In a widely-used simple approximation, the world of thinking and writing is a world of specific arguments. Readers collect their opinions by combining the arguments of those they read/hear with arguments that they think of for themselves. In this approximation, the only reason to interpret a writer/speaker is to figure out what exactly are his or her arguments.
Aug 13, 2023·edited Aug 13, 2023Liked by Robin Hanson
Mercier and Sperber claim reason evolved to win arguments, which also means reason is good at modeling other people's minds, so as to argue against them.
Joe Henrich claims that our minds evolved to copy cultural knowledge which we ourselves do not understand why it works. So we are programed to copy winners, and assume we can't figure out exactly why they are winners.
The combination of these two theories is our brains are tuned to model and be able to predict the actions of other high status people around us. And this means turning them into a sanded off mental model. A persona. The contradictory reality of what an actual person thinks is not what we meme and follow. Partly we meme persona the high status so we can blindly imitate what their persona would do, and partly so we can argue against that persona in status combat.
There is no evolutionary selection to actually understand logical arguments stand alone. It's more of a spandrel.
The test would be for people are good at abstract argument not tied to a persona, do they over history and prehistory tend to have fewer successful offspring or more. I suspect that capability is at best a non factor in reproductive success. And more likely negatively correlated with it.
> When I recently asked scholars who study Adam Smith what was their most controversial issue, they said it is whether Smith was on the left or right politically
Cringe. Partisan politics has broken peoples' brains.
Which is why I think your idea of historians striving for accuracy/consistency/truth is a nice romantic vision, but ultimately naive. Many (most?) historians (e.g. Howard Zinn) strive to further their own agenda and partisan worldview; history is about power, not about truth.
Or maybe I am just jaded and the 1% of partisan historians has soured me on the 99% of honest historians.
This is what we do with political leaders. When we vote for someone, we get a grab bag of policies, some of which we may disagree with, but we hope that the persona we interpret will lead to better results in the future. We might like someone’s policies but distrust their personality, find them somehow weak or untrustworthy. Politicians engage in ad hominem attacks because, in the end, the argument isn’t over the facts, but about who can be trusted.
It seems natural to see versions of the same taking place in intellectual life. When people decide they trust someone, they’re motivated to interpret their views in ways that align with their own, and to create an aligned persona that gives their own views more authority. And that’s also why, when a commonly revered authority held or holds views that some find disagreeable, they seek to undermine trust in the person.
This seems related to the general phenomenon of name dropping, that within and across various circles, competing hierarchies of high status intellectuals and such emerge, and that by name dropping such individuals and tying ones own arguments to their personas, one can signal to various other people what circles you belong to aswell as your status/depth within such a intellectual tradition. A good case study would be the various circles around philosophy, academic, analytic, continental, public, etc. With names such as Korsgaard, Nussbaum, Parfit, Wittgenstein, Putnam, Russell, Kant, Hegel, Plato, Derrida, Rand, Peterson, Harris etc. serving as signals as to which circles you occupy and how far you have researched into them. The value of such name dropping is further amplified when such characters have a cool sounding name, picture and story to go along, which is the reason as to why people such as David Lewis or Saul Kripke are only mentioned in certain highly selected circles, whereas some random French or German philosopher with the above is name dropped even in fairly mainstream circles.
To understand how personas work, I am tempted to start from the assumption that personas are not created by individual people but by groups of people (societies, tribes, subcultures, etc). (Of course, an individual is a group of size one, so individuals are a special case of groups.) If we then find a good example where the same person has different personas in different subcultures, maybe we can use our knowledge of the subcultures to shed light on your question of how personas work. For example, I would expect that personas serve a purpose in the subculture that is quite independent of the actual person. Ideally the example would oppose two personas (of the same person) that are equally powerful but quite far apart from each other.
I'm not sure I can think of any thinker who could really be described as something as complex as a persona. Figures in intellectual history who survive are reduced to characters in a poorly written play - if they're very influential, characters with a few lines, or even characters in a couple of different plays, but ultimately they're just a means of putting across a story.
And their most important attributes aren't necessarily the lines they speak. They're often on stage primarily to give the general narrative a sense of status, of class, by virtue of the things the audience will associate them with. It's no accident that most of the great thinkers are ancient Greeks from around the time of Alexander, Augustinian Romans, or Europeans from periods when their nation was particularly successful.
Aug 13, 2023·edited Aug 13, 2023Liked by Robin Hanson
Attaching oneself to a prestigious historical thinker is certainly most of the motivation. There is also signaling that you are a person of historical education, who understands the ideas of history's greatest thinkers. Anybody who questions you will have to deal with the rhetorical trick of an historical reference that they may not be aware of. People who make historical references are considered to be intelligent. Luckily for them, they are also among the least assailable in their contentions, except by others who've devoted their time to the study of history trivia. That sort of history mining, is generally in the pursuit of confirmation bias. One can find anything one wants to find, in historical anecdote.
Role model imitation is the default way most people learn. It's hard to think logically or think for themselves, and their peers or superiors may not approve of the results anyway. So people evolved to pick a high-status member of their tribe and just try to act like them. This is at the level of an instinct.
That's why many people are so concerned with interpreting what the highest-status intellectuals "would" do. If you like Adam Smith, if you think he's a high-quality high-status thinker, then you want to mimic him and be just like him. In order to do that it is necessary to model what he would do or think in different situations. Then you can just follow the model. The implicit inference is always, "Adam Smith would do or think X about Y, if he were asked to consider Y. Adam Smith is authoritative. Therefore, we should all do or think X about Y, and look down on deviants who don't." It's the same idea as, "What would the founding fathers think?"
Of course, nobody says this reasoning out loud, because when the whole argument is laid out like that, a moment's reflection shows it to be stupid. An argument rests on its own merits; we should be concerned with arguments for what is true, not with what some high-status person said or would say, except insofar as what some high-status person said is useful to determine what is true. High-status people are often wrong, just like the rest of us. But this high-status-mimicking behavior is at the level of an instinct, so logic doesn't matter.
I find your article very interesting. I hope you dont mind if I describe a personal experience that comes to my mind. I first encountered Jordan Peterson in blogs that others wrote about him. A somewhat coherent picture emerged. Using your concepts, I'd say now that this picture had little to do with Peterson the thinker and concerned Peterson the persona (relative to a certain subculture). Only after several friends talked about him was I intrigued enough to listen to Peterson the thinker. A completely different picture emerged. To understand the world we need to create abstractions and shortcuts, so creating personas is probably unavoidable. But can we develop a culture in which these personas are *useful* (as opposed to misleading) abstractions?
Human thinkers are only occasionally the ideal version of themselves
Ayn Rand is famous for lots of thinking, and an awful lot of it was solid -- but she also had LOTS of opinions that were "extraneous" -- and she was bombastic besides.
I find the question of what the ideal Rand thinks about a question more interesting than the question of whether Rand had a specific problem with Mozart.
----
If I'm thinking as a Hansonian -- I almost always start from the famous "Hansonian Credo"
- Applying models/theories to new situations is difficult, and the heuristic that the best person to do this is the person who developed said model/theory feels right. Imagine that I've read The Republic and I come away very impressed and convinced that Plato basically got it all right. Now I'm asked to weigh in on whether a proposed change in our social structure is good or bad. I look through my notes, but they suddenly seem too abstract to apply to this practical problem. Instead I can ask myself the age old question of 'What Would Plato Do?', of course if anyone is going to apply the ideas correctly it's going to be the person who came up with them in the first place. This works doubly well if I can find some record of Plato facing a similar situation.
- Evaluating models/theories is difficult, the heuristic of dismissing ideas from people who hold other bad/weird ideas feels right. This is one is pretty self-explanatory, and sounds like something to be avoided, but I think it has its uses.
It is essentially never the case that anyone, yourself included, writes out every logically necessary step of an argument they make. Not even math proofs do that consistently. Written arguments in natural language rely on the reader to fill in gaps. How those gaps should get filled in depends in part on the overall mindset of the writer, so readers construct personas in part to help them do that better. These personas also provide a general sense of how reliable, consistent, and coherent a thinker is. They help see when someone has changed their mind on something, which is relevant for knowing whether they still endorse their own past arguments. Since no one has the bandwidth to deeply engage with everyone's arguments on everything, thise knowledge helps us to decide who to bother reading in the first place.
The persona is often more useful to me than the real thinker. Maybe professionally I’m trying to imagine, what features is company X likely to add to their product. I can think, we’ll if they are a Steve Jobs type thinker, they will do this. If they are an Elon Musk type thinker, they will do that. Etc. The actual opinions of Steve Jobs and Elon Musk are not so useful to me because they are not giving their opinion on the question I care about.
A “persona” is a broad, general, even all-encompassing theory. In addressing a limited range of topics, an intellectual may be providing glimpses of an over-arching Weltanschauung that he embodies, perhaps without having articulated it, even for himself. This implicit theory-of-everything, if it exists, would be of interest to his readers; and if it seems defective to them now, they would also be interested in seeing the closest approximation to it that was not obviously invalid--a supposed rectified or “cleaned up” persona.
In Brazil, there is one guy, post-keynesian, that once wrote (it is in a published book's chapter) something like: "When Keynes said X, it was really thinking on Y". It's the most extreme case of interpretation (kind of an exegetics...) I ever seen.
Mercier and Sperber claim reason evolved to win arguments, which also means reason is good at modeling other people's minds, so as to argue against them.
Joe Henrich claims that our minds evolved to copy cultural knowledge which we ourselves do not understand why it works. So we are programed to copy winners, and assume we can't figure out exactly why they are winners.
The combination of these two theories is our brains are tuned to model and be able to predict the actions of other high status people around us. And this means turning them into a sanded off mental model. A persona. The contradictory reality of what an actual person thinks is not what we meme and follow. Partly we meme persona the high status so we can blindly imitate what their persona would do, and partly so we can argue against that persona in status combat.
There is no evolutionary selection to actually understand logical arguments stand alone. It's more of a spandrel.
The test would be for people are good at abstract argument not tied to a persona, do they over history and prehistory tend to have fewer successful offspring or more. I suspect that capability is at best a non factor in reproductive success. And more likely negatively correlated with it.
> When I recently asked scholars who study Adam Smith what was their most controversial issue, they said it is whether Smith was on the left or right politically
Cringe. Partisan politics has broken peoples' brains.
Which is why I think your idea of historians striving for accuracy/consistency/truth is a nice romantic vision, but ultimately naive. Many (most?) historians (e.g. Howard Zinn) strive to further their own agenda and partisan worldview; history is about power, not about truth.
Or maybe I am just jaded and the 1% of partisan historians has soured me on the 99% of honest historians.
This is what we do with political leaders. When we vote for someone, we get a grab bag of policies, some of which we may disagree with, but we hope that the persona we interpret will lead to better results in the future. We might like someone’s policies but distrust their personality, find them somehow weak or untrustworthy. Politicians engage in ad hominem attacks because, in the end, the argument isn’t over the facts, but about who can be trusted.
It seems natural to see versions of the same taking place in intellectual life. When people decide they trust someone, they’re motivated to interpret their views in ways that align with their own, and to create an aligned persona that gives their own views more authority. And that’s also why, when a commonly revered authority held or holds views that some find disagreeable, they seek to undermine trust in the person.
This seems related to the general phenomenon of name dropping, that within and across various circles, competing hierarchies of high status intellectuals and such emerge, and that by name dropping such individuals and tying ones own arguments to their personas, one can signal to various other people what circles you belong to aswell as your status/depth within such a intellectual tradition. A good case study would be the various circles around philosophy, academic, analytic, continental, public, etc. With names such as Korsgaard, Nussbaum, Parfit, Wittgenstein, Putnam, Russell, Kant, Hegel, Plato, Derrida, Rand, Peterson, Harris etc. serving as signals as to which circles you occupy and how far you have researched into them. The value of such name dropping is further amplified when such characters have a cool sounding name, picture and story to go along, which is the reason as to why people such as David Lewis or Saul Kripke are only mentioned in certain highly selected circles, whereas some random French or German philosopher with the above is name dropped even in fairly mainstream circles.
To understand how personas work, I am tempted to start from the assumption that personas are not created by individual people but by groups of people (societies, tribes, subcultures, etc). (Of course, an individual is a group of size one, so individuals are a special case of groups.) If we then find a good example where the same person has different personas in different subcultures, maybe we can use our knowledge of the subcultures to shed light on your question of how personas work. For example, I would expect that personas serve a purpose in the subculture that is quite independent of the actual person. Ideally the example would oppose two personas (of the same person) that are equally powerful but quite far apart from each other.
I'm not sure I can think of any thinker who could really be described as something as complex as a persona. Figures in intellectual history who survive are reduced to characters in a poorly written play - if they're very influential, characters with a few lines, or even characters in a couple of different plays, but ultimately they're just a means of putting across a story.
And their most important attributes aren't necessarily the lines they speak. They're often on stage primarily to give the general narrative a sense of status, of class, by virtue of the things the audience will associate them with. It's no accident that most of the great thinkers are ancient Greeks from around the time of Alexander, Augustinian Romans, or Europeans from periods when their nation was particularly successful.
Attaching oneself to a prestigious historical thinker is certainly most of the motivation. There is also signaling that you are a person of historical education, who understands the ideas of history's greatest thinkers. Anybody who questions you will have to deal with the rhetorical trick of an historical reference that they may not be aware of. People who make historical references are considered to be intelligent. Luckily for them, they are also among the least assailable in their contentions, except by others who've devoted their time to the study of history trivia. That sort of history mining, is generally in the pursuit of confirmation bias. One can find anything one wants to find, in historical anecdote.
Role model imitation is the default way most people learn. It's hard to think logically or think for themselves, and their peers or superiors may not approve of the results anyway. So people evolved to pick a high-status member of their tribe and just try to act like them. This is at the level of an instinct.
That's why many people are so concerned with interpreting what the highest-status intellectuals "would" do. If you like Adam Smith, if you think he's a high-quality high-status thinker, then you want to mimic him and be just like him. In order to do that it is necessary to model what he would do or think in different situations. Then you can just follow the model. The implicit inference is always, "Adam Smith would do or think X about Y, if he were asked to consider Y. Adam Smith is authoritative. Therefore, we should all do or think X about Y, and look down on deviants who don't." It's the same idea as, "What would the founding fathers think?"
Of course, nobody says this reasoning out loud, because when the whole argument is laid out like that, a moment's reflection shows it to be stupid. An argument rests on its own merits; we should be concerned with arguments for what is true, not with what some high-status person said or would say, except insofar as what some high-status person said is useful to determine what is true. High-status people are often wrong, just like the rest of us. But this high-status-mimicking behavior is at the level of an instinct, so logic doesn't matter.
I find your article very interesting. I hope you dont mind if I describe a personal experience that comes to my mind. I first encountered Jordan Peterson in blogs that others wrote about him. A somewhat coherent picture emerged. Using your concepts, I'd say now that this picture had little to do with Peterson the thinker and concerned Peterson the persona (relative to a certain subculture). Only after several friends talked about him was I intrigued enough to listen to Peterson the thinker. A completely different picture emerged. To understand the world we need to create abstractions and shortcuts, so creating personas is probably unavoidable. But can we develop a culture in which these personas are *useful* (as opposed to misleading) abstractions?
Human thinkers are only occasionally the ideal version of themselves
Ayn Rand is famous for lots of thinking, and an awful lot of it was solid -- but she also had LOTS of opinions that were "extraneous" -- and she was bombastic besides.
I find the question of what the ideal Rand thinks about a question more interesting than the question of whether Rand had a specific problem with Mozart.
----
If I'm thinking as a Hansonian -- I almost always start from the famous "Hansonian Credo"
https://www.overcomingbias.com/p/politics-isnt-ahtml
It's a lens that helps ME think, not a set of positions.
And ... well...that's how I approach all the ideas/positions?
Where does this direction of thinking take me ?
Two main reasons come to mind:
- Applying models/theories to new situations is difficult, and the heuristic that the best person to do this is the person who developed said model/theory feels right. Imagine that I've read The Republic and I come away very impressed and convinced that Plato basically got it all right. Now I'm asked to weigh in on whether a proposed change in our social structure is good or bad. I look through my notes, but they suddenly seem too abstract to apply to this practical problem. Instead I can ask myself the age old question of 'What Would Plato Do?', of course if anyone is going to apply the ideas correctly it's going to be the person who came up with them in the first place. This works doubly well if I can find some record of Plato facing a similar situation.
- Evaluating models/theories is difficult, the heuristic of dismissing ideas from people who hold other bad/weird ideas feels right. This is one is pretty self-explanatory, and sounds like something to be avoided, but I think it has its uses.
It is essentially never the case that anyone, yourself included, writes out every logically necessary step of an argument they make. Not even math proofs do that consistently. Written arguments in natural language rely on the reader to fill in gaps. How those gaps should get filled in depends in part on the overall mindset of the writer, so readers construct personas in part to help them do that better. These personas also provide a general sense of how reliable, consistent, and coherent a thinker is. They help see when someone has changed their mind on something, which is relevant for knowing whether they still endorse their own past arguments. Since no one has the bandwidth to deeply engage with everyone's arguments on everything, thise knowledge helps us to decide who to bother reading in the first place.
The persona is often more useful to me than the real thinker. Maybe professionally I’m trying to imagine, what features is company X likely to add to their product. I can think, we’ll if they are a Steve Jobs type thinker, they will do this. If they are an Elon Musk type thinker, they will do that. Etc. The actual opinions of Steve Jobs and Elon Musk are not so useful to me because they are not giving their opinion on the question I care about.
A “persona” is a broad, general, even all-encompassing theory. In addressing a limited range of topics, an intellectual may be providing glimpses of an over-arching Weltanschauung that he embodies, perhaps without having articulated it, even for himself. This implicit theory-of-everything, if it exists, would be of interest to his readers; and if it seems defective to them now, they would also be interested in seeing the closest approximation to it that was not obviously invalid--a supposed rectified or “cleaned up” persona.
In Brazil, there is one guy, post-keynesian, that once wrote (it is in a published book's chapter) something like: "When Keynes said X, it was really thinking on Y". It's the most extreme case of interpretation (kind of an exegetics...) I ever seen.
some minor notes, 2023-08-14:
" her or her persona ..."
"to engage most the messy details"