Wednesday, 11 April 2012

John Derbyshire.

William Saletan has written an excellent analysis of the John Derbyshire controversy, touching on what anyone who's known me more than ten minutes will have heard me railing against: the single biggest mistake people make when drawing conclusions from statistical data: "average" does not mean "normal".

Derbyshire thinks his data warrant his conclusions. But all his data references include the crucial term “mean” or “average.” They don’t tell you about the person walking toward you. They tell you what you can assess about the probability of danger when the only information you have is color. Look at Derbyshire’s point 10: “where you have nothing to guide you but knowledge of those mean differences … Avoid concentrations of blacks not all known to you personally … If accosted by a strange black in the street …” The common premise in all this advice is ignorance. Not ignorance of data, but ignorance about the person you’re facing.

Derbyshire relies on the same assumption in point 12: “[I]n those encounters with strangers that involve cognitive engagement, ceteris paribus the black stranger will be less intelligent than the white.” Ceteris paribus is Latin for “all other things being equal.” It assumes there’s no difference between a black person and a white person except that each has the average IQ test score for her race. In other words, the equation holds, as a matter of probability, only if you fail to notice anything about the person you’ve encountered aside from color.


I do want to comment on this one of Derbyshire's pieces of advice:

(10a) Avoid concentrations of blacks not all known to you personally.


I used to live in Govanhill on the Southside of Glasgow, in the middle of Glasgow's Asian population, and I have to admit I followed a similar rule: in the street, I would always avoid concentrations of young white men not known to me personally. Young Asian men, on the other hand, were not theatening to me at all, and it would never even have occurred to me to cross the street to avoid a group of them, no matter how large. This is not because I have anything against whites, as I am one, and it's not because I'm especially fond of Asians or of Muslims, both groups containing, like every other group of humans on the planet, large numbers of utter bastards — which is in fact Derbyshire's point 5. And I wouldn't necessarily follow the same rule in other parts of the world — Asian populations in England and Scotland are markedly different, for instance, and you'd have to be a bit strange to feel threatened by a large group of white men hanging around in an East Anglian market town such as Diss. My behaviour was guided simply by my knowledge and experience of Glasgow, a city whose natives are both wonderfully friendly and fucking dangerous. As Frankie Boyle said, "If I had to explain Glasgow to you, I'd say that if I had to pick a city in the world where I could depend on a member of the public to punch a man who was on fire."

Now, it strikes me that I was making, basically, exactly the same judgement as Derbyshire, except, since the group I avoid are white, I won't get much stick for it. Never having lived in the same part of the US as him — or any part of the US, for that matter — I am in no position to judge whether his advice is any practical use. But it does seem clear from his writing that what he's talking about is not his experience of his own life, but extrapolation from statistics, and he does seem rather obsessed with genetics. I don't think Derbyshire would make different judgements in different parts of the world. I don't think he's saying that large groups of black men in his area can be dangerous; he's saying that blacks are dangerous, all the time and everywhere. And that is of course bollocks. Mark Steyn:

He thought that neuroscientists and geneticists’ understanding of race trumped my touching belief in “culture.” I’m not so sure: Why is Haiti Haiti and Barbados Barbados? Why is India India and Pakistan Pakistan? Skin color and biological determinism don’t get you very far on that.


And anyone who's worked in IT will recognise this description of Saletan's:

But it tells you a lot about Derbyshire. It tells you he’s a math nerd who substitutes statistical intelligence for social intelligence. He recommends group calculations instead of taking the trouble to learn about the person standing in front of you.


For what it's worth, this is why I disagree with National Review's decision to sack him. Anyone who's read some Derbyshire will know that the man is not a racist. What he's guilty of here is the obsessive concentration on and extrapolation from easily measurable data and the exclusion of factors that are vaguer and trickier to measure — such as humanity — characteristic of the borderline-Asperger's maths and science nerds we all run into every day. But he was a good writer, and his sometimes overconfident faith in science regularly pissed off National Review's Creationist readers, which was surely a good thing.

He also once wrote this, which I love:

My ideal nursing-home attendant, auto mechanic, or president would be a cheerful, capable, well-motivated person who was thoroughly au courant with the theory of evolution — and indeed with all the most recent advances in astronomy, biochemistry, cosmology, dendrochronology, endochrinology, fluviology, geomorphology, hydrodynamics, ichthyology, jurisprudence, kinesiology, limnology, microbiology, neuropathology, ophthalmology, psychometrics, quantum chromodynamics, rocket science, seismology, trichology, urology, virology, wiretapping, xenodocheionology, yachting, and zoology.

Life, however, often consists of making a choice between unsatisfactory alternatives. Invited to choose between having my kids educated, my car fixed, or my elderly relatives cared for by (a) people of character, spirit, and dedication who believe in pseudoscience, or (b) unionized, time-serving drudges who believe in real science, which would I choose? Invited to choose between a president who is (a) a patriotic family man of character and ability who believes the universe was created on a Friday afternoon in 4,004 B.C. with all biological species instantly represented, or (b) an amoral hedonist and philanderer who “loathes the military” but who believes in the evolution of species via natural selection across hundreds of millions of years, which would I choose? Are you kidding?