In our world, we use many standard markers of status. These include personal connections with high status people and institutions, power, wealth, popularity, charisma, intelligence, eloquence, courage, athleticism, beauty, distinctive memorable personal styles, and participation in difficult achievements. We also use these same status markers for intellectuals, though specific fields favor specific variations. For example, in economics we favor complex game theory proofs and statistical analyses of expensive data as types of difficult achievements.
One alternative hypothesis is that the insight only becomes widely known once it is said by a high-status individual, even if it was said previously by others. If this is true, then currently insiders may actually perceive that high-status individuals also routinely have insight, infer a causal relationship between insight and status, and so pursue insight. In that case, exposing the hypocrisy would not change the inferred incentives, though it would make their beliefs more accurate.
The rate of insight production in the current system seems low enough that it can be explained as a random side effect, not at all selected for or intended.
I think I get what you're saying. When you talk about status and hypocrisy the way you do, you sound like someone airing moral grievances, although I've been reading your blog long enough that I should know better by now than to assume that. So I'm trying to think nonjudgmentally about how the status economy produces scholarship. To make an analogy with the money economy: people often lament that monetary incentives frequently motivate us to do things which are, according to their ethical perspective, useless or even harmful, even as dire problems go unremedied. Unfortunately, it turns out that it's often really difficult to change institutions, culture, or whatever to solve an incentive problem, without creating new incentive problems elsewhere in the system, and evaluating whether those tradeoffs are worth it is also difficult.
So, taking it as given that academics and their financial patrons are primarily motivated by status, I'm just trying to think about why the academic system we have produces any useful insight at all (because it clearly does), and what tradeoffs we would be likely to encounter if we were to implement changes aimed at better aligning status incentives with the production of useful insights. Returning to the analogy with money: one common complaint about the economy is that we have a lot of rich people spending a lot of money on luxuries, while a much greater number of people suffer from poverty. But societies which try to confiscate all wealth above some modest level, for redistribution to the needy, don't have a great track record for promoting human welfare, compared to societies which permit a significant degree of economic inequality. Presumably this is because redistribution requires swimming against the very powerful current of human greed, so at some point the undesired side-effects outweigh the benefits. Perhaps it would backfire in a similar way if we were too successful at decoupling the assignment of credit for new knowledge from traditional status markers. E.g. what if research funders took up your proposal to fund more careful documentation and promulgation of information about who deserves credit for scientific progress, and the result was that the activity of successfully advancing science came to be associated with people who lack markers of high status?
Wikipedia has some relevant lists. Here's their list of scientific priority disputes. If you scroll to the bottom of the page, you can find links to a list of examples of Stigler's law (discoveries named after people who weren't the first to make them) and a list of discoveries made by multiple people/groups working independently. Two examples that jumped out at me for the length of time between discovery and popularization:
"Benford's law, named after physicist Frank Benford, who stated it in 1938, although it had been previously stated by Simon Newcomb in 1881"
"In a treatise written in 1805 and published in 1866, Carl Friedrich Gauss describes an efficient algorithm to compute the discrete Fourier transform. James W. Cooley and John W. Tukey reinvented a similar algorithm [the fast Fourier transform] in 1965."
Roles and the people in them may not be neatly separable. Even if a scholar doesn't have any special title or formal position of leadership, they still might become high-status and influential just by being really good at social networking and anticipating trends in what is considered important in their field. (Although they could also become high-status and influential by fraud and underhanded political maneuvering, or just lucking into a position that wasn't very gatekept when they got it, but subsequently became important due to unpredictable events.)
Anyway, whether we assign credit to the people in powerful roles, or to the roles themselves, affects incentives; even if the person occupying an important position of influence could be replaced by a dozen other equally capable candidates, we still want that person to expect that they will be rewarded handsomely for doing a good job, and blamed for doing a lousy job. I suppose it's still hypocritical to credit such people for discoveries that they popularized but did not originate, and more generally to talk about them as if their status were solely the result of the quality of their contributions to the accumulated knowledge of their field.
You reply is entirely correct but I think it is beside the point. Certainly I tend to do what I am incentivized to do and popular histories of pretty much anything usually contain many misattributions. However, what I'm claiming is that my incentives aren't related to popular histories.
You claim that there is an enormous misalignment between the stated goals of intellectuals and what they are actually doing, a misalignment so large that correcting it could double productivity. I think that if this were the case, there would be more evidence of it than just a hypothetical about the role of popular histories.
Whatever your position, the official histories have a place for you, even if as a very minor character. If that place is wrongly described, and status is assigned differently as a result, then your incentives are different as a result.
Even if we accept the claim that intellectuals are generally misinformed about the history of their fields, it does not immediately follow that informing them would lead to increased production of "insight". Say I want to be the next Newton. If it turns out that Leibniz actually discovered calculus first, this just changes the name I associate with my ambition, not the ambition itself.
Furthermore, most competition for status among intellectuals is among people who never have made history and likely never will. Such people still do productive work, and some of them are better at it then others; ranking them is useful and generally does not involve comparing them to historical figures. In other words, my boss doesn't judge my work by comparing me to Von Neumann or Turing. I'm nobody compared to either Von Neumann, Turing, or whoever else really deserves credit for their discoveries if such a person exists. But I can still be better than the other guy being considered for a promotion.
To deserve credit, it isn't enough that powerful insiders filter and choose, they should also be much better at doing do than others who could sit in their powerful role. Otherwise it is the roie that deserves credit, not the people sitting in them.
The main empirical evidence supporting the hypothesis that we are hypocritical about what counts for status is that standard histories “tend to be especially wrong when they claim that a prominent figure was the first to take a position or make an argument. One can usually find lower status people who said basically the same things before.” Here is an alternative hypothesis which fits that evidence. Prominent figures really are good at advancing science. They have proven that throughout their careers. There is too much new information produced in any given field for every scholar to keep up with it themselves, so they filter by following prominent people who (in the view of the individual scholar) have proven that they are good enough to reliably filter the wheat from the chaff. When a low status person comes up with a new idea, of course most people ignore it, because it is impossible to read every article in the field and come up with your own explanation. But when a high status person picks up on that article, it is a reliable signal that it is really a good idea, and so it takes off.
The other claim made to support the hypocrisy argument is “high status accomplishments tend to be given more credit than they deserve in causing opinion change.” This is not fleshed out, so it is hard to know what counts as more than they deserve. In my hypothesis, both the low and high status person deserve some of the credit for causing the opinion change.
The Franklin story is often garbled. She was a thoroughgoing empiricist, so when Crick & Watson proposed their initial theory of its structure, she was skeptical. As she should have been, because it was wrong. But they were happy to then spin up another theory that turned out to be right, and again she was initially skeptical and thought more empirical work was needed. But there was also some bureaucratic bungling there where she'd been told she was in charge of DNA research at that university but her colleagues hadn't been so informed and couldn't understand why she was acting as if she was, leading to some badmouthing of her to C&W. But if you want a specific reason why she didn't share the Nobel (as Maurice Wilkins, who shared some of her research with them, did), she had died before it was given.
That's Stigler's law of eponymy (which Stephen Stigler credited Robert Merton with discovering, although I think the latter is more prominent than the former).
One alternative hypothesis is that the insight only becomes widely known once it is said by a high-status individual, even if it was said previously by others. If this is true, then currently insiders may actually perceive that high-status individuals also routinely have insight, infer a causal relationship between insight and status, and so pursue insight. In that case, exposing the hypocrisy would not change the inferred incentives, though it would make their beliefs more accurate.
The rate of insight production in the current system seems low enough that it can be explained as a random side effect, not at all selected for or intended.
I think I get what you're saying. When you talk about status and hypocrisy the way you do, you sound like someone airing moral grievances, although I've been reading your blog long enough that I should know better by now than to assume that. So I'm trying to think nonjudgmentally about how the status economy produces scholarship. To make an analogy with the money economy: people often lament that monetary incentives frequently motivate us to do things which are, according to their ethical perspective, useless or even harmful, even as dire problems go unremedied. Unfortunately, it turns out that it's often really difficult to change institutions, culture, or whatever to solve an incentive problem, without creating new incentive problems elsewhere in the system, and evaluating whether those tradeoffs are worth it is also difficult.
So, taking it as given that academics and their financial patrons are primarily motivated by status, I'm just trying to think about why the academic system we have produces any useful insight at all (because it clearly does), and what tradeoffs we would be likely to encounter if we were to implement changes aimed at better aligning status incentives with the production of useful insights. Returning to the analogy with money: one common complaint about the economy is that we have a lot of rich people spending a lot of money on luxuries, while a much greater number of people suffer from poverty. But societies which try to confiscate all wealth above some modest level, for redistribution to the needy, don't have a great track record for promoting human welfare, compared to societies which permit a significant degree of economic inequality. Presumably this is because redistribution requires swimming against the very powerful current of human greed, so at some point the undesired side-effects outweigh the benefits. Perhaps it would backfire in a similar way if we were too successful at decoupling the assignment of credit for new knowledge from traditional status markers. E.g. what if research funders took up your proposal to fund more careful documentation and promulgation of information about who deserves credit for scientific progress, and the result was that the activity of successfully advancing science came to be associated with people who lack markers of high status?
I'm saying they are mainly rewarded for showing marks of status, not for inducing intellectual progress.
Wikipedia has some relevant lists. Here's their list of scientific priority disputes. If you scroll to the bottom of the page, you can find links to a list of examples of Stigler's law (discoveries named after people who weren't the first to make them) and a list of discoveries made by multiple people/groups working independently. Two examples that jumped out at me for the length of time between discovery and popularization:
"Benford's law, named after physicist Frank Benford, who stated it in 1938, although it had been previously stated by Simon Newcomb in 1881"
"In a treatise written in 1805 and published in 1866, Carl Friedrich Gauss describes an efficient algorithm to compute the discrete Fourier transform. James W. Cooley and John W. Tukey reinvented a similar algorithm [the fast Fourier transform] in 1965."
Roles and the people in them may not be neatly separable. Even if a scholar doesn't have any special title or formal position of leadership, they still might become high-status and influential just by being really good at social networking and anticipating trends in what is considered important in their field. (Although they could also become high-status and influential by fraud and underhanded political maneuvering, or just lucking into a position that wasn't very gatekept when they got it, but subsequently became important due to unpredictable events.)
Anyway, whether we assign credit to the people in powerful roles, or to the roles themselves, affects incentives; even if the person occupying an important position of influence could be replaced by a dozen other equally capable candidates, we still want that person to expect that they will be rewarded handsomely for doing a good job, and blamed for doing a lousy job. I suppose it's still hypocritical to credit such people for discoveries that they popularized but did not originate, and more generally to talk about them as if their status were solely the result of the quality of their contributions to the accumulated knowledge of their field.
As I live in this world, is its quite obvious to me that we waste far more than a mere factor of two.
You reply is entirely correct but I think it is beside the point. Certainly I tend to do what I am incentivized to do and popular histories of pretty much anything usually contain many misattributions. However, what I'm claiming is that my incentives aren't related to popular histories.
You claim that there is an enormous misalignment between the stated goals of intellectuals and what they are actually doing, a misalignment so large that correcting it could double productivity. I think that if this were the case, there would be more evidence of it than just a hypothetical about the role of popular histories.
Whatever your position, the official histories have a place for you, even if as a very minor character. If that place is wrongly described, and status is assigned differently as a result, then your incentives are different as a result.
Even if we accept the claim that intellectuals are generally misinformed about the history of their fields, it does not immediately follow that informing them would lead to increased production of "insight". Say I want to be the next Newton. If it turns out that Leibniz actually discovered calculus first, this just changes the name I associate with my ambition, not the ambition itself.
Furthermore, most competition for status among intellectuals is among people who never have made history and likely never will. Such people still do productive work, and some of them are better at it then others; ranking them is useful and generally does not involve comparing them to historical figures. In other words, my boss doesn't judge my work by comparing me to Von Neumann or Turing. I'm nobody compared to either Von Neumann, Turing, or whoever else really deserves credit for their discoveries if such a person exists. But I can still be better than the other guy being considered for a promotion.
To deserve credit, it isn't enough that powerful insiders filter and choose, they should also be much better at doing do than others who could sit in their powerful role. Otherwise it is the roie that deserves credit, not the people sitting in them.
The main empirical evidence supporting the hypothesis that we are hypocritical about what counts for status is that standard histories “tend to be especially wrong when they claim that a prominent figure was the first to take a position or make an argument. One can usually find lower status people who said basically the same things before.” Here is an alternative hypothesis which fits that evidence. Prominent figures really are good at advancing science. They have proven that throughout their careers. There is too much new information produced in any given field for every scholar to keep up with it themselves, so they filter by following prominent people who (in the view of the individual scholar) have proven that they are good enough to reliably filter the wheat from the chaff. When a low status person comes up with a new idea, of course most people ignore it, because it is impossible to read every article in the field and come up with your own explanation. But when a high status person picks up on that article, it is a reliable signal that it is really a good idea, and so it takes off.
The other claim made to support the hypocrisy argument is “high status accomplishments tend to be given more credit than they deserve in causing opinion change.” This is not fleshed out, so it is hard to know what counts as more than they deserve. In my hypothesis, both the low and high status person deserve some of the credit for causing the opinion change.
Kolmogorov complexity was conceived by Ray Solomonoff.
The Franklin story is often garbled. She was a thoroughgoing empiricist, so when Crick & Watson proposed their initial theory of its structure, she was skeptical. As she should have been, because it was wrong. But they were happy to then spin up another theory that turned out to be right, and again she was initially skeptical and thought more empirical work was needed. But there was also some bureaucratic bungling there where she'd been told she was in charge of DNA research at that university but her colleagues hadn't been so informed and couldn't understand why she was acting as if she was, leading to some badmouthing of her to C&W. But if you want a specific reason why she didn't share the Nobel (as Maurice Wilkins, who shared some of her research with them, did), she had died before it was given.
That's Stigler's law of eponymy (which Stephen Stigler credited Robert Merton with discovering, although I think the latter is more prominent than the former).
I took Cosmology in college from Virginia Trimble, who constantly said "this is called the X effect/thing, though it was discovered by Y."