[Guest post by Aaron Worthing; if you have tips, please send them here. Or by Twitter @AaronWorthing.]
Apparently Andrew Sullivan is not new to the concept of crackpot theories. Long before he went spelunking in Sarah Palin’s womb to explain the incongruous event of a woman over forty years old giving birth to a child with Down’s Syndrome (note: I am being sarcastic), he was exploring the racial differences in IQ. Groan.
I admit I didn’t know that back The Bell Curve came out, it was Andrew Sullivan’s decision to give a cover story to an essay by one of the authors summarizing its findings. Sullivan justified it by saying, “the notion that there might be resilient ethnic differences in intelligence is not, we believe, an inherently racist belief. It’s an empirical hypothesis that can be examined.”
Well, first, actually yeah, that is kind of is racist. The only question is really whether or not it is true. And there is nothing wrong with a serious scientific inquiry into the subject. We should never shrink away from a question because we are worried what the answer might be. But The Bell Curve wasn’t that kind of serious inquiry. It was plainly an attempt to dress up regular old racism with a scientific gloss, something bigots had been doing since eugenics was fashionable.
The Bell Curve of course is the book that asserted that black people, on average, had a lower intelligence than white people, or so that is what its authors claimed it proved. What it actually proved, precisely understood, is far less useful: that the people considered by a myriad of persons to be “black” tended to have lower IQ scores than people considered “white.” And when you state it like that, you start to see the hidden assumptions that completely undermine their analysis.
First, exactly who gets counted as a member of what race? For instance our president is half black, half white. So if he is in the group being measured, what do we write him down as? Indeed, an ugly reality is that there is a great deal of “white blood” in most Americans descended from slaves and not all of that was the result of consensual unions. Slave rape was a painful reality in plantation life. So how on earth do you even hope to control for those factors?
Second, IQ test scores do not necessarily equal intelligence. There are many kinds of intelligence that is in fact difficult to measure. I believe, for instance, my parents are of roughly equal intelligence, but my mother is the classic “book smart” kind of person, while my father is the classic “people smart” kind. Both kinds of intelligence are valuable, but only one is relatively easy to measure.
And there are accusations that there biases in the tests. Don’t scoff, I have actually seen it myself. In one IQ test I took I was asked at one point what the Koran was and answered it correctly. Later in the same test, they asked me what the book of Genesis was and I answered it correctly. So you see in that example I was asked a question of general knowledge about the holy book of Islam, but a specific question of the holy book of Christianity and Judaism. If I happened to be Muslim, I would get no credit for my specific knowledge of the Koran, and I would be more likely to get the question about Genesis wrong (although arguably, it is well enough known in this culture that I suspect many Muslim Americans would know the answer anyway). Clearly that question favors Christians and Jews over Muslims. Cultural bias does exist in this context.
The clearest example to me that The Bell Curve was working toward a pre-determined conclusion instead of simply following the science where it led was in their treatment of a trans-racial adoption. I made a study of the Bell Curve’s conclusion, although it has been a while, so I really dug into the data at the time. In that study, the scientists tracked the IQ scores of black children adopted by white families and what they showed was fairly remarkable. It showed that at first the black children had scores comparable to white children of like age, but then as time when on their IQ fell down to something more typical of black children raised by black parents. And the amazing thing is that the authors of The Bell Curve read that as confirming their racial theories. The black adoptees, they said, had their scores artificially raised by exposure to their white parents only to revert to the natural level of all black persons. Even accepting that theory, that still requires you to believe that the IQ scores can be significantly affected by something other than good genes—a concession that undermines everything else they claim. And further, the same data could be entirely explained by factors besides genes. It could simply be the case that when they are younger the black adoptees are unaware of race and racism, and as they grow older they become aware of both and become discouraged.
And Sullivan gave a forum to that crap!
All of this is just a lead in to Rand Simberg stepping in today to defend Sullivan from charges of racism. You see, Sullivan has decided to weigh in on the subject again (here, here, here and here—the Daily Beast must be so proud) and Gawker decided to translate choice lines from Sullivan’s post, here, and Simberg decided to step in and defend him, here, saying:
One wouldn’t have thought it possible, but I actually largely agree with Andrew Sullivan. The notion that intelligence is not heritable is ludicrous, and if it is, the notion that every “race” is going to be equivalent in that regard is equally so.
Well, Simberg’s very limited assertion is correct. First, I am pretty sure the average anthropologist believes it is a “no brainer” that intelligence is inherited. Intelligence—indeed specific kinds of intellectual talent—runs in families. If intelligence was not inherited, then it would not be possible for the species to evolve into a more intelligent state. The evolution of human intelligence (assuming you do not buy into a creationist theory—and I am not putting you down if you do) required individual humans to become smarter than others, to enjoy a competitive advantage in the struggle for survival because of that greater intelligence and most crucially, for that intelligence to be passed in some way down to their children.
And second, I am sure that however one defines race, the average IQ of each group doesn’t come down to being precisely equal to the 1/1,000,000,000 of an IQ point. I am sure that if you had a good measure of intelligence and race, that you would find that one group edges out the others. I likewise believe that more than likely the difference is insignificant and shifts depending on random variations of the current “crop” of children and the “crop” of elderly that had just passed on, so that one year white people might edge out black people and another black people might and so on. It’s not precise mathematical equality but it is probably close, and indeed too close to provide you any guidance when deciding who to hire or admit to a law school.
But what Sullivan seems to be saying is a lot more than that mild and hard-to-dispute claim that Simberg assigns to him and really Gawker does a disservice to this discussion by resorting to crude caricatures that leave you having to believe that Gawker is being unfair. They are, but Sullivan is wrong, too. For instance, in the November 23 post I linked to above, Sullivan writes:
Two points: research is not about helping people; it’s about finding out stuff. And I have long opposed the political chilling of free inquiry into any area of legitimate curiosity or research. I’m not going to stop now.
First, Glen Reynolds, call your office. I think we are seeing the higher education bubble right there. Mind you, in his November 21 post, he relays a complaint that because of PC concerns there has been “an exodus of researchers away from the area, and a drying up of grant funding and research positions for researchers interested in IQ.” The last two pieces of this complaint concern a failure to provide university support for research for its own sake regardless of its value to society. The concept of “return on your investment” has no meaning to him.
But of course Sullivan is stalking a greater prey: affirmative action, writing in the November 23 post that
Secondly, I agree that there would be very little, if any, use for this data in our society, apart from the existence of affirmative action. But when public policy holds that all racial difference in, say, college degrees, are due to racism, a truth claim has already been made. So the p.c. egalitarians have made this a public and social issue by a statement of fact they subsequently do not want to see debated or challenged using the data. That’s an illiberal position, in my view.
So what he is really hoping for is that this strikes down affirmative action. In this his obsession on race differences bears some resemblance to his Trig Trutherism. In both cases you get the feeling that this isn’t really about the truth or falsity of the specific claim, but rather what damage it will do to something else—affirmative action in the case of The Bell Curve and the destruction of any chance of a Sarah Palin presidency in the case of Trig Trutherism.
But in this Sullivan gets things precisely backwards. For instance, take law school admissions. I don’t know hardly a lawyer alive who thinks that the LSAT (more or less the Law School equivalent of the SAT) is a good measure of aptitude for the profession. And I say that as someone who scored very well on the exam: it measures abilities that have almost nothing to do with my job. And yet opponents of affirmative action argue that we should use this flawed instrument that happens to result in racial disparities blindly. Particularly when a state school uses a tool like this to deny people opportunities that happens to have a racial disparity in its results, the burden should rightly be on the state to show that the test really truly relates to the relevant abilities. And I don’t know too many lawyers who could defend that test as a measure of the aptitude of lawyers with a straight face.
The logic puzzles section in particular seems to represent some pinhead’s idea of what legal reasoning is like, rather than what lawyers actually use most of the time. Here’s an example of this kind of question:
A university library budget committee must reduce exactly five of eight areas of expenditure—G, L, M, N, P, R, S, and W—in accordance with the following conditions: If both G and S are reduced, W is also reduced. If N is reduced, neither R nor S is reduced. If P is reduced, L is not reduced. Of the three areas L, M, and R, exactly two are reduced.
If both M and R are reduced, which one of the following is a pair of areas neither of which could be reduced?
(A) G, L
(B) G, N
(C) L, N
(D) L, P
(E) P, S
In analyzing statutes, case law, in arguing before courts, lawyers simply don’t think, don’t reason, this way. It fits certain unsubtle stereotypes of legal thinking, but not the reality of it. And denying the opportunity to enter the legal profession to a disproportionate number of racial minorities or what have you based on this P.O.S. test is just plain bad science. That doesn’t mean using a crude tool to correct it like Affirmative Action is suddenly a good idea, but the choice doesn’t have to be limited to either 1) blindly accepting the perfection of the law school admissions process as it is, or 2) using crude racial tools to fix the problem. There can be a third way.
Sullivan also digs himself in deeper in his November 28 post, writing:
No one is arguing that “that black people are dumber than white,” just that the distribution of IQ is slightly different among different racial populations, and these differences also hold true for all broad racial groups[.]
Um, no, when you argue that the real average intelligence of black people is lower white people, then, yes, you are arguing that black people are dumber than whites, especially when you argue that the difference justifies the dismantling of affirmative action. Own it, Andrew.
And you see in that same post where the caricatures really do harm this debate. Responding to the sarcastic hyperbole by Ta-Nehisi Coates that “[m]aybe the sterilizers and the slave-traders were wise beyond their years” Sullivan counters that “I don’t think any serious critic of my work could conjure up a defense of compulsory sterilization or slavery within it.” No, Sullivan, what you can find is a lazy justification for things like a law school admissions process that is unscientific and has the net effect of disproportionately shutting out people of color from the legal profession. It’s not slavery and sterilization, but it’s still wrong. And the fact that Coates was over the top in his denunciation of you, doesn’t suddenly make you right.
On the bigger picture, I don’t think there is anything inherently wrong with exploring the relationship between race and IQ in a clinical, scientific matter. I would for instance enjoy seing a lot of study into trying to explain racial achievement gaps and finding ways to bridge it without resorting to tools as crude and divisive as affirmative action. But having watched these debates for a number of years, I have little confidence that such science can be done in such a dispassionate manner.
[Posted and authored by Aaron Worthing.]