I don't see how the relationship system, socialization system, and status system would produce variation among twins. The purpose of these systems is supposedly to direct individuals to follow their strenghts and weaknesses and find areas where there is an unoccupied niche. The level of competition is likely similar if the twins are born into the same society at the same time, and the part about following strengths and weaknesses is circular reasoning, as differences in these had to already exist for the systems to find them. Also, I don't see how any of the functions attributed to these systems require a mind-reading module.

Re: socialization system -- I wanted to keep the focus on her central thesis, and the socialization system makes people similar rather than different. That's really the only reason I didn't mention it or any of the other really great stuff in the book (notably her evisceration of birth-order demagogue Frank Sulloway).

As for whether her proposal is or isn't a social-level version of DN, I should define exactly what I mean: DN is the orderly, non-random amplification of chance events, which is pretty much what she says with the phrase "widen the differences." It's like getting a random number from a pseudo-random number generator. For example, by the middle-square method, if you want a random number of 6 digits, you take any old 6-digit number (a capricious choice on your part). Then you square it, take the middle six digits of the result, and repeat the process a bunch of times (a step-by-step process utterly lacking in caprice).

Her proposal represents the non-random amplification part: e.g., if you witness someone trip into the salmon mousse, respond by pointing and laughing out loud; if you receive fewer than X% of all gazes in social interactions, respond by being more introverted than before; if the composite profile the Generalized Other has of you is of a wimp, respond by being less aggressive and cocky; and so on.

But whether MZ1 or MZ2 gets more gazes, etc., is either predictable -- in which case the cause can't be shared genes or shared environment, and must be some previous instance of nonshared environmental cause -- or it's not predictable -- in which case it's operating on chance, like a sampling error (e.g., 49 of 100 gazes vs 51 of 100 gazes).

So, I certainly didn't mean to imply that Harris' proposal is just chance! Just that DN even at the micro-level is this sort of non-random algorithm acting on a chance event. So, there's a chance disturbance in gene expression of proteins that guide axon growth, and the growth cone is pulled left instead of right -- a chance event -- but after that, a perfectly orderly process finds the growth cone a target, it synapses on the target as all others do, etc. (an orderly process).

I wholeheartedly agree that evolution doesn't favor this sort of thing if it has fitness costs -- if it allowed so much chance that the brain couldn't wire itself together -- but whether you're more or less introverted or more or less open to experience isn't of the "your ear is wired to your eye" type. The less crucial to fitness, the more DN is allowed.

Also, if I'm on the right track about microorganisms, then evolution doesn't select out disturbance of our "natural" personality -- since evolution doesn't care about us any more than the microorganisms trying to find a way around or through our defenses.

“Brain's Darwin Machine”

Jason Malloy provided this related link in GNXPForum:

http://www.latimes.com/features/...=la-home- health

“Together, the couple stalked an elusive sequence of DNA hidden in the heredity of every human cell. The wayward strand appeared to seek out developing brain cells and, like a virus, arbitrarily alter their genetic makeup.

In this way, it might be partly responsible for the infinite variety of the mind.”

Arosko, I think what she's saying is that these social niches exist in very small groups (for example, each class only has one class clown), and you really don't want to be competing with your own identical twin for the same social niche.

Theoretically, if their environments were sufficiently similar, identical twins rasied apart should be more similar to each other than would be identical twins raised together.

Just as a nitpick, there are times when it makes a lot of sense to randomize the behavior of an algorithm, especially when there are many copies of it in the world that might need to coordinate, or at least to minimize competition. (In fitness terms, twins don't compete, but I doubt their mental modules know anything about that.)

Comparitive advantage depends on speciailzation, e.g. the formation of differences. Even random differences might help.

By the way, it seems to me that Harris's hypothesis regarding niches and specialization suggests that children put in tracked classes should tend to regress to the mean more than children placed with less intelligent peers. After all, in a class of all smart kids, there would be less chance of a kid with an IQ of 120 or so filling the "smart kid" niche and hypothetically becoming an adult with an IQ of 130.

Finally a rational argument for eliminating tracking, if it proves true, but if true I would expect to have heard of it by now.

More generally, this seems to suggest that you should weaken any trait by surrounding its posessor with people who share that trait... that is, unless the relationship and status systems are counteracted by the socialization system, in which case Im not sure what if anything this hypothesis predicts.

So, would the implication here be that sending a tough kid to reform school would actually reform him, as the surrounding even tougher kids would take away the tough-guy niche? Perhaps a secondary effect of tracking is that it changes how many of each kind of niche there are. At the reform school, the only available niches may be tough kids and perpetual victims; at the gifted school the only available niches may be variations on "smart kids," though there may still be a couple kids that are still smarter than everyone else.

Thinking about it some more, you could easily imagine it working both ways. On the one hand, your identical twin could be a more direct competitor than anyone else imaginable. But on the other hand, whatever you're interested in, you'll probably get more "into" it if you have other people to talk to who are "into" the same thing, and who could be more likely to share your interests than your identical twin?

But it's probably true that whenever you have two people that spend a lot of time with each other, whether friends or relatives, most likely one of them will tend to be more the "leader" and the other more the "follower".

Re: tracking -- under Harris' view, this would have an effect on personality, though intelligence she doesn't discuss. The personality case sounds plausible: if the Generalized Other is composed of all badass peers and badass drill sergeants, their composite profile of you will be much wimpier than a random sample from the general pop would have. So it'd put you in your place.

But intelligence? Doubt it. You can't unconsciously adjust your processing speed (or whatever) downward. What it could affect is "crystallized" g (how big your vocab is, how many world capitals you know, etc.), or how nerdy you act. But not "fluid" g.

"By the way, it seems to me that Harris's hypothesis regarding niches and specialization suggests that children put in tracked classes should tend to regress to the mean more than children placed with less intelligent peers. After all, in a class of all smart kids, there would be less chance of a kid with an IQ of 120 or so filling the "smart kid" niche and hypothetically becoming an adult with an IQ of 130."

No. Harris specifically addresses this in her book. This doesn't occur because of the effect of the socialization system. While a gifted kid surrounded only by gifted peers would have her academic confidence decrease (since it's harder to compete for the "smart-kid" niche when surrounded by smart kids), her performance does not decrease because she's also being socialized; academic performance is the norm of her new group.

From pp 212-213 of No Two Alike: "A child of above average academic ability will have above-average academic self-esteem. But if we put this child in a highly selective school where she is only an average student, her academic self-esteem will drop. So does this lead to a poorer academic outcome? No, because something else is going on as well: the child is getting socialized. She's conforming to the norms of her group. A quarter century of research has shown that children learn more in schools and classrooms that have a relatively high concentration of able students."