another example of going meta.
Agnes:
And you asked a question. And you're just trying to refute my inference of
time. Well, maybe when we go meta, everything goes backwards or something.
Robin:
It's backwards day, so we should have a better day instead of backwards day.
Agnes:
You know how there's the- I think that on this podcast, I maybe only asked one
question until now, so we're even on this one, but I still have time. Okay.
Okay, so it's not the case that one person just asks questions, but I've
actually been on podcasts where people banter, Instead of asking questions.
Robin:
Trying to show a conviviality and friendship and fun relationship sort of
thing.
Agnes:
Right, so like on Night Owls, I ask questions and sometimes my guests are
like, well, couldn't we just have a conversation instead? And I think it's not
so easy to do that.
Robin:
Then congrats to us, because we're having conversations.
Agnes:
Well, let's let our listeners be the judge of that. Maybe we're bantering.
Robin:
Is there, what's the difference?
Agnes:
Oh, question, good job. I think that when you're bantering, you laugh more. So
we've been laughing a lot in this conversation, that's a bad sign.
Robin:
Is it a bad, only if bantering is bad.
Agnes:
I'm saying it's a bad sign for the idea that we're having a conversation as
opposed to bantering, right? So bantering, people laugh, they smile a lot, and
a lot of the conversational effort goes to getting along.
Robin:
People can't really accuse us of that, I think.
Agnes:
I don't know what they want to accuse us of.
Robin:
We're not trying that hard to get along.
Agnes:
I think that that's probably true.
Robin:
So good for us.
Agnes:
Because I think you're incapable of trying to do that, and I'm just actively
trying not to get along with you. And you're just indifferent to the thing I'm
doing. And so the effect is there's not much going along happening. So I think
that that's right.
Robin:
And yet we're still smiling and laughing.
Agnes:
Yes, that's because it's funny that we're not getting along. Yeah, that's
true. So it may actually be that at a meta level, we in some sense are getting
along. We are meta-factoring. We're both pleased by the fact that we don't
have to get along. And thus, in some strange way, we are getting along.
Robin:
And we are commenting on that fact in our podcast.
Agnes:
Therefore, we are being meta right here. In the meta podcast. It's only
permitted today. Today's the one day. We should have done this on April
Fool's. We could backdate it, maybe.
Robin:
Nobody knows when these things happened.
Agnes:
I think that in a bantering, you're just trying to pass the time until the
time is up and you're just like, okay, only another 10 minutes bantering to
fill in. But I think that in a conversation, there's something you're trying
to figure out and your time gets cut off. You're like, oh, we didn't, you
know, we didn't get there. So it's like, we might get to this end of this
conversation and we still don't know what meta is. Well, then we're screwed.
That's a problem. As opposed to we get to the end of this conversation and
like, oh, we managed to be smiling and sort of happy for the whole
conversation. So, yeah, I think we're having a conversation insofar as there's
something we're trying to figure out. I feel unsure in this particular case
whether there is anything we're trying to figure out, because I'm just not
sure that meta is a topic.
Robin:
Well, let me jump in and suggest a thing we try to figure out.
Agnes:
Okay.
Robin:
I think we've established what meta is. We agree on that.
Agnes:
You agree that it's my thing where it involves self-reference? Yes. Okay,
great.
Robin:
Okay. We agree on what meta is. We agree that we sometimes do meta. Yeah. I
think the simplest, most interesting question about meta might be, do people
do it too much or too little? Or when and where do they do too much or too
little self-reference? Or which sorts of people do it too much and too little?
Agnes:
Maybe narcissists do it extra. I'm a bit narcissistic, and I do it a lot.
Okay. Because narcissists are just inclined to think about themselves, and so
then they're going to be inclined to think about whatever they're doing. But
that's not a question of who does it. It's not a question of whether we do too
much, but maybe a certain group of people will do it more.
Robin:
So I think intellectuals are more capable of being meta, at least more
complicated ways. So they might be inclined to show off their abilities by
going meta. That might be part of the fun. It also might be if you're trying
to sort of make it difficult. So like if you have enough double negatives or
triple negatives off it, it's hard to like figure out what you're saying. And
I think there's a sense of which it's a bit hard to figure out exactly what
we're saying when we're talking meta.
Agnes:
I think there's another thing where meta is an escape. Say we're having an
uncomfortable conversation about one of our fundamental disagreements and
we're getting really annoyed at each other, we could be like, oh, that's an
interesting meta question that we're even having this disagreement. And then
all of a sudden we stepped out of the disagreement and we're looking at it.
Robin:
But in some sense that's true for all topics. It could always be used as a
distraction for some other topic, right?
Agnes:
Right. But in the meta case, it doesn't feel as though you're just changing
topic.
Robin:
It feels as though you're... So meta seems adjacent to every topic, maybe.
Agnes:
Yeah. Yeah, exactly.
Robin:
That makes it special, right?
Agnes:
It's a safe space you can move to from any topic where it doesn't feel like
you're changing the topic, but in fact, you are changing the topic.
Robin:
Now, the opposite of meta, I think, would be maybe calling it for an example.
OK, and that's also a way of distracting for having a difficult discussion. I
could say, well, let's think of an example. But is it really such a terrible
distraction to think of examples?
Agnes:
No, but I. I guess I'm less inclined to think that. Thinking of an example
serves as a sort of safety valve in the same way. Why not? Maybe because often
when a discussion is heated, it's heated precisely because you already are
thinking of some example and what you want to be doing is getting out of the
example.
Robin:
Maybe you could pick a different example to escape.
Agnes:
Yeah. Yeah, that's true. I think that it often will feel like the topic is
being changed when you pick a non But I don't know. No, I don't feel confident
about what I want to say about that one way or the other.
Robin:
Okay. So it's one of the many moves always available in a conversation. We
could think of it that way. That is, there's a range of conversational moves,
and some of them are only available in limited context. And the meta one, in
some sense, is always there, always ready.
Agnes:
Yeah. And, okay, I also think that There's a way in which meta does provide an
opportunity for something like impartiality or I wish I could think of an
example of this, but... I agree. Think about the very fact that you're saying
this.
Robin:
I think it's going to call to attention a class of things, which then might
call to attention other examples of the class, sort of forcing you to compare
your current situation to other situations that are similar. And that's, in
some sense, coming out of that situation.
Agnes:
That's true, but that's not what I was thinking of. I was thinking of
performative contradictions. So I was thinking of, like, question everything.
People sometimes say question everything, right? And you might say, did you
notice that when you said question everything, you were not asking a question?
You were making command. Right. So question everything, in a way, is a
performative contradiction. Because the act, a speech act, is an imperative,
but the content of the speech act is don't issue imperatives. And so I think
that going meta is often a way to expose performative contradictions.
Robin:
Of course, if it had said, I'm going meta, or sorry, you know, I'm questioning
everything, that- That's still not a question. It would become more of a
direct contradiction. Yeah, exactly. I'm questioning everything is a
contradiction, because at that moment, I'm not questioning everything.
Agnes:
That's true, right, right. And so that's in like the family of the liars
paradox.
Robin:
Or rather, I mean- Right. But it still seems to me the more interesting
question here is, where is going meta more useful? And when do people perhaps
use it when they shouldn't? When is it less useful?
Agnes:
Well, I'm saying I think it's useful for exposing performative contradictions.
I think that's one of the main ways that I use it is I'm saying, notice the
exact thing that you're saying. The thing you're saying is a counter example
to the content of what you're saying. The act of your saying it is in tension
with the content of what you're saying. That's a way that people routinely
contradict themselves. I do think that's a use of going meta.
Robin:
And it's different than other pointing out of the contradictions, because to
point out mere contradictions, you just look at the content of the claims. But
here you have to look at the context of the claims. And to look at the context
of the claims, you have to talk about our context that we're in.
Agnes:
Right. Right. Or, you know, even just like, I don't know, when people complain
about stuff like, oh, can't, let's say, can't people be like kinder and more
understanding? It's like, well, you're not being very understanding about how
people aren't understanding. You kind of can't be in that. You have to be
incomprehending about people's lack of understanding in order to complain
about it. So maybe it's also suggests that there are certain attitudes that
it's not just an accident that you get into some kind of contradiction.
There's a problem with expressing, and there's a problem with expressing the
thought, question everything, or everyone should be more understanding. It may
be that there's some very sophisticated way in which you could express that
thought, but all the obvious ways run into performative contradiction.
Robin:
So when people think about artificial intelligences and whether they're really
intelligent or about other things, I think they often are drawn to meta
discussions. Because in some sense, an ideal agent is able to talk about
itself as an ideality agent and what it's doing at the moment. So if you just
see an LLM talking, but if you ask it about itself and it doesn't seem to know
how to talk about itself, that seems a a bad sign. It's, you know, more just a
movie playing or something. And then, but there are all these difficulties
being very good at talking about yourself because you can easily make mistakes
where your descriptions, your meta descriptions don't actually work. So in
some sense, now it becomes a test of intelligent agency, how much can you
describe exactly what you're doing? And I guess there's a sense in which often
an ideal of rhetoric or of styles of writing or speaking is that the speaker
is able to concisely and directly describe their context of speaking in a way
that would set themselves up and help you attend to them, showing, you know,
I'm about to tell you this. I'm here to tell you this. The reason I'm here is
to tell you the following, and I will tell you only so long as you do the
following. I think we gain respect for a speaker to some degree, and it's more
commanding and more engaging if a speaker can Go meta sufficiently to
correctly or even aspirationally describe how it is that they're speaking to
you and why.
Agnes:
So I also think as a matter of style that certain kind of stylistic, I don't
know, flourishes or something, being wry or being self-aware, they're
considered virtues. being able to channel how you think the reader is
receiving you in a kind of tongue-in-cheek way or something.
Robin:
Exactly.
Agnes:
A form of going meta. And I guess a lot of humor actually involves going meta
in some way. Like comedians often in some way surprisingly invoke the very
context of speech. at a moment where you didn't expect it or something.
Robin:
Right. So I think there's a lot of engagement you get and respect and even a
sense of connection when a speaker or writer goes meta in a way that shows you
that they get you and that they get themselves and that they know how you
might react to them, how you might be skeptical, how you, et cetera. I think
that's sort of a big way we connect to each other. is by not just talking to
each other, saying things to each other, but talking in a way that shows the
other one that we realize how they're reacting, how they might think of
reacting.
Agnes:
And it's more than- It's because then I'm, like say I'm very self-aware about
how you're receiving me and I'm reflecting that back to you. And especially if
I'm on a stage or something and you can't talk, then I've like given you a
place in the conversation, right? I'm like, I'm not ignoring you because your
point of view has been included in my point of view. I'm somehow channeling
it. And so there's somehow something very, it's that in being aware of how
you're being received, what you're really being aware of is the point of view
of your interlocutor. And you've sort of incorporated that somehow.
Robin:
And of course, by making a statement on it, you're taking a risk that you
would be wrong. So the fact that you weren't wrong is showing a degree of
confidence and understanding. It's a way of signaling.
Agnes:
But there's a way in which being meta is being social, that is.
Robin:
Yes.
Agnes:
Because the attitude, the meta attitude that you're projecting is the one that
you anticipate your audience to have towards you.
Robin:
And it's a way of being social that doesn't require back and forth
conversation. So usually with two people talking, you can be sufficiently
social just by noticing what they said and having what you say depend clearly
on the fact of what they said. But if you're speaking to a wider audience, you
don't have that available as a way to be social. So being meta is one of your
only remaining options.
Agnes:
Right. Right. I mean, it's interesting because philosophers are not very
humorous or wry or self-aware or possessed of irony.
Robin:
They do often go meta in the sense of, OK, I know you're thinking of the
following objection.
Agnes:
That's true. Yeah, so maybe there's two different ways of going meta. One of
them is just going to be very explicit and it's not going to have much
affective content. And the other is going to be like with the kind of emotions
of the other person, that you're channeling the emotions of the other person
and not just their ideas. And so maybe philosophers do go meta in the
channeling the ideas of the person who's reading this. You probably have this
objection. But we don't go mad and like, you probably think this is absurd,
what I've just said, or you think I'm ridiculous. The ability to laugh at
yourself, I don't think is a kind of characteristically philosophical trait.
Robin:
So there's a common observation about ordinary conversation, which is that we
all know what the most important topics we could talk about are, and we don't
talk about those. We tend to talk about things a bit farther away from our
most important topics. And then, in some sense, we then more often connect
through these meta asides, that is, instead of directly saying, I love you and
I'm sure glad you're in my life in a literal, direct way, we prefer, I think,
to say things like that through these meta signs that I'm listening to you,
that I appreciate that you're here, that I know that you have other options or
things like that. And so there's a way in which, well, we do kind of say
sometimes the most important things, but not so directly. And these meta signs
are some of the ways that we allow ourselves to say some of the more important
things.
Agnes:
it seems to me talking about the more important topics and saying the more
important things are two very different, so like, at least as I was thinking
what the most important topics are, you know, I was like, okay, God, war, and
then you're like, saying the most important things, like how much we care
about each other, it's like, okay, that's a very different conversation than I
was imagining when I was imagining the conversation about God and war. And so
it's more like, you know, we don't directly, or often don't directly express
affection, because who would believe that or something? You have to like
express it through signals that you can't fully control, because then it seems
like it's coming from the deeper part of you. But that seems like a different
issue.
Robin:
So I long ago read this book called Clear and Simple as the Truth. It's a book
on rhetorical style, and it occurs to me that it's quite relevant here.
Rhetorical styles, in some sense, have to take stances on these meta issues.
Rhetorical style basically sets up defaults about meta claims about the
conversation. And Clear and Simple of the Truth says that a favorite style in
our world in the last few centuries is the style which basically presents the
meta-framing of, I'm just looking at something and I'm just telling you
exactly what I see. And there's no reason for you to doubt this. You might
just be informed by this and there we go. So this stance doesn't acknowledge
my uncertainty. It doesn't acknowledge that I might have had choices about
what to tell you about or where to look even. It doesn't acknowledge counter
arguments that might be available. This style sort of pushes those aside
because its stance is so directly compelling. And we tend to prefer it, but
the idea is that a writing style is itself a collection of claims about meta
of the conversation. And that's a way of understanding style. It's, it's a
package of metas. So instead of having to iterate through a bunch of specific
meta claims in order to get on board with UI and Vogue with style, and you can
immediately fill in all the rest of the meta expectations, we can get on with
our discussion.
Agnes:
Right. I was thinking about, um. I mean, like, I think just this style of
being very direct or something, if I think, oh, is that how Robin, like, is,
and then be like, yes, you're pretty much very direct and you're pretty much
always just explaining stuff, but actually that hides a variety under it.
Like, you're different, like, if we were just, if we were to end this
recording right now and then we were just to talk on Zoom, you have a
different manner. You're much more expository in a podcast, like you have a
Mr. Professor tone a little bit. Okay. And, um, And so that's, I guess that's
part of the style. And so you have a more didactic, I don't know, maybe I
differ too between podcasts and the regular conversation. I don't notice
myself differing, but of course I probably wouldn't. But it seems to me that
you can have a more didactic style would maybe make sense for a podcast
because you have a self-understanding that you're explaining stuff, not just
to me, but to a big variety of people, all of whom differ from one another, so
you have to make some average explanation that's gonna work. Whereas if it was
just me you were talking to, you'd just tailor the explanation to me. So when
you're just looking at something and you're just explaining it to someone,
still there's a big variety of ways you could do that.
Robin:
You're right. So yeah, so I think we professors have a lecture style, and it's
related to a writing style for a wider audience. That is, if I'm writing a
private email to someone, I might have a different style than writing an
essay. And there is the meta context of an essay or a lecture is, I'm speaking
to many people at once. And I can't address each of their idiosyncrasies very
well. You all have to understand that. So I'm going to maybe go over some
things. Some of you might not rather we went over and explain terms a little
more clearly than I otherwise would need to. And maybe anticipate responses
because you know, won't be able to respond and I'll have to, you know,
substitute for that by explaining more.
Agnes:
Right. And it could actually be that. Um, there's a kind of meta trade-off.
Um, so you could be someone like me who reflects on things being meta
relatively often. But, um, if you ask me, how many different styles do I have?
How many different modes of self-presentation do I have? People are often
very, very surprised to learn by experience, not just by me saying it, that
the answer is basically what? That is, when I'm lecturing in front of a group
of students, when I'm talking to you, when I'm talking to my children, when
I'm on a podcast, I kind of have the same mode of speaking. I don't really
have the ability to adopt different modes of speaking. And so in particular, I
don't have the ability to sort of sound more sophisticated or professorial, to
get rid of the word like from my vocabulary, people have complained about
that. I'm sure some of our listeners right now are thinking to themselves,
yes, it's very annoying. And so that may be just, I'm less meta in one sense
of meta, in the sense of not being really able to incorporate a
self-understanding of the peculiarities of the context into my
self-presentation, even if at a kind of conscious reflective level, I will be
very happy to go meta in terms of thinking about the very fact that I'm doing
that or not, as I am right now.
Robin:
Well, the more different styles you invest in, the less good you're going to
be at any one of them. So it might be that you're really good at a particular
style.
Agnes:
It's often the best style. Right. It involves saying, like a lot. Well, and
then because you- I got hooked onto the teenager style at the age of 17, and I
just never let go.
Robin:
No, but it's more having, if you have a very distinctive persona that you
project, it could be so vivid and compelling that it sort of fills in a lot of
details very quickly and automatically, and then you don't mind the fact that
it didn't vary with context.
Agnes:
Okay, I see that that's right.
Robin:
That is, you could be less trying to show that you get the context and that
you are aware of the context, but more just trying to show the listeners,
well, this is my metacontext and this is how the brand, my brand is very
stable. And distinctive and it's conveyed very, you know, vividly.
Agnes:
Right, though, interestingly enough, well before anyone knew me in any way, I
always had this brand. It was always the same in a variety of different
contexts. It could be that I developed the brand partly out of the incapacity
to go meta in this particular way.
Robin:
One of the parts of this sort of meta brand of a style that would seem to be
the least realistic is that Surely most people when they're talking are a bit
anxious about how they look and how they're coming across and whether people
like them and agree with them. And that tends to be a bad look in a style, to
be very vividly anxious and supplicant and flattering or whatever, especially
if it makes you look like you're very anxious and unsure about your stance or
relationship. So I think our meta-presentation is often going to suppress
those sorts of things. A favorable stance will be, I like you, I'm glad I'm
talking to you, and I'm not very afraid of you rejecting me, and I'm not
really very thinking much about whether you're criticizing me or disliking me.
I think that may in fact be a very common thing going on in our subconscious,
but we learn to suppress that
Agnes:
We learned to... Or at least we learned to suppress the visual manifestations.
Robin:
Right, exactly. So that's a meta context that we are avoiding. Right. So just
like clear and simple, the truth, you know, the critique was, well, it's
suppressing the fact about where the information came from and your
uncertainty about it and the different paths you could have taken to get there
and maybe the possible critiques. You know, here are quite often our ordinary
presentation style is suppressing the fact that we are or could be anxious and
insecure about our relationship to our audience.
Agnes:
Actually, I think there's a lot of different things about the social situation
that we're suppressing and not supposed to bring to our consciousness. I've
noticed this because I sometimes give talks in which I'll just use as an
example, okay, notice that everyone in this room is facing me. Right. Nobody
is talking to their neighbor and nobody is wearing a giant pink hat and nobody
like jumps up and starts doing a dance. Right. You're all just really well
behaved. And in fact, you all have the same expression on your face of just
kind of I'm a serious, attentive listener, like none of you have a grimace on
your face. And so you're you know, this this this giant performance of sitting
in the classroom has been coordinated across, you know, maybe in a lecture
room. and you're following a lot of rules about how to behave in this context.
And I'll say this to the students and they'll like, usually not to like my old
class of students, but it's like if I'm giving a lecture and some new group of
students who don't know me, and they'll be like a bit shocked that I'm
bringing that up. Like there's a look at their face, most violation and we're
not supposed to talk about this. So that's a kind of meta where, It's not that
they're insecure or something, it's just that they're not comfortable with
bringing the certain... The management of it. Yeah, the management of it to
the fore of their consciousness, to attend to it.
Robin:
Right, and so this is in some sense a way in which meta is especially potent.
So as you know, I have this book, The Elephant in the Brain, Hidden Motives in
Everyday Life, And you might think typically we find it hard to notice our
hidden motives, to notice things that we're hiding. But then I could have
these predictable formulas for taking some features of an environment and
going meta on that where it quite reliably will reveal hidden motives or
things that people don't want to talk about. So in that sense, meta is dense
with these sorts of things people aren't comfortable with.
Agnes:
Right. But it's almost like there's a There are these two aspects to it,
right? So what I'm doing when I point out to the students, notice that none of
you is faced in any other direction but towards me. That's not. Um, that's,
that's a kind of meta, right? Um, and then there's the fact that they're
facing me, which is kind of meta too, right? Their understanding of the
situation. And, um, it's almost like the, the, the second one is an issue of
style and style is like the meta that has become second nature or something.
that where they don't have to think about it. And part of what you're doing,
if you're being explicit about it, you're puncturing that by you're being
like, oh, we can just talk about all this. We can just bring it out into the
open. It doesn't need to be a thoughtless habit. It can be like something we
reflect on. And so it's almost like there are two points of view on meta.
There's the analytical point of view, a very conscious, very explicit from a
distance. And then there's the reflexive inhabiting of self-awareness that is
often involves a certain kind of ease, a certain kind of grace, or a certain
kind of irony, a certain kind of self-awareness, where that's, it's very much
not just like a set of conscious reflections about your situation. That's a
distinction in meta. And I'm not sure how to hash out that distinction, but
it's like there's style meta and analytical conscious meta or something like
that.
Robin:
Okay. So, I mean, I guess just one way to say it is that, uh, we often manage
context through these packages of expectations because it's a very high
dimensional situation. Most situations are pretty high dimensional and we
reduce their dimensionality by have a limited number of styles. And so if you
can invoke a style quickly, you can very quickly reduce this high dimensional
set of possibilities down to very small number, maybe even one possibility
that we all accept and agree on. And, you know, we go on. That's what a style
does.
Agnes:
Yes, but also, I mean, I feel like it reduces it down to a few and then it
reduces it even further down to zero because you don't have to think about it.
It becomes not something that you need to consciously reflect on at all. And
in fact, the thing that you're doing might be very, very complicated. You
might be managing, at one and the same time, like 10 different fronts in terms
of how you hold your head, and facial expression, and how long you take a
breath, and your attention. And then if it's gonna involve speech, it's gonna
be very, very complicated. But none of it is managed by thinking about, okay,
I'm in this situation, let me arrange my face in this shape or something. It's
not managed consciously.
Robin:
I mean, I've been thinking a lot about culture lately and about how much
cultural variety is possible, and this is a concrete example showing just how
easy it is to forget. So think about a play, for example, you sit down and
play. And at the beginning of the play, they're going to give you some hints
that put the play into some genre in your mind, set up some expectations, and
then you can quickly, you know, guess what the characters are and what they're
like and what they're about to do. And it's a comfortable world, even though
at the moment, you know, the play started, you might have had no idea. what
possible story this could possibly be. So there's this vast space of possible
stories, and yet quickly they're going to limit your expectations in a way
that lets you, you know, not be lost in the world of the play. But that's
hiding from you in some sense, the vast space of possible plays that could
have shown up that you would have not been able to follow and be completely
lost in. The fact that they invoked one of these standard styles is a way that
they then, you know, not only vastly reduced the variety possible, but made
you very comfortable with it and almost seemed to be obvious. These are ways
in which we forget how much variety is really possible. The first few notes of
a piece of music will pull it into a genre and a style of music from which you
can make pretty smart predictions.
Agnes:
Almost always, before you started listening to it, you decided to listen to
it. And so you knew even before the first few notes or even before the play
started. Right. If I go to a theater to see The Magic Flute, I know there's
going to be singing. And I don't just open my eyes to the set and be like, oh,
OK, I guess we're doing an opera. I knew we were doing an opera because I
Robin:
Even if you came to a concert with no description of what the music was going
to be like, and all of a sudden they start to play within a couple of notes,
your uncertainty will be gone. You will pretty much know what kind of piece of
music this is, and you'll know how to relate to it.
Agnes:
I guess like some kinds of avant-garde music were trying to make that not be
for the listeners by doing very surprising things like suddenly you'd have a
whole bunch of silence or something.
Robin:
Presumably some kinds of authors try to make stories that surprise, but
usually it doesn't go over well because people don't like it.
Agnes:
My son once made, he challenged himself to make a song that never repeated,
like it didn't, it had no repetitions in it. And it was terrible. And it kind
of felt like this long drawn breath that never got anywhere, which was like, I
don't know, five years ago. But yeah, I guess music especially relies on, I
mean, not just on telling you what genre it's in, but also on containing in
itself repetitions, motifs.
Robin:
Right. Quickly hinted at. And even lectures, of course, and essays.
Agnes:
Right. It'll be like. I discovered to my chagrin that I have like a writing
tick where there's things that always need to come in threes. Adjectives need
to come in threes. Sentences need to have three parts. The essay needs to have
three parts. My book is like three sections, each of which breaks into three
parts. And then I added some parts to disguise about myself. I feel like
things just need to come in threes. apparently, though I'm not aware that I
have this, but that's a stylistic feature or meta level or something where
I'm... I wonder if this is a thing that evokes, you know, fear or anxiety when
meta is brought up.
Robin:
If someone starts to explicitly try to talk meta, you might have the fear
they're going to break out of the style patterns and put you into a place you
don't know how to handle.
Agnes:
I kinda had that fear when we started this podcast. I was like, I don't know
what's gonna happen in this podcast. This could be anything. I have no set of
expectations about what, you know, like I have some idea about what our
podcasts are, but suddenly when we're talking about meta, I no longer know
what it is to do a podcast.
Robin:
Right, so in any conversation, if you stopped and said, do you think this
conversation's going well? It would be very awkward. Or any sort of very
simple question about that. You think we've talked enough? Should we stop now,
even?
Agnes:
Yeah. Okay, good. That's another thing we might talk about. But I feel like we
have 10 more minutes in this podcast. And I think in these 10 minutes, we
should try to figure out whether or not, it's a good meta question, is this
podcast good enough to post? Or we should at least address that question.
Robin:
I already think it is. What's the question? It's good. This is good. Post it.
Agnes:
Okay, fine. At least we addressed the meta question inside the podcast itself.
Robin:
Okay, but should we address the meta meta question of whether that was a
useful question? Feel free. It was a useful question.
Agnes:
Okay, now I think we should go back to the thing you were just saying about
how in conversations we don't say things like, should we stop? You and I both,
someone on Twitter adverted our attention to a study that says that basically
everyone stays in conversations longer than they want to. That is,
participants say afterwards, yeah, I would have done it shorter, but I thought
the other person wanted it, and everyone does that. And so we don't explicitly
manage conversations, and we don't explicitly ask ourselves, How was this
conversation going? What could we be doing to improve it, right? We don't do
any of that. So there are certain forms of meta that we don't go in for. Yeah,
why not?
Robin:
Well, the theory you just pointed to is the idea that we're comfortable with
styles. which answer a whole bunch of meta questions all at once as a package.
And as soon as you start to explicitly talk meta, you're breaking out of the
style.
Agnes:
I see, there's no style. But if we were to just create a style, that's the
meta style.
Robin:
But then you might think it might be a hard style to create because it's
referring, you see, to the other style, right? If it were a whole
self-contained style, we could just go to that separate style. But you see,
this style is talking about the other conversation and threatening to change
or question its style.
Agnes:
I think that that's right. So I think actually it is a bit like the stepping
back moves that I was making fun of earlier. There's something to that thought
that if I say to you, okay, is this podcast good enough to post or how's this
going or should we stop having our conversation? that that conversation, the
conversation where we're trying to decide, you know, how to manage our
conversation, we don't know how to have that conversation. It's like the Wild
West of conversation.
Robin:
Right.
Agnes:
We don't have, which it's hilarious, actually, that it's so easy to enter the
Wild West, like that our ability to talk is so thin that it's just contingent
on, you know, we have this thing that we're doing, we just keep doing it. And
so it's just, we may be, because I've long thought something like, oh, well,
it sends a bad signal or something. If I say, I don't know, this
conversation's kind of boring. Do you want to stop or do you want to keep
going? Or even if I say to you, a minute ago I said, shouldn't we post this
conversation? That could be a bad signal. You could be like, oh, Agnes was
doubting whether this is good, right? So that's one possible explanation of
why we don't go meta. But now we come up with a different one, which I like
better, which is just, we're really uncomfortable with the meta conversation
Um, because we, there's just, we just don't have a style in which to conduct
it.
Robin:
Now that sometimes there are like set styles for meta conversations, and it's
interesting to notice how they become perfunctory and kind of boring. You
know, for example, at the beginning of most every podcast conversation I've
had, they say, well, like, they have the thing, like, you can stop at any time
because we're going to record this. And then they say, you know, if there's
anything you didn't want to have said, we can cut it out. And how are you
today?
Agnes:
Like, we have like a whole... And you're always like, hi, Agnes, are we doing
a podcast today?
Robin:
Right. So, and then there's these after conversations that are also very
structured. So people get comfortable with these very structured before and
after conversations or even pre-meetings, pre-meetings. Yeah. So, but once
they have a standard structure, then you go through the standard thing and
they kind of know what the rules are about that. And so there's less anxiety,
I guess, as you're sort of managed these transitions through these transition
styles in some sense. So maybe that's a thing to notice. We have transition
styles. We have transition styles, which are exactly intended to, or for
example, you know, when you have a meeting once a year with your boss about
your performance review, well, there's performance review styles. It may be a
little awkward.
Agnes:
There's the beginning of the opening move of the performance review. So I
think that the thing about foreplay is interesting because you might think
what the foreplay has to do when you say, hi, Agnes, we're having a podcast
today, or they say, oh, you can stop at any time or whatever, is that it has
to mediate the situation. Say we just turn on Zoom and we're not talking yet,
right? and that we're in that Wild West where we don't know how to interact,
right? We have to get out of that somehow. That's our first job as
conversation partners is get out of the Wild West. And so then there are these
very safe places we can go to where we can step over to how was your day with
small talk something, right? Where what small talk does is at least you're not
in the Wild West, at least you're not in the- Right. how do we manage this
conversation, whatever. That would be terrible. If you started and you turned
on Zoom and you were like, how are we going to manage this conversation? That
would be the worst case scenario. We're getting out of that by saying, here's
some boilerplate. And then once we did the boilerplate, it's almost like, OK,
we feel safe. We know we're not in the Wild West. We've safely gotten out of
that. And so we can move over to this slightly freer form of conversation
where we're going to talk and answer questions and whatever. And we're willing
to go there precisely because we see that we're not in too much danger of
going back to the Wild West. And so it's like we're taking baby steps in our
conversations, but there is this like worst case scenario that we're avoiding
the whole time.
Robin:
So you read much more ancient texts than I do, but whenever I read ancient
texts, I'm really struck by they're just invoking different standard style
packages there, right? That is, we're less familiar with their conversation
style package. And we have to learn a bit about how their styles were, but
there weren't that many, so it's not that hard, but it's still really
striking. It just emphasizes to me, there's this vast space of possible styles
that we aren't exploring, but maybe our descendants will, they might just
drift into other kinds of packages of expectations of how the conversation
might start or go or transition.
Agnes:
Yeah, I mean, one of the challenges of reading ancient texts is that most of
our texts don't resemble conversations. They're not casual. Getting a text
written down costs a lot of money. So you're not just gonna write, you know,
our podcast would not cut it in terms of being worth recording. And so what
are you going to do if you want to understand how people talk? I mean, I think
the best thing in the ancient world is you're going to look at comedies,
plays, I guess, plays, comedies, not tragedies, right? Comedies and tragedies
are really different. People talk really differently in comedies and
tragedies.
Robin:
That makes me distrust both of them. I mean, why don't they talk the same?
Agnes:
Well, because the tragedy, it's like there's this like august poeticness to
it.
Robin:
And, you know, it's supposed to be... Maybe there's a special comedic fun
thing about the comedy, which also is... I agree.
Agnes:
I don't think comedy is necessarily that great. It's just probably the best
we've got. You know, Plato's dialogues are probably, like, not terrible also.
I mean, that is, they're weird philosophy conversations, so, but there are
little bits at the, you know, that are, that, so that's unusually good, but
mostly our ancient texts are just not gonna tell us much about conversation.
That's just not the kind, that's not the style.
Robin:
And they didn't even really have transcripts of trials.
Agnes:
There are not too many transcripts going on. I mean, when literacy is not big,
when texts are very expensive to produce, you're not just going to do
transcripts of any random thing. You're going to copy Homer out another time.
Robin:
All right. Well, is it time to go meta here?
Agnes:
It's time to stop talking, which is not the same thing.
Robin:
Well, but by talking about the fact that it's time to stop talking is going
out for a fiora.
Agnes:
Yes, that is true.
Robin:
Ta-da. So, the end. Okay, the end. For an awkward way to be different at the
meta level here. We could put little letters across the screen, the end.
Agnes:
We can ask Will to do, yeah.
Robin:
Yeah, that might work. All right, stop for now.