[SEQ RERUN] Why Are Individual IQ Differences OK?
Today’s post, Why Are Individual IQ Differences OK? was originally published on 26 October 2007. A summary (taken from the LW wiki):
People act as though it is perfectly fine and normal for individuals to have differing levels of intelligence, but that it is absolutely horrible for one racial group to be more intelligent than another. Why should the two be considered any differently?
Discuss the post here (rather than in the comments to the original post).
This post is part of the Rerunning the Sequences series, where we’ll be going through Eliezer Yudkowsky’s old posts in order so that people who are interested can (re-)read and discuss them. The previous post was No One Knows What Science Doesn’t Know, and you can use the sequence_reruns tag or rss feed to follow the rest of the series.
Sequence reruns are a community-driven effort. You can participate by re-reading the sequence post, discussing it here, posting the next day’s sequence reruns post, or summarizing forthcoming articles on the wiki. Go here for more details, or to have meta discussions about the Rerunning the Sequences series.
I’ve noticed that even on Lesswrong, there is such a thing as knowledge that it is deemed better not to know. Apparently this is referred to as the basilisk’s gaze (I’ve yet to manage to read anything deemed dangerous here before it was deleted, so I could be wrong in the details of that).
It seems to me that a lot of the “Don’t suggest that there are racial differences in IQ” position is actually based on a hidden belief that looking at the possibility of racial differences is gazing at a basilisk.
Suppose you are an employer hiring for a position, using an examination where performance is correlated with intelligence. It is essentially harmless to take the position, “My prior is that whites have higher IQs on average than blacks, so I expect the average score of the white applicants to be higher than the average score of the black applicants.”
What the opponents of acknowledging racial differences are worried about is that the employer will also take the step of saying “This particular black applicant scored exceptionally well on the examination, but since I know that blacks in the aggregate have lower IQs, I’m going to treat my prior and the examination as separate bits of knowledge and scale my assessment of the candidate’s intelligence downward from what the exam alone would suggest.” As opposed to having the prior be swamped by the examination.
This is on top of (legitimately) expecting that the average person won’t understand the difference between the layman’s concept of “race” and the more scientifically rigorous concepts of “population” and “cohort.”
In the wider world, unlike on Lesswrong, openly coming out and saying “Considering this idea is like gazing at a basilisk” would end disasterously. So people go with “This idea is false” instead.
I’m not sure the conventional kind of basilisk qualifies as knowledge, as such: Langford’s original story was about an image that crashes onlookers’ brains through a defect in image processing, not through anything to do with verbal or logical parsing. Most other treatments of the concept have done the same, more or less, although there are some ambiguous ones (like the “Funniest Joke in the World” Python sketch).
There are various presentations of knowledge which clearly aren’t mind-safe (and presentation and context usually matter more than the content), but their danger generally comes in the form of bias and related issues like priming effects, which of course is rather well-traveled ground on this site. I think we can explain the effects of bringing up racial IQ differences and other politically sensitive ideas perfectly well within that framework, without having to invoke any more fundamental problems; in fact, we do.
Upvoted for bringing the Langford story to my attention; I was not aware of it.
I continue to believe my explanation accurately describes the approach of a significant proportion of highly educated individuals expressing anti-racial differences views, whether or not those individuals are even aware of the general concept of mindkilling.
This story: http://www.infinityplus.co.uk/stories/blit.htm ?
There’s a similar plotline in one of the Star Trek episodes with the Borg, where (IIRC) they discuss the morality of crashing all the Borg’s minds with some bug they’ve found in their information processing, decide not to, and then discover they already sort of did that by accident....
That’s the one I had in mind, yes.
The “basilisk” is of a very different character btw. It’s more of a game theoretic issue, sort of like how if you can’t understand the language the blackmailer is using to try to communicate you can’t be blackmailed.
Yes, I’ve yet to encounter a legitimate example but that’s the name for the hypothetical concept. ;)
What about those nasty shock sites you sometimes see on the ’net, or really sickening jokes/short stories where you wish you’d not seen/heard them?
They are not basilisks. Basilisks do real damage to you, they don’t make you nauseous. To get the general idea of the concept see RationalWiki. That illustrates what apparently constituted a basilisk for certain psychologically vulnerable individuals. Of course others would warn you that you shouldn’t discover what an alleged basilisk is on the off chance that it actually is one!
Guess I was going for stuff I wish I hadn’t seen rather than stuff that it would have actually been better not to know.
OK, I did a search on RationalWiki, and found this:http://rationalwiki.org/wiki/LessWrong#The_ugly which strikes me as odd.
Are there any useful links to stuff on this concept? I expect that being exposed to ideas that mess with your mind might be a good way to develop the ‘mental immune system’.
That particular event is a rather strange bit of early LW lore. It seems to have largely passed out of the site’s public consciousness now, but for a while it cast a long shadow: doing a site search on “forbidden topic” ought to give you an outline of the opinion surrounding it.
I’d advise against deliberately seeking out allegedly harmful knowledge in order to expose yourself to it: I’m aware of no particular evidence that the mind responds to memetic threats (as opposed to non-memetic stresses) by hardening itself against them, and with confirmation bias and group identification behavior in mind there are a number of reasons why that might not be the case. I should probably temper that by admitting I haven’t always followed my own advice here, though.
On the other hand, I do think there’s room for a more general theory of harmful knowledge. While some of the groundwork has been laid, and we have a few ad-hoc guidelines in place, we don’t yet have a good consensus on epistemic safety, as the comments on the World of Warcraft thread (to say nothing of this one!) demonstrate. About as close as I’ve seen anyone get is Nick Bostrom’s 2009 paper on information hazards, but it limits itself to typology. Contributing to such a theory might be a valuable thing to pursue, if you’re determined to risk your sanity.
As far as I can tell, people’s vulnerability to memetic hazards that drive some people but not others insane should be very predictable. Granted that there are problems with retrospectively changing one’s outlook to try and defend against some of them, it shouldn’t be too hard to test someone to see if they already have appropriate cached response defenses up without exposing them to the idea itself.
I don’t think I’d go as far as deliberately risking my sanity (such as it is).
So has knowledge that is harmful in more than specific situations been demonstrated to exist, or are you referring to theorising?
Depends what bounds you want to put on it. Basilisk-like knowledge (what the Bostrom paper calls a neuropsychological hazard) affecting the human cognitive architecture has not as far as I know been demonstrated to exist. Several other context-dependent but still fairly general informational hazards (ideological, for example) do clearly exist, though, and many of them seem poorly understood.
The forbidden topic in particular seems to belong to an interesting family of reflective hazards that hasn’t gotten much attention at all, although for the sake of local norms I’d rather not devote too much attention to it here.
Doh. Maybe I’m too tired so my brain is working less well than I’d hope, but I hadn’t noticed the link to the Bostrom paper there. I need to try to more carefully read through the stuff people say to me.
I’ll give the paper a read-through tomorrow.
[edit] I scanned the paper, but the tiny section on neuropsychological hazard seemed to tend toward the low-level (photosensitive epilepsy as one example), rather than the Lovecraftian (as I might have expected it to, if I had thought carefully about it, since I don’t place much credence in high-level ideas that could blow your mind that way)
Observe that in the recent crisis, blacks and hispanics had two or three times higher default rate, even when controlled for income and credit rating. So had bankers applied that policy, they would have been right. A protected minority candidate with the same apparent credit worthiness as a white candidate is far more likely to default.
La Griffe du Lion has claimed that the same is true in academic achievement—that blacks with the same IQ and GPA as whites have lower levels of achievement, though I do not recall what evidence he presented for this claim.
Herrnstein and Murray on the other hand claimed that controlling for IQ, blacks had similar levels of accomplishment, though I seem to recall they were controlling for IQ and intact family
All three claims could be simultaneously true if we suppose that accomplishment reflects IQ and character, and that assessing an IQ indicator alone is not sufficient to swamp one’s priors.
I am not convinced this situation is at all analogous. Consider the following three facts: 1) The geographical distribution of blacks, hispanics, and whites is not random—there is considerable segregation by race; 2)In the aggregate, blacks and hispanics have lower average credit rating than whites; 3)If your neighbor defaults/is foreclosed, your own property value falls.
This would suggest that more higher-credit minorities would get dragged down by their neighbors than would white homeowners with equal credit scores. But an individual’s intelligence is not dependent on the intelligence of his neighbor, at least not at remotely the strength of causation that his property value is related to the property value of his neighbor.
GPA is a good basis for comparison within a school, but it is not a good basis for comparison between schools, so this is not surprising, considering that the average majority-minority school is less well-resourced than the average majority-white school. (I would suggest that standardized measures like the SAT are better than GPA, but unfortunately that is a highly imperfect solution as well.
I would say that the claim “Blacks on average have worse character than whites” is far more dubious on a simple empirical level than “Blacks on average have lower IQ than whites” (and the precise determination of “character” is doing a lot of work in this formulation). This makes the situation worse, not better.
So instead of evidence that the bankers should have redlined members of certain groups, this then would be evidence that they should have redlined certain neighborhoods.
Which of these questions do you think would have served the banks better:
A)Will this applicant remain financially solvent if the average home in their neighborhood drops in value by 30%?
B)Will this applicant remain financially solvent if the average home owned by a black family drops in value by 30%?
I do not think it correct to term it redlining unless the answer is actually going to be “no” for any individual in a given neighborhood regardless of their financial position.
A person in a white neighborhood was substantially less likely to experience a thirty percent drop in value. (Compare East Palo Alto with Palo Alto west of the freeway.)
Homes in areas with large numbers of Hispanics and/or blacks, primarily those with large numbers of Hispanics had the largest proportion of foreclosures, and such neighborhoods had the most severe drops in price, for example Gilroy in California, so discriminating by neighborhood or race or both, regardless of the individual merits of the applicant, would have served the banks better than a race blind or neighborhood blind policy
Would you now?
Would you also arrange for your daughter to wait for you on Martin Luther King Boulevard?
Assume that Prismatic was making a claim about “worse” being different for each English speaker, and each of our models of others differ, and argue against that point instead.
So what your saying is that if it’s not safe for your daughter to wait in a certain neighborhood that doesn’t qualify as the neighborhoods residents having worse “character” for your and/or Prismatic’s definition of “character”?
Sam0345 asks:
Sam0345 is making the assumption that I am white. This happens to be correct (at least, as far as most people are concerned since sometime in the mid-20th century), but I don’t think there was anything in my analysis to justify that assumption.
Thus we can assume that Sam0345 thinks my putative daughter is white. The problem is, while it may be unsafe for my putative white daughter to wait on the corner of MLK Blvd, there are also majority-white neighborhoods where it would be unsafe for a black individual to linger (and I mean because they risk being assaulted, not because the police would harrass them). This makes “neighborhoods you wouldn’t want to linger in” a muddled proxy for “average character of the residents.”
Would statistically speaking your Black daughter be safer on Robert E. Lee Boulevard or on MLK Boulevard?
I am not making that assumption: Blacks are race realists—they know what neighborhoods are dangerous better than anyone. If you search twitter for racist references to recent violent incidents, most of the people complaining about the violence in explicitly racial terms are black.
Don’t be silly. If blacks were in danger of racist attack, you would have a better poster boy than Emmet Till. Till was not attacked by a white mob for being black, but by a husband for groping that husband’s wife, something that is apt to happen regardless of the race of groper and gropee.
Every day there are incidents where a black mob attacks a random white screaming racist epithets, indicating that the attack is motivated simply by whiteness. If the equivalent thing had ever happened to a black, that black would be the poster boy, not Till. Till was killed by a husband for making a pass at that husband’s wife, not by a white for being black, while every day whites are beaten and often killed purely for being white.
P(H|e)=P(H)P(e|H)/P(e), P(e|H)/P(e) ~ 1
ETA: what this means is that assuming I wouldn’t arrange for my daughter to wait for me on MLK blvd., that doesn’t mean MLK blvd. is worse than other places unless I would arrange for my daughter to wait for me at those places. So I am criticizing the form of the argument.
Above, Konkvistador at least phrases it properly. My problem with sam0345′s post is that he seems to sacrifice clarity for offensiveness.
This is unlike normal arguments over political correctness, in which all sides agree some tradeoff between positive values of communicating clearly, being correct, not making people feel offended, etc. is appropriate.
My problem with the above post is that it sacrifices accuracy for offensiveness, as if that were a positive value, rather than, say, a worthless thing, which might be the typical extreme anti-PC position.
I think it’s very hard for people to overcome their priors here even after getting contradictory evidence. Does being a Bayesian and thinking of them only as priors really work on all levels of your mind?
No, of course not! And that’s part of the problem—people don’t want to admit that they have certain implicit associations, so they try to signal the opposite, often at the expense of correctness. (E.g. for some employers, not appearing racist may be more important than selecting the best applicants.)
I would agree with your explanation.
Also, in the job example once you get to interview/test stage the observations should indeed clearly swamp out all priors based on what group the candidate belongs to. However earlier in the process (when sifting through thousands of similar resumes) could these priors still retain some importance?
Basically I would separate 2 types of discrimination:
(1) I will not hire a person from group B because I don’t like people from group B. Or I believe people from group B will almost certainly perform less well than people from group A.
(2) I know the prior distribution of job performance for groups A and B (A is higher on average). After taking into account my obervations (looking at a resume) about 1 candidate from each group, the posterior distribution indicates that the candidate from group A is expected to perform better. So I hire A. Had I ignored the prior I would have hired B.
(1) is sub-optimal clearly unacceptable. (2) seems theoretically optimal and appears to be used for many groupings, like [went to a top university] vs. [medium university—same gpa/experience]
However (2) is completely unacceptable for other groupings (like race). Possible explanations:
It has no impact anyway. For these groupings any differences in priors would be so tiny that they would immediately get overwhelmed by the slightest job application relevant info
These are groupings for which people have absolutely no control. It is unfair that top group B people need to systematically overcome this prior.
In practice no one will be able to apply this properly and everyone will end up amplifying priors and giving them way too much importance, so it is best to not go near it.
People don’t have control over their IQ either.
Yes they do. What the cannot do is increase their IQ by a significant amount. But there is a whole range of IQ over which they are free to choose. Approximately the range [default IQ + 5, minimum measurable IQ]. Beating your head against something should do the trick but excessive drug use is probably more fun.
Good point (acknowledging wedrifid’s caveat) but one could argue IQ is often directly relevant to job performance, whereas race is not (“discriminating” based on ability-to-do-the-job is probably ok, even if mostly genetic).
It seems that using factors that cause good/bad job performance is normal hiring procedure whereas using factors that only correlate with good/bad job performance is statistical discrimination (thx for the link Emile)
So using things like test scores, impressions from interviews, etc., is statistical discrimination?
hmmm. Yes that statement is probably not correct. I guess your examples are observations that correlate with factors that cause good/bad job performance. Why is it more acceptable? Maybe because the link is much clearer/ correlation is much stronger?
Because you’ve drilled as far as you can before making a determination.
(2) is statistical discrimination.
I can’t even imagine what sort of hell would break loose in my Politics class if I were to profess a belief merely in the possibility of measurable differences in intelligence between races. Any logic would be ignored, immediately branded as justification for a bigoted agenda. Politics truly is the mindkiller.
Same way that some people can talk about the statistical correlations between e.g. IQ and race, most other people nowadays have learned to correlate instead those people who so correlate these things with evil people who want to oppress other races.
And certainly that’s actually the rational thing to do. If you hear someone seek to correlate races with IQ, you ought adjust upwards the probabilty of them wanting to oppress other races. Because there is a positive correlation between people who speak about lower intelligence of blacks and evil people who so want to oppress other races.
No, it is a rationalization of what they happened to do.
I would think it’s usually both.
By a small amount I think it is.
Is there any research on whether people can achieve evidence-based prejudice, or if (as I’m inclined to suspect) they overshoot and overestimate the effects of differences?
Interesting question. I’ve generally assumed that this was basically just belief in belief and that people’s actions already basically take into account the data (since it is ridiculusly hard to not notice it poping up everywhere if you have even a bit of pattern recognition capability).
I think people are remarkably good at ignoring data in favor of socially reinforced ideas.
This is a good point. Though what CJ says about actions already reflecting the data seems to be a pretty reasonable observation. Perhaps people don’t notice the data, perhaps their actions and behaviours arise purley through social mobility and memetic evolution.
Take for example the upper middle class and compare their stated ideals with their actual lifestyles on say divorce and other issues. It is hard to miss that they often behave as if they only had belief in belief but organized their actual lives by a different model. Someone being slightly cynical migh call this class warfare or perhaps handicap signaling.
But there might be a different better explanation. If say a society promotes ideals that make it harder for people to stay in a certain socioeconomic neiche, the behaviour of people that remain in that neiche will still on average be pretty decently adapted to it (or at least more dapted to it than anyone else) even if they state, belive and truly aren’t using the now low status principles to organize their lives. You don’t always need the human brain to come up with rationalizations, Azathoth can do it for you!
This is where I do think something like the handicap principle comes back into play. Evolution’s beneficiaries are perhaps the people who in far mode most distain the rational self interested reasons or the easiest to grok rationalizations which recommend the course of action that actually keeps them in their neiche. Everyone notices what kind of people tend to have what kind of beliefs. Stabilising competence signaling ensues.
There’s research to suggest people are racist to begin with, so giving them evidence may not affect their predictions or expectations at all!
That’s a powerful one. Several of the Rationalist Bootcampers had similar reactions.
Was this part of a “lets try to kill the minds of the bootcampers and see how they handle it” or did the subject just come up?
Subject came up.
How did the subject come up? I have never ever heard this subject discussed outside of two contexts:
The question under discussion is “What can’t we say?”; or
One of the interlocutors feels it’s important to take civil rights away from people.
EDIT: To be absolutely clear, I’m certain the second thing is not true of any of the bootcampers, and I’m almost certain that racial IQ differences came up in conversation only as an example of something people don’t consider rationally, not as a subject of interest in its own right.
Group differences in average IQ are relevant in discussions regarding whether or not policies and institutions that were successful in some jurisdictions should be recommended in other jurisdictions where the people have different genetic backgrounds. They can also legitimately arise in discussions about the evolution of intelligence.
How well the relevant groups in those discussions match folk notions of race, is another question. Some aspects of our folk racial categories seem obviously silly in this regard—like how children of one white parent and one black parent are considered black.
I have never seen that happen ever.
Well, since this is the first reference in this thread to “What can’t we say?”, which of the commenters would you say “feels it’s important to take civil rights away from people”?
But seriously, you should get out of the habit of assuming sinister motives of people who disagree with you.
Ummm, Eliezer Yudkowsky’s post, on which this discussion is based, is about “What can’t we say?” ie. why can’t we say there are racial differences in IQ.
So this thread doesn’t seem to be evidence against Nisan’s statement.
I feel like I’ve been misunderstood somehow.
It comes up for people considering adopting kids.
Was anyone’s mind destroyed, or did people get over it?
Might want to do an intro to statistics at the end if day one, where the mass of each soda bottle ever produced by Pepsi and Coke is calculated. Then find the average bottle mass for each company.
Then wait.
I don’t see what you’re getting at with the Pepsi and Coke bottle thing—could you explain a bit?
It’s a test of the universal law that two random different things are never miraculously equal and never equal unless there is a spectacularly good reason. This applies even when there is a spectacularly good reason to think that they would be roughly equal, and also when summing and taking averages.
As a close analogy, consider the mass of each bottle to be the IQ of each person in a group, and the bottle types produced by each company to each comprise a group.
That’s not a universal law. A random partition of a large set of objects may well produce two sets in which the distribution of all properties is the same as in the original set. The same is true if the set is partitioned according to some property that doesn’t correlate with anything else.
The controversies on this issue are about whether certain properties that can be used to partition human populations do have correlations with various other relevant properties, what is the reason for these correlations if they do exist, and what should be their wider implications.
What I believe you meant to say is that the results of two different processes “are never miraculously equal and never equal unless there is a spectacularly good reason.”
If I wanted to raise this possibility for discussion, I would likely leave “race” out of the discussion.
I’d probably start out by raising the possibility of measurable differences in intelligence among individuals; if that were successful I’d move on to the notion that there might be other shared differential characteristics among high-intelligence and low-intelligence communities; if that were successful I’d move on to the question of what those other characteristics might be… for example, age, or childhood nutritional regimens, or geographic region of birth, or various other things.
If, instead of doing that, somebody starts out by privileging a hypothesis that “race” correlates with intelligence for some particular definition of “race” and framing the conversation in those terms, I’d want to know what leads them to privilege that hypothesis before I was willing to invest much in that discussion, in much the same way that if someone starts out by privileging a hypothesis that the Old Testament God created the world in seven days I’d be inclined to reject their conversational framing.
Part of the problem is that “race” isn’t really a biological classification, although people pretend it is. It is a folk taxonomy that appears to be based partly on phenotype (e.g. skin color) but also partly on socially constructed facts such as language, socioeconomic status, and religion. So if you want to talk about phenotype or ancestral origin, talk about these; race is at best a biased label for a disguised query.
In say the American context race does basically match ancestral origin in the first approximation. Say African Americans are a hybrid population of West Africans and North-Western Europeans. Or for example European Americans are mostly North Western Europeans mixed in with a bit of Southern and Eastern European. The Hispanic category (which isn’t even a race) is basically an euphemism for Mexican American (who the majority of US Hispanics are) which is because of economic differences between Mexico and the US and the pattern of migration basically a euphemism for Mestizo (which as a European-Native American hybrid population with a relatively uniform amount of admixture does make sense as a biological entity).
Not only that, those who have older roots on the continent may have gone through common selective pressures. To quote Gregory Cochran
Absolutely agreed, which is why I put the word in scare quotes in the first place.
Sure you can, just call them blondes or gingers, instead of that other explosive thing. You can reveal the truth at the end of the discussion.
I think the problem is that even though people in their heart of hearts know that the chance of IQ distributions being 100% equal between arbitrarily divided groups is impossibly low, we confuse the idea of accepting that with acting upon it. Distinguishing between individuals based on any arbitrary indicators is seen as discrimination. Strictly speaking, it is discrimination in the strictest sense of the word, but the word discrimination is indexed to something intrinsically wrong and immoral, in modern American English at the very least.
There is a standard answer to how “adding race makes it worse”, which this article doesn’t address at all. In simple and blunt terms, the standard answer runs as follows:
It’s bad that some people are dumb. However, given that there are dumb people, it’s okay to treat them like they’re dumb.
It’s bad to treat smart people like they’re dumb.
If race-based differences in intelligence exist, then people will treat even the smart people in a dumber-on-average race as though they were dumb. This is bad because of point 2 above, at least.
But, if only individual differences in intelligence exist*, then people will only treat dumb individuals as dumb. This is okay because of point 1 above.
Therefore, “adding race makes it worse”.
* It’s not at all clear that this is a coherent hypothesis.
This will happen anyway, in fact it will happen more often if relevant information is discarded. The difference is that the victims will no longer be correlated with race. Thus we are still left with the question of why adding race makes it worse.
Edit: Another way to express what I’m trying to say is that your argument, if it works, shows we should avoid using any data that’s correlated but not perfectly correlated with intelligence, e.g., test scores, grades, job performance, pretty much anything really.
Yes, that is a fatal flaw in the above argument. It “proves” way too much, namely that we should be disturbed by the idea that there is any observable property of people whatsoever whose correlation with intelligence is neither zero nor one.
I agree. My footnote was trying to get at the same problem. (Though I’m not sure that Eugine_Nier was making the same point.)
I was.
If humans were perfect reasoners, your objection would be valid. But people are irrational, and will tend to over-discriminate based on race if minor discrimination (based on the amount of info race does provide) is socially allowed.
Here’s a thought experiment: have a bunch of typical humans guess the intelligence of a group of subjects, knowing their test scores, grades, and job performance, but not race. Then have the same judges guess the intelligence of a second group of subjects, using all the above info plus race, and the fact that “on average, whites are smarter than blacks” without any quantitative data on by how much*. I consider it likely that the second group, despite having more information, will be less accurate than the first. Therefore, if race-based IQ differences exist, we should try to ignore them unless we know their magnitude and are confident in our own rationality.
“On average, whites are smarter than blacks” is all most people will remember from an article on race and IQ, and they won’t think to look up by how much.
What will happen at various degrees of minor discrimination not being socially allowed?
I don’t have any data to hand on this issue, but there is some optimal amount of disseminated info+set of social norms that maximizes people’s ability to correctly judge others’ merit. Ceteris paribus, let’s go with that.
I agree, but that means the optimum norms are a balance rather than to be maximally vigilant hunting down minor discrimination. Since the maximum is wrong, this means that complaints that something is the type of thing that should be socially disallowed are highly suspect.
Yes. I don’t mean that it should be totally disallowed. The optimal thing would be to weight it appropriately. However, it may be that there can’t be any other states in society other than “ban it/make things race-blind” and “allow it/don’t make things race-blind”.
Just to be clear, the argument I outlined (but did not endorse) is about why it would be worse if race-based IQ differences existed in fact. Note that the hypothesis in point 3 was “If race-based differences in intelligence exist...”, not, “If we explore the possibility of race-based differences in intelligence...”. The argument doesn’t conclude that we should ignore relevant information. The argument’s conclusion is that, in a “juster world”, racial information wouldn’t be relevant.
(It’s not clear to me whether you meant to imply otherwise, but I thought that I should clarify that point.)
Just to be clear, my objection applies to that version as well.
Could you elaborate on how you see your objection applying to that version? To be honest, I don’t yet see that the hypothesis in point 4 is coherent enough to judge whether your claim would be true of it.
I think that you’re saying that, in a world where fewer observable properties correlate with dumbness, there will be more false positives — i.e. more smart people falsely identified as dumb. Is that right?
That is correct. Notice that we could “simulate” such a world by simply ignoring some of the correlates.
But that isn’t true in general. It might be true under some additional plausible assumptions, but I haven’t worked out what those assumptions would be.
The following toy model is a counterexample. Suppose that intelligence is measured by a quantity between 0 and 1. People are paid according to their employer’s best guess of their intelligence. (We assume universal employment.) More precisely, the employer computes an expected intelligence E (between 0 and 1) for the employee and then pays that employee at a rate of E utilons-per-hour.
Define “dumb” to mean “intelligence less than 0.5“. Define “smart” to mean “intelligence greater than or equal to 0.5”. Define “treating a smart person as dumb” to mean “paying an employee at a rate less than 0.5 when that employee’s intelligence is greater than or equal to 0.5″.
Now consider the following two possible worlds. In both worlds, intelligence is distributed uniformly, in the sense that the proportion of individuals with intelligence between a and b is b − a. World 1 is a world with no observable correlate for intelligence. World 2 is a world that does have an observable correlate for intelligence. I claim that, in both worlds, half the people are paid below their intelligence, but, in World 2 alone, some smart people are treated as dumb.
In World 1, the employer has no information about the employee’s intelligence, beyond the uniform prior distribution. This yields an expected intelligence of E = 0.5 for each employee, so everyone is paid exactly 0.5 utilons-per-hour. Thus, in World 1, half the people are paid below their intelligence, but no smart people are treated as dumb.
In World 2, the population is split half-and-half into f-people and g-people. Employers know the actual distribution of intelligence among both sub-populations. An employer can identify an employee as an f-person or a g-person with perfect reliability, but the employer knows nothing else about that employee’s intelligence.
The f-people’s intelligence satisfies the distribution f, where
f(x) = 4⁄3 for 0 ≤ x < 1⁄2, and f(x) = 2⁄3 for 1⁄2 ≤ x ≤ 1.
Hence, f-people are dumber on average. If I computed correctly, the f-people have expected intelligence E = 5⁄12. Thus, the f-people are all paid 5⁄12 by their employers. In particular, some smart f-people are treated as dumb.
Meanwhile, the g-people’s intelligence is distributed according to the distribution g, where
g(x) = 2⁄3 for 0 ≤ x < 1⁄2, and g(x) = 4⁄3 for 1⁄2 ≤ x ≤ 1.
Hence, g-people are smarter on average. I compute an expected intelligence of E = 7⁄12 for the g-people.
If N is the total population size, our assumptions say that there are N/2 f-people and N/2 g-people. I compute that the number of f-people paid below their intelligence is 4N/18. I get that the number of g-people paid below their intelligence is 5N/18. Thus, in World 2, half the people are paid below their intelligence, but some smart people are treated as dumb.
Your scenarios implicitly assume that anyone whose expected intelligence is bellow median will get treated as dumb and that this is somehow much much worse then what happens to people whose expected is exactly median. Furthermore, even under this assumption you will find that your example falls apart if there is any way besides race to obtain information correlated with intelligence.
Well, actually, I thought that I made this assumption generously explicit. Evidently, you had implicit assumptions behind your claim that taking correlates into account would always lead to fewer false positives. What were these additional assumptions?
I did not make any assumption quantifying how much worse it is. It need only be marginally worse.
No. I can construct similar counterexamples where there are two observable properties (which you can think of as black/white, male/female), corresponding to four populations (black males, …, white females). You will need to make your assumptions more explicit if you want to rule out these kinds of counterexamples.
Sorry, I was assuming a utility function that summed over the amount of suffering each person experienced. You seem to be using a Rawls-style utility function base on minimizing the suffering of the worst of individual. (BTW, that’s a very stupid function to use in anything outside a very simple toy model.
Only by fiddling with the parameters very precisely.
If your assumption is that people whose expected intelligence is bellow the median (or really the nth percentile for any n) will be treated as dumb, the only way a counter-example like yours can work is by having lots of people exactly tied for the nth percentile. And the more other information is available the more the numbers in the scenario must be jiggered for that to happen.
You will need to make your assumptions more explicit if you want to rule out these kinds of counterexamples.
Your claim was that more correlates means fewer false positives. This is an abstract mathematical claim about epistemic probability. Utility functions don’t enter into it, at least not explicitly. It’s a claim about some class of probability distributions and criteria for categorization (“positives”). I’m just trying to figure out what class of distributions and criteria you’re talking about.
My counterexamples show that your claim doesn’t apply in full generality. You now claim that such counterexamples require “fiddling with the parameters very precisely.” I take this to be the claim that all scenarios satisfy your claim, except for some measure-zero subset (with respect to some natural measure). Can you prove this?
I’m not sure how to make sense of this. It doesn’t seem to reflect an understanding of my example.
I argued in the continuous limit. A measure-zero subset of people are tied for exactly the nth percentile. Recall that I said that “the proportion of individuals with intelligence between a and b is b − a.” So, the proportion of people whose intelligence is exactly tied for any value x is x − x = 0.
Of course, the continuous limit is only an approximation of the discrete reality. But I can find discrete examples where this proportion is arbitrarily small. It’s never “lots” relative to the size of the entire population, if that population is of any significant size.
I meant lots of people tied for the nth percentile in terms of your estimate of their intelligence, which was happening in your scenarios because the amount of information available was discrete and very small.
Okay, good. That makes a lot more sense.
What you say is true of the counterexamples I’ve described explicitly so far. But it is just an artifact of their being the simplest representatives of their family. I can construct similar counterexamples where the number of subpopulations in World 2 is arbitrarily large, and each subpopulation has a different expected intelligence. The proportion of people tied for any given expected intelligence can be arbitrarily small.
ETA: Also, these counterexamples work even if we redefine “treating smart people as dumb” to mean, “treating someone in the top 1% as if they were in the bottom 1%”. We still have a World 1 where no one smart is treated as dumb, and a World 2 where some smart people are treated as dumb.
I believe that the argument in my previous comment applies to any case satisfying the following.
Assume again that intelligence is measured by a quantity between 0 and 1. Assume there are two worlds, both with a prior distribution p for intelligence applying to the entire population. Furthermore, in the second world, the total population is divided into two equal sub-populations, f-people and g-people, with respective posterior distributions f and g for their intelligence. Assume the following about these distributions:
The support of each distribution p, f, and g is the entire interval [0,1]. That is, these distributions are all nonzero over the entire interval.
The prior distribution p is symmetric about 0.5. That is, p(x) = p(1 − x) for 0 ≤ x ≤ 1.
The distributions f and g are mirror images of each other about x=0.5. That is, f(x) = g(1 − x) for 0 ≤ x ≤ 1.
The expected intelligence for the f-people is below 0.5.
I believe that these assumptions suffice for the conclusions in my previous comment to follow. That is, in both worlds, exactly half the people are paid below their intelligence. But, in World 2 alone, some smart people are treated as dumb.
(Here I use the definitions from my previous comment, which I repeat here for convenience: Each employer computes an expected intelligence E for an employee and then pays that employee at a rate of E utilons-per-hour. “Dumb” means “intelligence less than 0.5“. “Smart” means “intelligence greater than or equal to 0.5”. Finally, “treating a smart person as dumb” means “paying an employee at a rate less than 0.5 when that employee’s intelligence is greater than or equal to 0.5″.)
Individual IQ differences are, in general “not okay”; racial IQ differences are downright verboten. I won’t discuss either in certain company for fear of attracting any number of labels, with the exception of the effect of lead on IQ, which is a soapbox I mount often.
As ArisKastaris points out, those labels should adhere to you more often than not. I tend to think that this is because the rest of us have never developed a decent realm of discussion which includes IQ. I get the same feeling with the “not everybody should attend college” crowd, since it too often seems to consist of high status inviduals from elite schools vastly overstating their case. That is, signaling their status. (In fact, unless you’re super interested in becoming a professor, you probably shouldn’t attend college in the same way non-priests shouldn’t have taken up seminary.)
Well, really, there are different classes of people who should attend college: would-be professors, poor autodidacts, akrasics, people who need separation from society for a few years. Probably others. The irony is that the people I thnk should be least interested in college are often those that are most interested in explaining why they but not others should attend.
I’m not confident I understand what you are saying.
Not confdent about which part? Skeptics or college-as-seminary?
I don’t think I understand either (illusion of transparency and all that). You seem to be saying that people from elite schools say not everybody should attend college, whereas in fact, you shouldn’t attend college unless you want to become a professor. (That after reading it five times I’m still not sure of what you meant is not a good sign)
Less clever analogies, more clarity please!
If you want to be an engineer, or have any other career that requires technical education, going to college is generally a pretty good idea.
Which would seem to make the entire metric either worthless or ‘bad’. :P
College is a good idea if you’ll have more job opportunities and get paid more for having gone. In other words, before you can convince me that I don’t need a college degree you’ll need to convince a few million hiring managers in my field of the same thing. The fact that their opinion may be poorly supported by evidence doesn’t change anything for me.
Downvoted for motivating an aggressive sort of skepticism: you’ve denied context agreement and therefore sent us straight to the pyrrhonic depths.
I don’t understand
1) What your position is the value on going to college in the top comment (and from lessdazed’s comment, I’m not the only one)
2) Whether you disagree with jhuffman
3) Why you downvoted (context agreement? aggressive skepticism? pyrrhonic dephts? what?)
Either you’re overestimating how much other people understand what you wanted to say, or I’m particularly stupid.
(1) It is a bad idea for everyone to go to college, at least as college is currently (4 years, etc.). College is foremost a technology for learning; it has advantages and disadvantages. If you need the advantages of this model, then go. However, it’s a well-known fact, at least in the arts, that it is not ideal; that field also contains “schools” and “institutes” with differing educational models and environments.
The problem with Huffman is he has decided here to break with the plain meaning of my statement within the context of the debate; when people discuss this topic, they do not track through adversarial dynamics involving job markets. Instead, you cut straight through to the optimal outcome where if college is not good for everybody, then it is something we shouldn’t demand, either, unless it really is the only/best source of a skill set.
(2) I don’t, but his comment is also an irrelevant extension of what we’re discussing. It’s as if I was trying to model the orbit of Mars well enough to find it with a telescope only to have someone criticize that Newtonian mechanics is superseded by relativistic mechanics. It’s true, I agree, but it is not important to what I’m doing and just makes things unnecessarily complicated therefore. This habit is common amongst analytical people, it should be guarded against.
(3) Context agreement is where we establish a limited domain of possibilities before proceeding. This is why when I talk say “endian” in a programming course, I don’t need to worry much that a hand will shoot up to ask “do you mean Native Americans or people of the Subcontinent?” In conversation, it limits confusion; in argument, it prevents global skepticism because when I say “I know I’m in Los Angeles” we agree that we’re talking in a “naive” sense and there’s no need to interject with “but how do you know you know?”
When we break context agreement in an argument, we must constantly and hopelessly reconstruct justifications. These are the pyrrhonic depths. After the sort of skepticism. I consider this to be “aggressive” in that it rapidly makes conversation unworkable; other types just create hidden problems we can safely ignore or introduce avenues we need, albeit with potholes we’ll need to grin and bear for a bit. “Aggressive skepticism” simply opens a sinkhole which swallows the whole town.
This is a good explanation but I think your comments on college were extraneous to begin with—that itself is reason enough not to respond to them but I don’t really agree we had clear context for the discussion.
If you want to talk about whats wrong with the expectations that individuals and businesses have of college its odd to start by singling out classes of people that “shouldn’t go to college” in a hypothetical world where it didn’t have the present instrumental value.
Depends, you have to make sure that the extra pay is enough to offset the cost of your student loans.
Mormons believe that you existed in a pre-mortal life. I don’t know if it’s doctrine or not, but it’s commonly believed that your conditions here are chosen based on how you acted in the pre-mortal life. I’ve been told that the most faithful were given disabilities. I’m not sure if someone has to be disabled for the ineffable plan to work or what, but they don’t seem to focus that much on that part.
I have asked this to a Mormon friend and a guy married to a Mormon, and they said that the reason it is considered a blessing to be born with stuff like Down’s Syndrome is that you are incapable of sinning, because sin requires knowledge of good and evil. So you basically have a free pass to the top levels of Heaven. So yes, to a Mormon, being born retarded is a great thing, while being born with dark skin is a terrible punishment.
But, knowledge of good and evil was supposed to be a gift. It was the entire point. It can’t be a gift to have it and a gift to not have it.
Also, if it was better to be born that way, wouldn’t everyone be born that way?
I’ll have to ask them about that. My understanding of Genesis is likely completely different from the LDS interpretation, but I thought that this knowledge was forbidden to humans, which was the catalyst for being banished from paradise and forced to suffer. If only they ate from the Tree of Immortality first! Rational Adam and Eve could have been gods!
According to Mormon doctrine, it was supposed to happen that way. I’m not sure why God told them not to eat it.
There they go again… as soon as you posit an omniscient character in a story, you run headlong into contradiction.
What contradiction? I understand that in most Christian churches, they weren’t supposed to eat from the tree of life, but according to Mormons, they were. If they somehow didn’t, that would have thrown off the plan. Or am I misunderstanding you?
Sorry, I was a bit vague. I was thinking about how claiming it is part of the plan was a dodge around questioning why god would be surprised of this outcome, but that it throws any claims of benevolence under the bus. But also how god often had these tests for people (Abraham, Job, Moses), and that were he omniscient, this would be either stupid or evil to do. I suppose were I to be more precise with that quip, I should say that any character that is omniscient will run into contradiction if you posit them to have just about any additional abilities.
Do you know if that leads Mormon communities to take better care of their disabled than other groups? In some parts of the US, abuse of the mentally disabled is endemic in the institutions intended to care for them.
I’m not sure what you meant to imply by the second sentence of this comment, but the worse non-Mormons treat the disabled the less interesting a same-sized gap in treatment by Mormons and others is.
If most are good and Mormons are great, we can look to them for lessons. If most are awful and Mormons are bad but better, they are less relevant.
Most non-Mormon institutions range from acceptable to awful. Mormons, with their particular respect for the disabled and their ability to coordinate for the good of the community, may have actually good solutions we can learn from.
In short, I think non-Mormon and Mormon institutions may be on opposite sides of the line between “good” and “bad.”
OK, great, I was just confused by “better”.
Strange name to call it. One would think all life is essentially pre-mortal.
In some religions that was a past life on Earth. Having it be a pre-mortal life seems like an extravagance.
This makes the essay a self-refuting argument.
It could be that your mother was on crack when pregnant, but you are not on crack.
I kid, I kid.
Sorry if I missed this in the coments already here, but BOTH of these statements are true: 1) White people are smarter than black people 2) Black people are smarter than white people.
In The Bell Curve there are reams of data on these questions leading to the conclusion that there is a difference in the mean intelligence of black and white Americans that is not accounted for by differences in education, income, or a variety of other things. But the other overwhelming fact is that the distributions of intelligence (measured by IQ in that book) were highly overlapping. So for the vast bulk of us, we could find people who were smarter than us be they white or black and be we white or black.
I think the provocation of the statement comes from our bias against thinking in terms of distributions. Who cares if women are smarter than men on average as long as I am smarter than my wife. Who cares if white people are smarter than black people if the black guy down the hall is kicking my butt at analyzing things at work.
Another remarkable component of the original EY post is the equation of genetic intelligence among humans with “fairness” or “justice” to the point that EY asserts god is a criminal if she has created us ith different intelligences. (Presumably a hypothetical as presumably EY doesn’t believe god created us at all anymore?)
What a narrow view of fairness! I suspect the richest people in the world have small overlap with the most intelligent. I suspect the most popular people in the world have small overlap with the most intelligent. I am virtually positive that the most laid (sexually) people in the world have little overlap with the most intelligent. As far as concerns along the “problem of evil” dimension are used to convert people to atheism, we might want to stick with the problems of young children treated absolutely horribly by sadistic perverts rather than much narrower injustice of being born to get a B in history and have to attend a public college.
Downvoted: You seem rather confused in your thinking—and you seem to be projecting some confusion to the rest of us as well. I don’t think anyone here has confused the meanings of the sentence “Group A has lower average intelligence than Group B” and the sentence “All members of group A have lower intelligence than any member of Group B”—that’s a confusion which you spend the three first paragraphs needlessly disentangling for the rest of us.
As for your last two paragraph, you seem to be thinking that for Eliezer to mention one example of the universe’s unfairness, means that he necessarily considers it the prime example of said unfairness. Hardly. It just came up, because another person mentioned differences in intelligence as a sign of an unjust God—but they limited said thinking to only racial differences, not individual differences.