Yesterday, I said truth alone wasn't sufficient grounds for framing test scores this way. I said we should do it only if classifying and comparing by race, rather than using some other classification system or judging each person as an individual, does more good than harm. So far, I've argued only the first half of the equation: that racializing the problem is unnecessary and unhelpful. But what about the second half? What's the harm?
There is none, says McWhorter. Commenting on the New Haven case, he writes,
Of course, the question we are not supposed to ask is whether the failure rate suggests that black people are less intelligent. However, there is no need to fear here. The reason black people of unaffluent origin tend not to do well on standardized tests is a matter of language and how it's used—and the issue is less about color than class, and in the global sense, about what it is to be human.
Less about color than class? That pretty much concedes the point we're debating here. But let's set that aside and go straight to the ugly part. McWhorter casually dismisses the less-intelligence theory and its blogger-advocate Steve Sailer, with whom I tangled yesterday. Why? Because McWhorter is confident that his alternative theory, based on language, can explain racial gaps in test scores. In his commentary on the New Haven case, McWhorter lays out the theory: Working-class blacks and whites communicate orally rather than in writing, and they're unfamiliar with the art of answering direct questions. I'm sure there's truth in this theory. But McWhorter offers no quantitative evidence for it. Nor does it address some of the most difficult evidence presented by proponents of the genetic theory: whites outscoring blacks even when the class factor skews the other way. In his rebuttal to my original article on the NAEP data, for instance, Sailer notes:
Here's the 2007 8th grade Reading scores broken down by race and income. White kids whose parents are so poor that they are eligible for the National School Lunch Program outscore affluent black kids by four points and affluent Hispanic kids by one point. The gap between poor whites and poor blacks is 19 points, and the gap among not poor whites and not poor blacks is 21 points. That's what you normally get—sizable racial gaps anyway you slice it.
Is Sailer a nice guy? No. Does he display an unhealthy interest in categorizing people by race or ethnicity? Yes. But the problem here isn't Sailer, James Watson, Charles Murray, or anybody else you feel like dismissing as a racist. The problem is the evidence these people quote. Condemnation won't make it go away.
Don't get me wrong. Genetic and environmental explanations aren't mutually exclusive. In the case of IQ, everybody accepts environmental factors, and there's plenty of evidence and argument against the hereditarian view. But that's just one battle in a larger war. Beyond the march of test scores, there's the onslaught of genetic research. We've already identified genes that correlate with traits and vary in prevalence between ethnic groups. Are you confident that intelligence will turn out to be exempt from this list? Confident enough to leave no backup plan, no understanding of equality that can withstand a partial role for heredity? Confident enough to keep tallying and reporting test scores by race? And if intelligence turns out not to vary genetically between groups, do you imagine that we'll get just as lucky with every other significant mental trait?
If you want to know why I keep writing about this subject, Mr. McWhorter, there's your answer. No, I don't care about the merit badge. I'm staring over your shoulder at an oncoming train. It starts with genomic differentiation of populations around the world, and that's just the locomotive. If you turn around and look, you'll see that the first few cars are already in view: genes that affect mental traits, genes that affect abilities, and variations between populations in the prevalence of these genes. No genetically distinguishable population will be spared. We're sitting in the path of this train, tied to the tracks by a literalist conception of equality that can't accept hereditary differences between group averages. I suggest we free ourselves.
Under these circumstances, do I think gaps between average white and black test scores should "shed less than positive light on black people"? No, I don't. Each of us should be judged by his own performance, not by a stereotype. Genetic variation between averages doesn't alter that moral truth. Nor does it give anyone an excuse.