ETA: I’ve updated my position since writing this comment and no longer endorse it; see this comment.
I agree with Lumifer and shminux’s comments. More generally, this does not seem like the sort of thing that you should be making conclusions about without taking a look at actual data when actual data is easily available. I recommend Neihart (1999). At best, the picture is a lot more nuanced than you’re making it out to be, and at worst, you’re regurgitating a cached thought: “Gifted children are poorly adjusted compared to their peers.” The closest thing to this in the literature is the correlation between highly creative adults and psychopathology, particularly mood disorders, psychosis, and suicidality; not more general giftedness and social competence. There isn’t enough evidence to conclude that this generalizes, in terms of populations, to highly intelligent individuals of all ages or to highly creative, or more generally, highly intelligent, children; or that this generalizes, in terms of individual characteristics, to social competence.
As for giftedness and measurements of psychological adjustment besides social competence, most studies have found little correlation between depression, anxiety, or suicidality and giftedness, and what correlation there is errs on the side of gifted children being better-adjusted than their average peers. Gifted children are also far less likely to engage in deviant behavior.
As for social competence, it’s diverse among gifted persons. We can’t conclude that either intelligence or personality factors are the primary causal factor in any purported correlation between the two and social competence. That is, your model is “Gifted children are highly intelligent. Gifted children are dissimilar from their peers because they are highly intelligent. Their peers ostracize them because they are dissimilar. Ostracized children engage in less social interaction. Gifted children have poor social skills because they have less experience with social interaction.” A similarly plausible model is: “Some gifted children feel dissimilar from their peers. Gifted children who feel dissimilar from their peers are less likely to interact with their peers. Gifted children who interact less have poor social skills.” And it can be both, and/or something else. There’s also some evidence that verbally precocious children are less socially competent, or at least perceive themselves so, than mathematically precocious children. Finally, a big confounder in this entire field of inquiry is the correlation between high socioeconomic class and giftedness; it’s hard to get a large control group that isn’t also more socioeconomically heterogeneous than the experimental group.
If we’re going to armchair psychologize, then I much prefer Paul Graham’s model. High-IQ societies suffer from the same deficiency as public schools and prisons: their selection criteria are neither the motives of their members nor individual characteristics highly correlated with particular motives, groups composed of individuals with diffuse motives are directionless, and directionless groups that are prevented from decomposing into smaller more purposeful groups degenerate into unadulterated status games.
This community has a tendency to speculate wildly, especially about psychology and sociology. This can be useful. Sometimes there’s no data to look at. But when there’s data to look at, you should look at it. Jonah, do you disagree that this article has unnecessarily low epistemic standards? Usually when this criticism is made, you point out that you are still explaining and not yet defending, but there are clear misunderstandings of psychopathology, as detailed below by Lumifer, and direct contradictions with empirical evidence, as I have detailed above, so I don’t see how that response would be applicable in this context.
If we’re going to armchair psychologize, then I much prefer Paul Graham’s model.
Honestly, I think that is the worst model. I remember how we, the 3-4 nerds in the class, how bitter we were about the popular kids and how we have gladly sacrificed our intelligence or books or anything to be like them, be boys who are respected by boys and loved by girls. Our lives were characterized of bitter envy of the popular kids, with the occasional desperate sour-graping.
Paul Graham is used to productive nerds, teenagers who at 16 are already writing useful software or learning science and really ignore what others do. Perhaps this model is useful for them.
But it is not useful at all for the larger number of less productive, more escapist nerds in the local D&D or MtG club. (And the difference is not intelligence but personality type, interests and so on, there are people in the Mensa who are like this.)
Yeah, I do consider it armchair psychology. He seems to be generalizing too much about the psychological characteristics of the empirical cluster to which he refers as nerds. I found the interesting part of the essay to be the hypotheses about perverse incentives explaining the similarities between the milieux of American prisons and American public schools, and that’s really why I brought it up here.
I was under the impression that US schools have far more surveillance. I mean my school experience was like that, but it was a poor school at the ass end of Europe, no campus police no security guards no cameras no staff nothing, just teachers and kids, and teachers usually spending the breaks in their lounge, drinking coffee, so nothing kids did to each other was noticed and punished, pretty much free reign. I sort of expected US schools to be under far more surveillance. Prisons, too. There are no privacy rights in prison, none at all. If a country can afford it, there is supposed to be camera in every corner?
That depends. US schools are run by local (not even state) governments, usually towns, and there is huge diversity in what a school might look like.
Public schools in big cities (NYC, Philly, etc.) basically are prisons with wire fences, metal detectors, cameras everywhere, etc. I’ve seen a school with no windows—it was designed this way so that there were no windows to break. But schools in e.g. rich suburbs are nice places, similar to college campuses.
I don’t quite understand your point. At any rate, in my experience, I only encountered two surveillance cameras that were privately owned by tech-savvy teachers, there was usually one (sometimes two) ‘SRO’ (school resource officer), a police officer assigned to the campus, and we were almost always supervised by adults. Graham’s hypothesis is plausible even given poor supervision. The only requirements are 1) adults and children not being invested in the ostensible purpose of the educational system and 2a) the presence of obstacles to refusing to participate in or 2b) reforming the educational system.
ETA: I’ve updated my position since writing this comment and no longer endorse it; see this comment.
I agree with Lumifer and shminux’s comments. More generally, this does not seem like the sort of thing that you should be making conclusions about without taking a look at actual data when actual data is easily available. I recommend Neihart (1999). At best, the picture is a lot more nuanced than you’re making it out to be, and at worst, you’re regurgitating a cached thought: “Gifted children are poorly adjusted compared to their peers.” The closest thing to this in the literature is the correlation between highly creative adults and psychopathology, particularly mood disorders, psychosis, and suicidality; not more general giftedness and social competence. There isn’t enough evidence to conclude that this generalizes, in terms of populations, to highly intelligent individuals of all ages or to highly creative, or more generally, highly intelligent, children; or that this generalizes, in terms of individual characteristics, to social competence.
As for giftedness and measurements of psychological adjustment besides social competence, most studies have found little correlation between depression, anxiety, or suicidality and giftedness, and what correlation there is errs on the side of gifted children being better-adjusted than their average peers. Gifted children are also far less likely to engage in deviant behavior.
As for social competence, it’s diverse among gifted persons. We can’t conclude that either intelligence or personality factors are the primary causal factor in any purported correlation between the two and social competence. That is, your model is “Gifted children are highly intelligent. Gifted children are dissimilar from their peers because they are highly intelligent. Their peers ostracize them because they are dissimilar. Ostracized children engage in less social interaction. Gifted children have poor social skills because they have less experience with social interaction.” A similarly plausible model is: “Some gifted children feel dissimilar from their peers. Gifted children who feel dissimilar from their peers are less likely to interact with their peers. Gifted children who interact less have poor social skills.” And it can be both, and/or something else. There’s also some evidence that verbally precocious children are less socially competent, or at least perceive themselves so, than mathematically precocious children. Finally, a big confounder in this entire field of inquiry is the correlation between high socioeconomic class and giftedness; it’s hard to get a large control group that isn’t also more socioeconomically heterogeneous than the experimental group.
If we’re going to armchair psychologize, then I much prefer Paul Graham’s model. High-IQ societies suffer from the same deficiency as public schools and prisons: their selection criteria are neither the motives of their members nor individual characteristics highly correlated with particular motives, groups composed of individuals with diffuse motives are directionless, and directionless groups that are prevented from decomposing into smaller more purposeful groups degenerate into unadulterated status games.
This community has a tendency to speculate wildly, especially about psychology and sociology. This can be useful. Sometimes there’s no data to look at. But when there’s data to look at, you should look at it. Jonah, do you disagree that this article has unnecessarily low epistemic standards? Usually when this criticism is made, you point out that you are still explaining and not yet defending, but there are clear misunderstandings of psychopathology, as detailed below by Lumifer, and direct contradictions with empirical evidence, as I have detailed above, so I don’t see how that response would be applicable in this context.
Honestly, I think that is the worst model. I remember how we, the 3-4 nerds in the class, how bitter we were about the popular kids and how we have gladly sacrificed our intelligence or books or anything to be like them, be boys who are respected by boys and loved by girls. Our lives were characterized of bitter envy of the popular kids, with the occasional desperate sour-graping.
Paul Graham is used to productive nerds, teenagers who at 16 are already writing useful software or learning science and really ignore what others do. Perhaps this model is useful for them.
But it is not useful at all for the larger number of less productive, more escapist nerds in the local D&D or MtG club. (And the difference is not intelligence but personality type, interests and so on, there are people in the Mensa who are like this.)
Yeah, I do consider it armchair psychology. He seems to be generalizing too much about the psychological characteristics of the empirical cluster to which he refers as nerds. I found the interesting part of the essay to be the hypotheses about perverse incentives explaining the similarities between the milieux of American prisons and American public schools, and that’s really why I brought it up here.
I was under the impression that US schools have far more surveillance. I mean my school experience was like that, but it was a poor school at the ass end of Europe, no campus police no security guards no cameras no staff nothing, just teachers and kids, and teachers usually spending the breaks in their lounge, drinking coffee, so nothing kids did to each other was noticed and punished, pretty much free reign. I sort of expected US schools to be under far more surveillance. Prisons, too. There are no privacy rights in prison, none at all. If a country can afford it, there is supposed to be camera in every corner?
That depends. US schools are run by local (not even state) governments, usually towns, and there is huge diversity in what a school might look like.
Public schools in big cities (NYC, Philly, etc.) basically are prisons with wire fences, metal detectors, cameras everywhere, etc. I’ve seen a school with no windows—it was designed this way so that there were no windows to break. But schools in e.g. rich suburbs are nice places, similar to college campuses.
I don’t quite understand your point. At any rate, in my experience, I only encountered two surveillance cameras that were privately owned by tech-savvy teachers, there was usually one (sometimes two) ‘SRO’ (school resource officer), a police officer assigned to the campus, and we were almost always supervised by adults. Graham’s hypothesis is plausible even given poor supervision. The only requirements are 1) adults and children not being invested in the ostensible purpose of the educational system and 2a) the presence of obstacles to refusing to participate in or 2b) reforming the educational system.
There quite possibly is. Not that it stops prison gangs from dominating prisons.