Sowell's wrong
I don't have time (for a few days at least) to deal with Thomas Sowell's
column 1 and
2 on race and IQ, so I'm going to post a letter I wrote to a interested correspondent on the topic. I want to begin by giving Sowell major credit for mentioning the race/IQ issue in public in the first place. Even though he's wrong (as I'll briefly detail), he's done us all a service by daring to broach the issue at all. Ok - let's get to the nitty gritty:
The Flynn effect is the major theoretical problem now in intelligence testing. There's a discussion on it here (search on the page for flynn) and in the link on the left hand side under "Flynn Effect".
First and foremost: it is incorrect for Sowell to claim that the Flynn effect means that genetic difference are immaterial. The black-white IQ gap has remained steady at 1 SD while absolute IQ scores have risen. As an analogy, heights for both asians and whites have increased over the past century, but that does not mean that asians will ever catch up to whites in average height.
There are several possible explanations for the Flynn effect. Most people accept that it's environmentally caused rather than genetically caused, b/c the gains are very large indeed and have (seemingly) not been accompanied by any visible genetic changes. Some possibilities:
1) Real intelligence has risen just as real height as risen, due to a better environment/nutrition/etc.
2) Only test scores have risen, but real intelligence has not changed. This might happen if the change in environment was "training for the test".
3) Hybrid vigor caused by intermarriage among previously isolated groups (due to mass transportation) has lowered the frequency of homozygous recessives and caused a real genetically caused gain in intelligence. Note that this does not have to be black/white intermarriage; Irish/Italian intermarriage would also work.
I only heard theory 3 recently, after the post I linked above. It *may* be a non-environmental explanation for the effect, but I find it implausible.
The only way to distinguish these possibilities is to move to an "absolute" metric for intelligence, so that we can measure *real* intelligence. This is the impetus behind the development of mental chronometry: measuring intelligence in units of inverse time, just as we do for computer processors. See this link for more details, and also check out the interview with Arthur Jensen linked on the left hand side.