Wednesday, May 22, 2013

all races being equal but all people certainly not being

This Zack Beauchamp missive on the Richwine affair has things to like and things to dislike about it. I want to focus on one thing, and in that way shift away from talking about race to talking about parentage.

Beauchamp goes hard on the notion that environment trumps everything when it comes to IQ. Indeed, he goes so hard on that attitude that most readers will likely think that there is nothing to the notion of a genetic basis for IQ. That's simply not in keeping with the large majority of the data. For example, that adopted children have IQs that correlate far more highly with their biological parents than their adoptive parents has been replicated repeatedly. (See, for example, Plomin et al. from 1997, for just one.) James Flynn, who I will remind you is deeply committed to social justice and is also the preeminent researcher in IQ, wrote in 2007, "The most radical form of environmental intervention is adoption into a privileged home. Adoptive parents often wonder why the adopted child loses ground on their natural children. If their own children inherit elite genes and the adopted child has average genes, then as parents slowly lose the ability to impose an equally enriched environment on both, the individual differences in genes begin to dominate." That Flynn piece, I think, is really excellent as a discussion of how to think through and understand the interactions between genetics and environment in IQ. It is not defeatist, and could never be called racist. But it is far more sober and clear about the relationship between genetics and IQ than Beauchamp's piece.

 Beauchamp quotes the well-known Turkheimer et al article that indicates that environment depresses genetic potential for IQ, but he doesn't point out that Turkheimer et al lends more credence, not less, to the fact that identical twins who were separated at birth and placed in unequal environments are far more alike than random strangers. Nor does he really explore what the Turkheimer article demonstrates. It's perfectly possible for environment to depress the IQ ceiling of a twin relative to his brother (I've been arguing that for years!) and for both brothers to have low IQs thanks in part to genetic lineage. Indeed, Turkheimer himself wrote in 2000, "the effect of being raised in the same family is smaller than the effect of genes." No one who reads Beauchamps piece without having been previously exposed to these issues would know that in fact Turkheimer has made some of the strongest arguments for the heredity of IQ. To come away from reading an article on the heredity of IQ thinking that Turkheimer is on the side of environmental causes is a big problem. The entire article lacks similar context.

Beauchamp quotes Richard Nisbett in arguing for early-life interventions as a way to improve IQ. I support early-life intervention. But Beauchamp does nothing to tell the reader what these interventions entail or what kind of resources they require. Nisbett is likely referring to studies of intensive preschool programs like that found in the famous Perry project. But Perry and similar other studies have typically suffered from small sample size and conditions that can't be replicated. What's more, actually investigating the results of that study shows that while program participants had statistically significant improvements compared to non-program participants, they still lagged far behind national medians in every tested category. Finally, Beauchamp himself quotes Nisbett in admitting that the IQ gains (which is what is at issue here) fade by adulthood.

Most egregiously, Beauchamp speaks about all of this by referring to a "new consensus," as if the notion that IQ is dominantly environmental rather than genetic is broadly shared in developmental psychology. This simply is not true. Early in his piece, Beauchamp says that he performed " dozens of interviews with subject matter experts." Well, let's be clear: he conducted dozens of interviews with subject matter experts who are inclined to be sympathetic to his preexisting commitments. I don't think that Beauchamp is being dishonest. I think he's a journalist who approached this question looking to find a particular position and found it. I'm not complaining about media bias or any such thing here, and again, in broad strokes Beauchamp and I agree on Richwine's argument for Heritage. It's okay to both perform journalism and have a particular point of view. But speaking as someone who has been reading academic journal articles and chapters on these issues for years, Beauchamp is presenting a deeply misleading portrait of current opinions on the relationship between heredity and IQ.

I have argued (again, for years) about the limitations of what IQ is and what it means in a human, social context. But it has to be said: if IQ had as little genetic basis as is suggested in that article, it would be extraordinarily out of the ordinary for measurable human attributes.

Why do I point this stuff out, when I've been making the case against the Richwine argument for the past week, and against the general race and IQ argument for years? Well, first, because I think that Beauchamp is simply presenting an inaccurate picture of the extant evidence. He is pointing to the fact that controversy exists on controversial questions and acting as if the existence of controversy is dispositive one way or another. He's cherry picking particular researchers who have particular stances (as all researchers do) and treating their opinions as necessarily dispositive. He's treating the existence of criticism in the literature as disqualifying of the papers Richwine cites, when of course criticism exists in the literature for the studies that Beauchamp himself cites. And he suggests a minimalist relationship between genetics and IQ that almost no one in developmental psychology actually believes. A reader who comes to that article without a background in these subjects will believe in a "new consensus" that doesn't exist.

But more to the point: we don't have to do this. We don't have to misrepresent the importance of genetic parentage to IQ to recognize the importance of environment. Beauchamp makes some very good points about what it means to be Hispanic and about what a race is. I myself have written four times in the last week or two about why we shouldn't listen to Jason Richwine. By misrepresenting the actual extant evidence, well-meaning people play into the hands of those who work tirelessly to establish the idea of a conspiracy to hide the truth.

Removed from the emotional grindhouse of race, why does all of this matter? It matters because our educational debates are dominated by a piety that almost everyone argues but almost no one believes: that all people are of equal ability. If you think that's an exaggeration, consider No Child Left Behind, which insists: 100% must achieve the standard, 100% compliance. Here in the real world, 100% of people will never reach the standard in anything at all. Yet this notion that our problems can all not only be improved upon but literally erased permeates education at all levels. It is the most glaring orthodoxy in our educational debates: you must never suggest that anyone will ever fail.

But here in real life, failure is always an option, and half of people will always be below average at everything, and both the data and the actual lived experience of everyone tells you that individual human beings have radically different abilities.

That's not much of a problem for me: I'm a socialist. I don't believe that human welfare should be subject to the vicissitudes of chance, which includes both environment and genetics. I don't believe in the notion that someone has to deserve material security and comfort. So I don't mind pointing out that human beings are substantially unequal in their abilities because I don't think that this should condemn anyone to a life of poverty. I've long advocated to the "scientific racist" crowd that we can conduct a really powerful experiment: raise the standard of living of black and Hispanic Americans to that of the white American middle class through the brute force of mass redistribution. Then, if their living standards start to decline after they've finally reached parity with the white middle class (which they've never enjoyed at any point in American history), we can go from there. If you think that human beings need to deserve things like housing, education, food, etc., then yes, inequality in ability might be a problem for you. But it certainly isn't for me.

8 comments:

dporpentine said...

So: like Steven Pinker--who takes a similarly "hey, those are the facts, dude" attitude toward the ability of women to do advanced math--you argue here that we should just get comfy with the existence of an overclass. The only difference is that in your case you seem to think that's mitigated by tacking on the word "socialism" at the end and making a case for something you know full well could never happen in the US. Such a brave position!

Ethan Gach said...

"the end and making a case for something you know full well could never happen in the US."

That's your position, not his.

Freddie said...

While I'm sure it's fun to simply ascribe positions to people that they don't hold when it suits you, I'm afraid it isn't a very responsible way to argue. Maybe you should read some more of my opinions before you claim that I think "we should get comfy with the existence of an overclass."

Steve D. said...

Great post -- I didn't realize that adoption studies showed such strong heritability for IQ.

N. Eugene said...

"you argue here that we should just get comfy with the existence of an overclass."

That argument exists nowhere in any of Freddie's writings.

Brett said...

Me neither.

@Freddie
I've long advocated to the "scientific racist" crowd that we can conduct a really powerful experiment: raise the standard of living of black and Hispanic Americans to that of the white American middle class through the brute force of mass redistribution. Then, if their living standards start to decline after they've finally reached parity with the white middle class (which they've never enjoyed at any point in American history), we can go from there.

It's a big country. You could probably find enough households in a particular income strata from different ethnicities, control for household wealth, and then examine the rate at which they drop down with each generation.

Freddie said...

Yes, but my version has the benefit of dragging a lot of people out of poverty. (Sneakily!)

Dan Crawf said...

When are social scientists going to realize that they have nowhere near enough humility when it comes to their own understanding genetics and heretibility? The past decade of genetics research has shown that epigenetic markers have huge impacts on development, occur both prenatally and even can even have external causes but be heritable (such as a link between grandparent's diet in youth and heart disease in the grandchild).

I have no doubt that there are heritable and genetic components to intelligence, but the genetics is far more complicated than nearly anyone who professes to be able to speak on this subject can comprehend. Researchers are rapidly learning that the more we know about genetics, the less we understand.

Why does this fact never seem to be reflected in these social science conversations? I think the answer lies in ignorance and a complete failure to understand that comparing a Bell curve level of knowledge on genetics to the present is like comparing Newtonian mechanics to modern physics.