I.Q. Aint Everything
In 1984, social scientist James Flynn noted what became known as the Flynn Effect, the general trend in I.Q. scores to rise over time. The effect was quite broad, and fairly smooth. The effect is no longer in doubt as a statistical fact, but what it means is less clear. Malcolm Gladwell writes about it in the New Yorker, 12/17. Recently Flynn debated Charles Murray, one of the authors of "The Bell Curve", at the Manhattan Institute, on the black/white I.Q. divide. Gladwell takes the occasion to look over what has been learned since the effect was first noticed. It seems it is a specific result of certain types of information and thinking becoming more pervasive over time, so the effect really shows when a given version of I.Q. test was "normed", the questions chosen. The Wechsler Intelligence Scale for Children is the common form, and has been re-normed three times since its first version. Every time it has been reset, I.Q. differences between groups (that would share in that fact set and worldview) shrink to small or trivial. Also, the percentage of testees rated as retarded decreases over time, until the new norm is employed, when the percentage jumps up again. An example of the type of knowledge that can escape WISC categories is what Ameicans would say about dogs and rabbits now, as opposed to last century. Now we would likely call them both mammals, but a hundred years ago the naturalists view would not have been as likely as the farmers or hunters view, that dogs prey on rabbits, so one is a farm tool and the other a resource. Psychologist Michael Cole gave a version of the WISC similarities test to the Liberian Kpelle tribe. A basket contained food, tools, containers, and clothing. The tribesmen were asked to place the items in groups. Annoyingly they always chose functional pairings, such as the knife with a potato. "A wise man could only do such," they said. When asked what a fool might choose, they easily sorted items into the "right" categories of type. Flynn likes to say I.Q. measures the extent to which we have "put on scientific spectacles" in our worldview. When Flynn looked at Richard Lynns data that seemed to argue an inherently higher I.Q. for Asians, he found that the demographics had not been set equal, with a broad American sample compared to a narrow upper-class urbanized Japanese sample. Recalculated, the Japanese tested at an average of 99.2. Chinese-Americans data turned out to be from the Lorge-Thorndike test, but that had been normed in the 50s and was used in the 70s. No wonder it was a piece of cake. Re-assessed the Chinese-Americans tested at 97 verbal and 100 nonverbal. They worked harder, their homework practices in the family were more effective, but they werent whiz kids. At the debate, Flynn pointed out that the data showed little difference at earliest ages, and the gap between black and white widened with age. This was not an expected genetic pattern but fit well with the different cognitive environments experienced by the subjects. He also showed that data from mixed-race marriages and adoptions did not fit a genetic model either. Children of black American GIs in post-war Germany that were raised by German mothers tested at the same levels as children of white GIs with German mothers. Flynn says "the brain is much more like a muscle than weve ever realized ". But another view is possible, that what WISC and I.Q. are measuring is not intelligence but shared logical categories more than anything else. Consider that computer expert systems are common now, from Deep Blue to Aegis defense systems. What is hard to emulate is precisely social intelligence, fitting in smoothly with the milieu, laughing at the right jokes, and not doing something stupid like trying to use a knife to eat rice.




