In his long post on John Derbyshire (henceforth “Derb”), my colleague John O’Sullivan (henceforth “John”) says many things that deserve thoughtful consideration. Without addressing John’s entire position, I will comment on the following two paragraphs, which contain a criticism of me:
If principled anti-racism requires people to ignore considerations such as personal safety, then racism will not seem all that terrible. In a world of muggings, stabbings, rapes, and murders, it would look almost like a technical offense, born of excessive prudence, provable only as a result of close reasoning, and shared with black parents who talk the Talk. Of course, we are under no obligation to define racism in so broad and idealistic a fashion. Once we reject doing so, however, the case against Derb collapses. . . .
. . . [S]omething worse is therefore needed to convict him. My colleague Jason Steorts thinks he has found it in his “readiness to assume that statistical differences between races — e.g., between their incarceration rates or average scores on IQ tests — are due to innate psychological and cognitive differences” and in his lack of interest in other explanations. This readiness strikes Jason as redolent of racism in itself but also perhaps a sign of a wider animus toward black Americans. Now, John certainly holds the view quoted by Jason. It is one of the main line-items in the indictment against him. Far from assuming it, however, he has argued it at length over a number of years.
In fact I deliberately avoided using the words “racist” and “racism” in making this point. What I actually wrote is that I found Derb’s defenders “insufficiently troubled” by his treatment of the issues under discussion, and I left open the possibility that the only troubling thing is his lack of interest in making a thorough case for his view.
What I mean is this. Over the years, Derb has written at length about the rationale for evolutionary/genetic explanations of group differences between races, but he has written almost nothing about how these explanations should be weighed against alternative hypotheses. What specific evidence is there for or against these alternatives? If individually they are insufficient to explain group differences, how far do they go jointly? How much work is left over for evolutionary/genetic explanations to do at the maximum, and how open should we be to the possibility that yet-to-be formulated non-evolutionary/genetic hypotheses might still do some of this work? The American Psychological Association, for example, has held since 1996 that environmental factors are insufficient to explain the racial IQ gap, but that there is not enough evidence to establish the existence of a genetic cause, and that the question has no scientific answer. Among writers who favor evolutionary/genetic explanations, there are those who give due consideration to the questions I have raised, but I continue to think that Derb does not. Here is a representative quotation, from remarks he delivered earlier this year:
In group difference of outcome between races, the Standard [Social Science] Model [which Derb defines as the doctrine that “all observed group differences are the result of social forces”] was . . . not preposterous fifty years ago. Here in the U.S.A., nonwhite citizens labored under well-known legal and social disabilities. In what we were just beginning to call the Third World, nonwhite populations had been humiliated and subordinated by decades of colonialism. Remove the disabilities, dismantle colonialism, and socio-economic group differences would surely melt away. The Standard Model as applied to race was not preposterous fifty years ago.
I think it is fair to say that it is now, half a century on, well into the zone of preposterosity. Segregation and colonialism have long since been dismantled; trillions of dollars have been spent to rectify past wrongs; countless helping-hand policies have been enacted; yet still the world, and our individual nations, are deeply stratified by race. “The harvest is past, the summer is ended, and we are not saved.”
However intuitively plausible you find the second paragraph, I think it is fair to say that it falls well short of the empirical and analytical rigor that Derb usually aspires to. It is not based on any study. It does not explain why half a century is too long for the effects of segregation to persist. It does not examine any of the “countless” policies. It is a 72-word summary of 50 years of history. And the thesis it opposes is that social forces explain all observed group differences. The question “If not all, then how much?” is not even raised. It is not raised anywhere in Derb’s writings with which I am familiar.
So I do think Derb seems — as I first put it — “uninterested” in explanations other than those he endorses, or at least in explaining why he does not give more consideration to the alternatives, and this does trouble me. My one regret is that I spoke of his “readiness to assume”; here I accept John’s criticism, and in retrospect I wish I had written “readiness to conclude.”
Regardless, I did not lay nearly as much stress on this point as John’s post suggests. For me, the main item is and always has been the practical recommendations Derb made based on his perception of the relation between statistical likelihood and personal safety. (I called this the “most overtly racist” aspect of his article. By this I meant to suggest the idea of racism as continuum, since reasonable people can disagree about the point at which it begins. But my words can also be read as implying that there were less overtly, or covertly, racist elements to Derb’s article, and this may be what led John to his “redolent of racism.” If so, the fault is mine, and I should have chosen a clearer expression.)
On this main item, I think two points are crucially important. 1) The law of large numbers does not apply in an individual person’s practical decisions the way it does in, say, an actuary’s calculations for an insurance company. The instances are comparatively few, and the relevance of a group mean is comparatively small. This does not mean that statistical averages are irrelevant to, and never borne out in, personal experience; if that were so, there could be no such thing as poker. But some of the scenarios about which Derb counsels us are very, very specific (acting the Good Samaritan to a black person in distress). The examples he links to are even more specific (acting the Good Samaritan to a black woman in a case of domestic violence). How many such encounters will any of us actually have? Enough to justify taking Derb’s advice? 2) At least some of Derb’s recommendations (as well as Jesse Jackson’s comment about muggers) are based on implicit probabilities about what a kind person will do on this occasion (mug me; kill me; whatever). But these implicit probabilities are justified with statistics about the things that kinds of people do over long periods (mug someone, sometime; murder someone, sometime). You cannot go from the latter to the former with any precision.
Now we can grant points 1 and 2 and still make a defense of Derb (and Jesse Jackson). It boils down to, “Better safe than sorry.” (I.e., if a person from Group X is more likely to do a given thing than is someone from Group Y, then I am more likely to have that thing done to me when in the presence of someone from Group X than when in the presence of someone from Group Y.) That’s fine, if you present your position as — and it honestly is — a general (i.e., non-race-specific) and extreme prudence that goes beyond what can be justified with statistical rigor. But I don’t think Derb has ever done that, and I find the thought incredible that he would be — that anyone could be, this side of insanity — so careful in all his affairs (think of all the things we’d end up avoiding: getting in cars with older or newly licensed drivers; biking by night on an unlit street; eating shellfish; skiing; skiing at resorts that serve beer; etc. etc. etc. etc. etc.).
So then I ask: Why does Derb do this specifically with race? And why does he go on and on and on about racial group differences? The charitable answer is that our culture is obsessed with race, goes on and on and on about it; that Derb is just joining the conversation; and that often he is doing so in reaction against the sort of political correctness that would stop us even from noting group differences or investigating their origin. I believe there is something to that. But his recent article, in which group means are used to justify default treatments of individuals, crossed a new threshold. Given that he said he has offered its advice to his children in bits and pieces over the years, a claim he repeated in his Gawker interview, I was not able to dismiss the article as satire, irony, and role reversal. And in any case, Derb has disavowed any satirical intent.