A nice piece in yesterday's NYT by psychologist Richard E. Nisbett on the race/intelligence controversy sparked by The Bell Curve and reignited recently by James Watson. (Thanks to Michael Bowerman for pointing it out).
Nisbett cogently and concisely deflates the claims that intelligence has been proven 60-80% "heritable" by twin and adoption population studies. By the end of the piece, the reader is left with the strong impression that those studies are flawed and misleading. They don't stand up, even on their own terms.
What Nisbett doesn't do -- understandably, because it's a lot trickier -- is explain how those population studies also fly in the face of our modern understanding of genetics. It's not biologically possible for someone to directly inherit, via genes, a certain level of intelligence. Genes don't work that way. Everything about our genes is mediated through interaction with the environment. The dichotomy of "nature vs. nurture" actually does not exist.
Our popular discussion of genetics, intelligence, talent etc is stuck in a very strange place. We use terms, concepts and metaphors which lead us astray. In order to get unstuck, we're going to need a whole new way to frame the discussion. That's what I'm working on (struggling with) in my book.