On Sat, 28 Aug 2004, ravi wrote:
> i think you may be missing the point: height is well-defined, at least
> in some sense. gould's and others' criticism is that the notion of IQ is
> conceptually not well-defined. if its nothing but how one performs on
> particular tests, surely the onus is on the other camp to prove that
> such things are indeed hereditary? gould and the related camp could be
> wrong about their belief. but you may be constructing a strawman above.
>
I think this is conflating two questions:
1. Is IQ performance heritable?
2. What do IQ tests measure?
Granted, the questions are related, but they are distinct. The first one turns out to be the easier question: Twin and adoption research clearly supports the claim that variation in IQ performance is in part influenced by genetic variation (e.g., the correlation in IQ score between a biological parent and one of their offspring raised in an adoptive household is greater than 0: it's hard to cook up a nongenetic explanation for that.) Where our new friend drives me nuts is with the spiel about "genetic determinism" (Luke, what were you saying about people not being this naive about genetics?).
The second question's the tricky one, hotly debated by research psychologists. Some psychologists claim that IQ measures intelligence (more or less by fiat, your IQ score is an valid representation of how smart you are). Other psychologists argue that IQ tests assess academic skill ("book smarts"), but not general intelligence. In my view, a lot of data supports this latter view: e.g., Ceci's found that successful bookies intuitively apply complex mathematical models in their odds-setting, but they cannot solve formal representations of the same problems on a standardized IQ-style test. (Just so with comparison shopping: if people have a lot of experience shopping, they can pick out the best deal, even if they cannot do the relevant math on a standardized test.)
So to use a cliche: IQ measures book smarts, but not street smarts. (There are many types of practical, everyday intelligence that people use in our society that are not accurately measured by IQ score. However, if you want to know who's going to succeed in academic settings, IQ is a relevant predictor.)
Miles