Monday, May 13, 2013

intellectual lineage and defeating ideas

Ta-Nehisi Coates point out that, despite the common claim that research about race and intelligence is verboten, it has a long lineage at institutions that were and are considered prestigious and mainstream. One thing that struck me was Coates invoking Karl Pearson's indefensible stance as an explicit eugenicist and racist. It's a really shameful history, one which deserves telling, and also one which should open up some uncomfortable questions about our intellectual heritage, or heritages.

Let's be clear: Karl Pearson was not a statistician and a eugenicist. He was the statistician, and a eugenicist. It's hard to name a figure who has had more impact on his or her particular field than Pearson had on statistics. It's similar to the impact of Newton on physics or Chomsky on linguistics. If you've ever used the word "correlate," you are working in Pearson's lineage. If you've taken any statistics classes, whether a freshman intro or a doctoral seminar, you've worked in frames that were developed by Karl Pearson. In a very real way, if you are a modern human who understands the world, in part, through quantitative or statistical means, you are in the lineage of Karl Pearson. From educational research to climatology to baseball to economics, Pearson's influence is everywhere.

Now the simplest thing to say in regards to that is, Wagner wrote some good tunes. It's a banal matter of human life: brilliance and morality are separate attributes, and I'm afraid that the former does not require the latter. The reality is that you cannot pursue any intellectual lineage very far before you find that the progenitors of necessary ideas held some other views that were deeply wrong and deeply morally problematic as well. Do we toss out the good along with the bad? No. We utilize the good for the good and we rebut and reject the bad. Pearson advocated for white supremacy, and now we can use his techniques to show why we should work harder to solve racial inequalities that produce injustice. Be honest about the ugly stuff while you utilize the positive stuff. That goes for Karl Pearson and for Thomas Jefferson and for Jesus and for Aristotle. You don't need to reject bad lineages entirely because you can rebut bad ideas.

When I look to the comments of that Ta-Nehisi Coates post, I feel a little despair. Because you have good, well-meaning people who appear to be working furiously to confirm the frame that the "race scientists" prefer. 

Here's a typical example, from commenter Barry_D:
The idea that certain immigrant groups have lower IQ's and that this is inherent/genetic and won't change is, as anybody should know, very old and very wrong in US history. And so far it's been dead f-ing wrong.
There are three separate claims here, that should be considered separately. That IQ is inherent or genetic is deeply controversial, and asserting it as fact simply does not bear scrutiny. To say that it won't change is to be even more wrong. To say that certain groups have scored lower on IQ tests is not wrong. It is actually a fact acknowledge as true in a broad range of scholarly disciplines by a broad collection of researchers, many of whom are political leftists, some of whom are nonwhite themselves, and almost none of whom think that these differences mean that races are inherently superior or inferior to each other.

Let's look at a different intellectual lineage, that of the APA report I cited the other day. I really want to endorse it again, because it say a lot of smart things: about how group tendencies can never be determinative of individual outcomes when statistics are involved; about how race is at once an artificial construct which we nevertheless identify socially (including the way in which people self-identify into racial categories); about outcome bias; and about the evidence that (at the time the report was published, I'm afraid it's quite old) the racial IQ gap was shrinking. All of that is interesting and important and, I think, a model for those like me who dispute the notion of scientific racial inferiority. But the report also speaks plainly and straightforwardly: the results of IQ tests across several decades, in many forms, conducted by many researchers on many groups and in many contexts, have found consistent and statistically significant difference between racial groups in IQ. That's not what I'm saying, that's what the APA's literature review is saying. It could not come as a surprise to anyone who has studied these issues for any length of time.

Once again: this is not an endorsement of the conclusions of those who call themselves race scientists. Quite the opposite. As I have said, I find that IQ is not a reasonable or useful proxy for what we casually think of as intelligence, that intelligence is a multivariate and complex phenomenon that exists along many different axes which do not scale adequately onto simply metrics, that these tests exist in a sociocultural and economic landscape which cannot be corrected for through simplistic stratification of income level.... But none of that changes the accuracy in reporting the findings.

So that lineage. The report, which was commissioned by the APA's specifically as a response to The Bell Curve and its surrounding controversy, has eleven authors. They came from Emory University, the Educational Testing Service, the University of Minnesota, Howard University, Wesleyan University, Cornell University, California State University, the University of Pittsburgh, the University of Texas, the University of North Florida, and Yale University. My question to the people commenting at Coates's blog is simple: are the authors racist for reporting that finding? Is the APA a racist organization? Is Howard a hothouse of racism? Is Wesleyan a fever swamp of conservatism? Remember, this was not original research. This was a literature review, one that looked at the broad swath of research on this topic. It's not an exaggeration to say that if the opinion that there are observable racial differences in IQ test results is racist, then essentially the fields of modern psychology, education, anthropology, and assorted others are racist.

All of that is precisely why the space must exist to spell out why the simplistic assertion of racial inferiority via IQ tests is wrong, to explain how much difference there is between reporting the observed scores and accepting them as dispositive evidence for the relative intelligence of individual members of difference races. It is precisely because the research is not being undertaken merely by fringe cranks that its findings must be discussed and criticized carefully.

We have no problem with this when it comes to educational statistics. Nobody mistakes the fact that Hispanic students drop out of high school at far higher rates than white students as evidence of inherent and genetic inferiority, and nobody mistakes social scientists accurately reporting that statistic as somehow being indicative of fringe views and cryptoracism. I don't know why IQ provokes such profound misunderstanding on either side.

One commenter seems to have a straightforward take. MindOverMatter writes, "Agreed. It make one wonder the kind of literature review that goes into the 'scientific research' conducted by these 'academics' (racist hacks)". I wonder if s/he is aware of how many people s/he is indicted with that comment. Others endorse Stephen Jay Gould's The Mismeasure of  Man, which has been thoroughly dismantled. Also, there's an interesting exchange between commenters John Strider, Justin van Wormer, and DoctorJay. Van Wormer says, "Is anyone making the argument that the data doesn't "say what is says?" The correlation is there. My understanding is that the discredit comes from the fact that the variables being measured are not rigorously defined and that the correlation is therefore essentially meaningless as an empirical finding." That is more or less my own position. But it is strange to see van Wormer saying that; there are many people in the comments who plainly don't think that the correlation is there, and who find merely discussing it racist. Like MindOverMatter above. The distinction is hugely important, and yet there is no indication among most of the people commenting that they are willing to consider it.

Most discouraging for me, though, is the moderator Sandy Young. She has deleted some comments and admonished others with threats of banning. Maybe those actions were perfectly correct, I don't know, what with the comments being gone. But when she says stuff like this, "This is my line - it may or may not be the party line. To steal a phrase from a writer I admire: we will not debate people's humanity. There's a whole big internet where you can do that. And you are right, I have no intention of being 'neutral' about this," she is directly contributing to the cause of race scientists. She simply could not do a better job of fulfilling all of their presumptions, could not do more to work within their frame. 

I will repeat: the biggest effort that people like Charles Murray and Richwine undertake is not to prove that black and Hispanic people are inferior. The biggest effort they undertake is to claim that liberals are incapable of defending their views on the merits, that they forbid discussion rather than win debate, and that the "PC police" squelches dissent it can't rebut. Indeed, that frame is precisely what Coates is trying to fight against. When Young says that she does not want to hear these arguments and that she cannot be neutral about it, she is saying precisely what her opponents want her to say.

Coates and his commenters have often invoked their perfectly reasonable right to not engage with arguments they don't want to. It's part of the burden of racism that people of color are constantly in a position where they are asked to represent their races, and constantly in a position where they have to either rebut racism or let it go. I cannot imagine how aggravating and annoying it is, or how tiring it must be to constantly have to make the same fucking arguments to the same fucking idiots. Coates (and The Atlantic) can have whatever commenting policy they want. Neither Coates nor Young has a responsibility to undertake the kind of rebuttal I'm calling for.

What I am saying, though, is that someone has to make that rebuttal. Because as much as I might want to side with Young in saying that there are some arguments which shouldn't have to be rebutted neutrally or dispassionately, I'm afraid those standards are so deeply ingrained in our society that to refuse to respond to the social scientific arguments for racial inferiority with social scientific rebuttal is to forfeit and hand them the appearance of victory. And whoever does make the case has to make it as strong as possible. Don't do things like reference The Mismeasure of Man, which is the lowest of low-hanging fruit, the rebuttal of which just gives ammunition to the people who are wrong on the merits.

I believe that the case for scientific racial inferiority is wrong, and I think that the only way to ensure that it doesn't spread is to argue against it, and I think that the only way to argue against it is to be willing to look at the arguments in favor of it, without fear of what you might find.

15 comments:

Andrew said...

Even saying "This sort of study could hypothetically be done in a non-racist manner" is a long way from saying "this study was done in a non-racist manner". The racism in Jason Richwine's study is evident from the thesis statement.

In particular, the fact is that certain racial categories, particularly "hispanic" and "black" are basically useless from a genetic perspective. African populations are massively diverse, such that different populations on the continent are as genetically distinct from one another as they are from European or Asian populations, and "hispanic" includes at the very least white, mestizo and black populations. If you ever look at a study that uses race data in a legitimate fashion (eg, studying a heritable disease), they'll be very distinct about which African population they're sampling, and they'll basically never use "hispanic" as a category.

Use of "hispanic" and "genetics" in the same sentence (and Richwine does, Matthews pulls a line from his thesis) in particular should be a dead giveaway that you're not dealing with any sort of rigorous study. You literally don't have to know anything else to start laughing a guy like Richwine out of the room.

Freddie said...

I agree, entirely.

lou said...

You already know too much.

pigmundia said...

I really appreciate the methodical way you lay out your arguments. I have only recently begun to read your blog and listen to you on Bloggingheads, as a result of the conversation you had with Friedersdorf after the horriffic deaths of the workers in Bangladesh. For most of that conversation you and Friedersdorf had a pretty civil discussion. But that final ten minutes was pretty hostile when you took issue with Friedersdorf presuming the good faith of people who argue in support of different labor laws for third world countries. Your reaction to Friedersdorf on that issue seems to me to be on par with TNC's reaction to "racecraft". I can only imagine that for a blogger like TNC who really converses with his audience, to constantly refute the same tired arguments would be toxic to his health and career.

Q said...

When discussing the commenting policy and argumentative approach of one of the only mainstream, race-conscious POC blogs in existence about one of the most noxious, recurring racist debates in the past couple of decades, I wish your thinking would have more clearly started here:

"Coates and his commenters have often invoked their perfectly reasonable right to not engage with arguments they don't want to. It's part of the burden of racism that people of color are constantly in a position where they are asked to represent their races, and constantly in a position where they have to either rebut racism or let it go. I cannot imagine how aggravating and annoying it is, or how tiring it must be to constantly have to make the same fucking arguments to the same fucking idiots."

Instead of here:

"When I look to the comments of that Ta-Nehisi Coates post, I feel a little despair. Because you have good, well-meaning people who appear to be working furiously to confirm the frame that the "race scientists" prefer."

I'd like think your post and perhaps even your conclusions about how some liberals/POC's choose to discuss this topic would have gone differently if it had. I also think the disconnect between those two observations - themselves an example of clear argumentative choices and moral/emotional prioritization - are illustrative of my recurring issues with your writing on this topic.

Q said...

I'd like to think*, of course.

Brian said...

First of all Sandy Young is a he. Second, the rules of the blog are clear. You can't attack a moderator for following the rules of the blog.

The idea that "Blacks" "Whites" or "Hispanics" are anything other than socially constructed groups has no basis. Anything that attempts to make genetic inherency claims about any of these groups is ridiculous. This line of study is bogus because there are no significant genetic similarities within the groups, and there are no boundaries between the groups with any statistical significance.

The moral problem that you are not seeing is that even if you want to study populations, the results of these studies, all of them, have been used to make policies that affect individuals. The purpose of this line of study is to make policy affecting individuals. There is no other legitimate basis for this line of study.

imnotherzog said...

Andrew,

Of course, you don't know what you are talking about. Race, from a "genetic perspective" is a perfectly fine concept, including the use of Hispanic:

http://www.ln.edu.hk/philoso/staff/sesardic/Race2.pdf

AND

http://infoproc.blogspot.com/2008/11/human-genetic-variation-fst-and.html

You can deny the science all you want, but Richwine's data was correct:

http://westhunt.wordpress.com/2012/12/31/generations-of-exclusion/

Fred Smith said...

I sincerely don't understand the claim that the races are a "cultural" construct with no biological significance.

What we call "races" are groups of people with physical characteristics similar to each other and distinct from other groups, those characteristics themselves being passed down genetically.

Right? I mean, I feel like a real idiot sometimes reading a debate like this. Something like what I've written above seems so obvious to me, but very smart people are saying its wrong.

Barry said...

" Here's a typical example, from commenter Barry_D:

The idea that certain immigrant groups have lower IQ's and that this is inherent/genetic and won't change is, as anybody should know, very old and very wrong in US history. And so far it's been dead f-ing wrong."

I'm Barry_D. Please read Sowell's rebuttal to 'The Bell Curve'; he lists numerous examples of convergence of groups.


"One commenter seems to have a straightforward take. MindOverMatter writes, "Agreed. It make one wonder the kind of literature review that goes into the 'scientific research' conducted by these 'academics' (racist hacks)". I wonder if s/he is aware of how many people s/he is indicted with that comment."

Who cares how many. It's pretty clear that Harvard is a corrupt institution, having some excellent people but a large boatload of hacks.

Andrew said...

imnotherzog, neither of your first two links says a single thing about "hispanic" as a racial category, nor does the third link say anything about genetics. Certainly, any categorization that includes Marc Gasol, David Ortiz and Javier Hernandez (as "hispanic" does) isn't going to be a useful categorization.

Your second link says that the author can find genetic clusters among East Asians, and Nigerians specifically. Note that I specifically singled out "black" as a useless racial construct - Nigerians will genetically cluster with other ethnically similar Nigerians, but if the survey had used a wide variety of black Africans you'd see two clusters and then a spread across the plot.

You can find genetic differences between different racial categories, if you're specific and careful enough - no geneticist will ever deny that. No rigorous amount of care, however, will ever include "hispanic" as a racial category because it's such a genetically heterologous group of people. Which is why Richwine is a hack.

(Additionally, all of Freddie's critiques of IQ as a measurement are relevant but not really my area of expertise)

Brian said...

Fred Smith-

"I sincerely don't understand the claim that the races are a "cultural" construct with no biological significance."

To have biological significance, you h would have to show that there is a genetically distinct group that correlates with skin color.

http://onlinelibrary.wiley.com/doi/10.1525/aa.1997.99.3.534/abstract;jsessionid=720D4EB8C1610F3E8A1D81DD98E2B12B.d01t03

Modern Genetic studies, like where you go in and study the actual genes of the different groups, show that skin color and facial characteristics have no correlation at all with genetically similar or different groups. Dark skin correllates with living near the equator for 60-100 generation or more. That's it.


"Black" people in Africa exhibit more genetic diversity than all other people outside of africa combined. It's an overly broad construct. When we discuss African AMericans, there is so much genetic material from both European groups and African groups in both "Black: and "White" Americans, there is no way to tell from a genetic profile how that person will present in American society.

WHat seems obvious is not backed up by the science.

Brian said...

Fred Smith-

"I sincerely don't understand the claim that the races are a "cultural" construct with no biological significance."

To have biological significance, you h would have to show that there is a genetically distinct group that correlates with skin color.

http://onlinelibrary.wiley.com/doi/10.1525/aa.1997.99.3.534/abstract;jsessionid=720D4EB8C1610F3E8A1D81DD98E2B12B.d01t03

Modern Genetic studies, like where you go in and study the actual genes of the different groups, show that skin color and facial characteristics have no correlation at all with genetically similar or different groups. Dark skin correllates with living near the equator for 60-100 generation or more. That's it.


"Black" people in Africa exhibit more genetic diversity than all other people outside of africa combined. It's an overly broad construct. When we discuss African AMericans, there is so much genetic material from both European groups and African groups in both "Black: and "White" Americans, there is no way to tell from a genetic profile how that person will present in American society.

WHat seems obvious is not backed up by the science.

TGGP said...

The one-drop rule has been somewhat self-fulfilling in that self-identified blacks in the U.S average 80% African ancestry:
http://blogs.discovermagazine.com/gnxp/2013/05/why-race-as-a-biological-construct-matters/
It's different in places like Brazil where there's less correlation between observable phenotype and ancestry, but in the U.S self-identified whites are overwhelmingly European in ancestry.

Pittsburgh Mike said...

I'm one of the posters who referenced the Mismeasure of Man, and the link you posted "debunking" that book talks only about one part, having to do with skull measurements, which is completely irrelevant to the discussion of the persistence of IQ differences.

My main point, which still applies here, was that at the beginning of the 20th century, the groups that did most poorly in IQ tests included Asians, Jews and Eastern Europeans. The scores of all of those groups have greatly improved, such that many subgroups are now "above average" in IQ.

These are, as you say, facts. And even though some of these differences persisted for generations, they eventually disappeared completely and even reversed themselves.

There's no excuse for someone like Richwine, doing research in this area, to pretend that any IQ differences will continue as groups with lower IQs get better integrated into American society.

With so many examples of groups whose IQs increased by huge amounts over a few generations, I think it's fair to call "racist" those people who continue to ignore the transitoriness of these differences.