Both of us have explored the elbow room left by our genetic endowment and evolutionary history. Traditionally this is attributed to some vague entity called "the environment" or "nurture," variously identified with some parenting, culture, schools, conditioning, statistical regularities in sensory input, and the mass media. In writing The Blank Slate I was repeatedly struck by how mysterious this leftover category is. So much energy has gone into attacking the possibility that nature matters at all that the nurture side of the causal matrix has been left unproblematized, as the postmodernists say.
Judith Harris was among the first to call attention to the fact that we know much less than we think we do about "environmental" influences on personality and behavior. In The Nurture Assumption she points out that once you subtract out the effects of shared genes on correlations between parenting practices and children's outcomes (which few psychologists do), there isn't much evidence that parenting shapes children's personalities. It's not all in the genes, but the part that isn't from the genes isn't from parents either (siblings separated at birth end up no more different than siblings reared together, and adopted siblings end up not similar at all). She amassed data that people learn social skills and knowledge from their friends and colleagues (which in the case of children we condescendingly call their "peer group"), not from their parents. She also suggested that the nongenetic variation in personality comes from how we learn to specialize in social niches within that circle of colleagues.
If she is right about personality, then much of the "environmental" influences are consequences of random tosses of the dice: which niche is left open in a peer group one happens to join and how well one can fill it. It made me think of how many other openings there are for Lady Luck to affect our fates—for example, our responses (perhaps genetically constrained) to unpredictable events such as being targeted by a bully as a child, or coming across a pamphlet describing a possible career, or making a public commitment that it is too hard to back away from. Even before our life trajectory unfolds, chance events must affect the way our brains develop in utero and the first few years of life, because there is not enough information either in the genes or in sensory input to wire the brain down to the last synapse. Indeed, the paradox I mentioned at the end of my first posting taunts us with the necessary conclusion that chance events in large part shape who we are: Identical twins reared together, who share genes, family, schooling, siblings, birth order, peers, media influences, and every other measurable environmental factor are nowhere near being perfectly correlated in their personality, intelligence, or behavior.
Yet another mysterious nongenetic factor is what social scientists call secular trends ("secular" not in the sense of "unreligious" but of "extended in time"). The most famous example is the Flynn Effect: IQs have risen three points a decade for most of the past century, all over the world, and no one knows why; every plausible explanation has been disproved. Another is anxiety: People have become more anxious. A third is obesity: We've been getting fatter, and once again, none of the usual suspects (exercise, dietary fat, and so on) can explain it. Rates of violent crime yo-yo from one decade to the next—up in the 1960s, down in the 1990s—but probably the overall trend over the centuries has been downward.
These secular trends, like other mysterious nongenetic factors we call "the environment," show that there is more in heaven and earth than is dreamed of in our social science. The possible good news is that they show that some things can change, even when they are undoubtedly influenced by the genes, which don't change (certainly not over a few decades). As James Flynn pointed out, the black-white IQ gap would be completely explained without appealing to genes if the "environment" of blacks in 2000 is comparable to the "environment" of whites in 1950 (though we can't put our fingers on what in the environment makes such a difference). You mention that being overweight can't be cured by dieting or any other environmental manipulation we know of—yet if we put the average American in a way-back machine and sent him a couple of decades into the past, he'd be thinner. Perhaps he's be less anxious as well.
How optimistic should we be about these areas of wiggle room? Can we hope to bottle the causes of the random variation in personality, or of the secular trends? Or might they consist of thousands of microcauses, perhaps growing out of people's choices in a free society, that sometimes push together in one direction, or trigger a series of tipping points, but that no mortal scientist will ever catalogue in full?