= 27d567bc2d48d9f40e9ccc20ea370b15)
Eugine_Nier comments on [LINK] Why I'm not on the Rationalist Masterlist - Less Wrong Discussion
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (866)
But it might affect how rational you are.
It's possible.
Why are you bringing it up, though? As an aspiring rationalist, I believe it should be possible in principle to discuss whether one sex is more rational than the other, on average. However, it makes me feel uncomfortable that a considerable number of people here feel the need to inject the topic into a conversation where it's not really relevant. If I were a woman, I can imagine I would feel more hesitant to participate on Less Wrong as a result of this, and that would be a pity.
It's an interesting topic, the moreso because it is taboo, and not exactly tangential to the subject, I think.
It affects your argument that there is something wrong with having a skewed gender balance here.
Compare with Cosma Shalizi on the heritability of IQ (emphasis mine):
At this point I would have to conclude that the guy is either very deliberately blind or is lying through his teeth.
He, of course, knows very well what the consequences for his career and social life would be were he to admit the unspeakable.
What you & Anatoly_Vorobey have quoted is talking about heritable IQ differences between individuals ("who do not have significant developmental disorders"). Is it possible you're conflating that with talking about heritable IQ differences between races or sexes?
That you use the word "unspeakable" suggests you are, as does the fact that your two cases of scientists suffering career consequences (Gottfredson & Cattell) are cases where they suggested genetic racial differences as well as genetic individual differences. (In fact, if I remember rightly, both went further and inferred likely policy implications of genetic racial differences.)
That's a good point, I think the two issues got a bit conflated in the discussion here.
However I can't but see it as a reinforcement of my scepticism. My impression is that the partial heritability of IQ in individuals is well established. At most you can talk about doubting the evidence or not believing it or something like that. Shalizi says he "has no evidence" which is not credible at all.
Yes, I think it supports your dim view of what Shalizi wrote. I also think it detracts from your implication that he's simply evading saying the "unspeakable", since heritable IQ differences between individuals are a much less contentious topic than heritable racial (or sexual) IQ differences.
You're wrong.
First, about the consequences: the theatrics of the "unspeakable" are getting a little tiresome. Shalizi is a statistics professor at Carnegie-Mellon. The Mainstream Science on Intelligence was signed by 52 professors and included very clear statements about interracial IQ differences, lack of culture bias, and explicit heritability estimates. I would ask you to name the supposedly inescapable and grave "consequences for career and social life" these 52 professors brought on their heads.
Second, about the subject matter: this quote comes at the end of a long post in which Shalizi challenges the accepted estimates of IQ heritability, and criticizes at length the frequent but confused interpretation of heritability as lack of malleability. In his next post on the subject, he criticizes the notion of a single g factor as standing on a shaky ground, having been inferred by intelligence researchers on the basis of factor analysis that is known to statisticians to be inadequate for such a conclusion. Basically, Shalizi criticizes the statistical foundations employed by IQ researchers as being statistically unsound, and he carries out this critique on a much deeper technical level than what normally makes it into summaries, popular books and blog posts. On the face of it, this isn't a completely ridiculous idea: we know that much of psychology and medicine routinely misuses statistics in ways that make experts wince, although we might also expect IQ researchers to have their statistical shit together much more decisively than your average soft-psychology paper.
There have been replies to Shalizi's critique on the same technical level, and further debates. Frankly, most of this goes over my head. I know just about enough basic statistics to understand most of Shalizi's critique but not assess it intelligently on my own, and certainly not to follow the ensuing debate. I doubt, however, that your dismissal of Shalizi's honesty is based on a solid understanding of the arguments in this debate about statistical foundations of IQ research.
That flat and unconditional statement seems to be mismatched with your sentence a bit later:
Given that you say you lack the capability to "assess it intelligently on my own" and given that I don't see the basis on which you decide I am statistically incompetent, I am rather curious why did you decide that I am wrong. Especially given that I was talking about my personal conclusions and not stating a falsifiable fact about reality.
P.S. Oh, and the bit about consequences for career? Try Blits, Jan H. The silenced partner: Linda Gottfredson and the University of Delaware
You're wrong because your conclusion that Shalizi was either blind or lying rested on two premises: one, that heritability in racial IQ differences has been proven, and two, that for Shalizi to admit this fact would be uttering the "unspeakable" and would carry severe social and career-wise consequences. I wrote a detailed explanation about the way Shalizi challenges the first premise on statistical grounds, in the field where he's an expert (and in a way that's neither blind nor dishonest, albeit it could be wrong). I gave an example that illustrates that the second premise is wildly exaggerated, especially when applied to an academic such as Shalizi. That's why you are wrong.
Your response was to twist my words into a claim that you are "statistically incompetent", where in fact I emphasized that Shalizi's critique was on a deep technical level, and that I myself lacked knowledge to assess it. That is cheap emotional manipulation. You also cited a paper about Gottfredson that wasn't relevant to what I said. Given this unpromising situation, I'm sure you'll understand if I neglect to address further responses of that kind.
How could you possibly do that for a subject about which you said that "most of this goes over my head"?
Short memory, too. Your words: "I doubt, however, that your dismissal of Shalizi's honesty is based on a solid understanding of the arguments in this debate about statistical foundations of IQ research."
Oh, I'm the understanding kind :-P
That's a locked-up paper printed in a journal operated by a political advocacy group.
Linda Gottfredson doesn't seem to have been "silenced", though. (But I have a libertarian, rather than a left/right partisan, view on that concept. Someone who takes grants from wealthy ideological supporters instead of from government institutions is not thereby silenced; on the contrary, that would seem pretty darn liberating.)
The "Look Inside" button will give you the first two pages. I am not sure why the publisher of the journal is relevant unless you're going to claim the paper is an outright lie.
As reasonable as that person sounds, I feel the need to point out that IQ differences between race has little or nothing to do with IQ differences between sexes (and even less with rationality, but I guess we gravitated away from that). Even if there is a "stupid gene", to phrase it very dumbly, there is still no reason to believe that someone with 2 X chromosomes would inherit this gene while someone with the same parents but with a Y chromosome would not.
If you (or anyone) want to argue that women naturally have lower IQ than men, I would go with an argument based on hormones instead. Sounds much more plausible to me.
Where do you think the differences in hormone levels come from?
Food, genes, certain types of activity such as sports and competitiveness in general, the environment you grow up in, being in a position of authority, to name some factors that influence hormone production.
It's certainly not just the gender divide. If you think that testosterone makes men smarter than women on average, you would also have to accept the conclusion that women with more testosterone than men will be smarter than men on average. All other things being equal, of course.
Testosterone levels in men and women are in completely different ballparks, and there is no overlap in healthy individuals of the different sexes beyond puberty. This would make me think the difference is mainly genetic.
I'm not arguing for anything beyond this point, so we don't have to go there.
I stand corrected on the testosterone levels: The difference is indeed greater than I thought. I will accept that the difference is mainly, but certainly not solely, genetic.
You are absolutely correct on the facts, and in a saner world I could leave it at that, but you seem to have missed an unspoken part of the argument;
The common factor isn't genetics per se but rather an appeal to inherent nature. Whether that nature is the genetic legacy of selection for vastly different ancestral environments or due to the epigenetics of sexual dimorphism is very important in a scientific sense but not in the metaphysical sense of presenting a challenge to the ideals of "equality" or the "psychic unity of mankind."
When Dr Shalizi writes the rhetorical question "why it is so important to you that IQ be heritable and unchangeable?" in the context of "'human equality' and 'genetic identity'" his tone is not that of scientific skepticism of an unproven claim but rather an apologetic defense of an embattled creed. Really, why is it so important to you what the truth is? After all, we don't have any evidence to suggest that the doctrines are wrong, so why not just repeat the cant like everyone else? Who else but a heretic would feel need to ask uncomfortable questions?
For the most part, scientists writing against the hereditarian position don't bother debating the facts anymore; now that actual genetic evidence is starting to come out they know it'll just make them look foolish in a few years, and the psychometric evidence has survived four decades of concentrated attack already. It's all about implications and responsibility now, or in other words that the lie is too big to fail. It's hardly important to them if the truth at hand is a genetic or an hormonal inequality, they just want it to go away.
I read Shalizi differently, as asking something like, "Really, is it because you care about the truth qua truth that you find this particular alleged truth so important?" Far from apologetic, he is — cautiously, because there is a counterfactual gun to his head — going on the offensive, hinting that the people insistently disagreeing with him are motivated by more than unalloyed curiosity. It is not, of course, dispassionate scientific scepticism, but nor is it a defensive crouch.
My interpretation could be wrong. Shalizi isn't spelling things out in explicit, objective detail there. But my interpretation rings truer to my gut, and fits better with the fact that his peroration rounds off ten thousand words of blunt and occasionally snarky statistical critique.
I think you misinterpret Dr Shalizi, and do him a disservice. I think his answer is perfectly reasonable from a bayesian point of view. Basically, I see three common reasons to spend time researching difference between races:
A) People who are genuinely interested in the answer, for pragmatic or intellectual reasons
B) People who are a racist and want to hear a particular answer that fits their preconceived views
C) People who are trying to be controversial/contrarian/want to provoke people
Certainly there are people who are genuinely curious towards the answer, purely for intellectual reasons (A). I am somewhat interested myself. However, the fact of the matter is that many others are interested purely for racist reasons (B). Many racists aren't open in their racism, and as such mask their racism as honest scientific inquiry, making B indistinguishable from A. Showing interest in the subject is therefore Bayesian evidence for B as much as it is for A. Even worse is the fact that everyone knows that everyone realizes this on an intuitive level, which causes most As to shut up for fear of being identified as Bs, while Bs continue what they are doing. This serves to compound the effect. Meanwhile, Cs arise expressly because it is a hot button topic. As a result it is entirely rational to conclude that someone who is constantly yelling about race and inserting the subject into other conversations is more likely to be a racist on average than others. And of course, it's incredibly frustrating if you are an A and just want an honest conversation about the subject, which is now impossible (thanks, politics!).
I think Shalizi deals with this messed up situation admirably: Making clear what he believes while doing everything to avoid sounding controversial or giving fuel to racists. Of course this doesn't work very well because people who call others racist fall into two categories themselves:
D) People who are genuinely worried about the dangerous effects of racist claims.
E) People who realise they can win any argument by default by calling the other a racist
And people who fall under category E do not, of course, care about the truth of the matter in the slightest.
Kind of tempted to write a top-level post about this, now. Hmm...
I think that the fact that there is a debate and that the "good guys" use name-calling instead of scientific arguments, increases also the number of people in the group A.
It's a bit like telling people not to think of an elephant, and then justify it by saying that elephant-haters are most obsessed about elephants, therefore thinking of an elephant is an evidence of being an evil person. Well, as soon as told everyone not to think of an elephant, this stopped being true.
Actually, it is more like not being allowed to talk about the elephant (...in the room. See what I did there?). Not talking about a subject is much easier than not thinking about it. And because everybody knows that talking about the elephant will cause you to be called an elephant hater and nothing good whatsoever will come of it in 95% of cases, the only people who continue to talk about elephants are people who care so strongly about the subject that they are willing to be called an elephant-hater just so that they can be heard. So that leaves people who either really hate elephants, and people who really can't stand being told that they're not allowed to say something (and super-dedicated elephant scientists I guess, but there's not very many of those).
In the same sense that showing interest in medicine is Bayesian evidence for me wanting to poison my neighbors.
I'd say that the percentage of people showing interest in medicine that want to poison their neighbour is rather lower than the percentage of people talking about genetic differences between race being racist.
That depends on the definition of "racist" used.
Yes, Shalizi was talking about something completely different, but his attitude was similar to yours. He was saying: "sure, I could imagine that it might be so (that there might be a heritable difference), but why are you so invested in believing in that? Why do you fight for it so much?". I meant for my quotation to bolster your case.
Ahhhh, you're right, I completely misunderstood your intent. In that case we are in agreement.