SubscribeActually, there is a whole lot more to worry about with twins studies. Their results are expressed in terms of the heritability of properties, or proportion of variance supposed due to genetic factors. There is already a worry, since by the time of birth the twins’ genes have been expressing themselves in identical environment for nine months, and the time of separation and its extent are confounding factors (many “separated” twins are brought up within the extended family). The results of this research have included such gems as the heritability of milk and soda intake (high) or of fruit juice and diet soda (not so high). What is not usually stressed, and not stressed here, is that any measure of heritability is highly contextual. In a world of clones, the heritability of properties is zero; in a world of absolute sameness of environment, it goes to 100%. That is, if iron is put in a uniform environment, differences of rust are 100% due to difference of composition, but if identical samples of iron are put in a variety of environments, differences of rust are 100% due to environment. Heritability has also little or nothing to do with the malleability of the trait in question. In Swedish twins studies, heritability estimates for regular tobacco use was given as three times as great for men as for women, but for women it also ranged from zero to sixty percent in three different age cohorts, presumably because of changing cultural pressures on female smoking. Pinker is either not aware of the health warnings attached to this kind of research, or suppresses mention of them.
Although it is not always necessary to ask the evolutionary question regarding the adaptiveness of every stable trait, the presence of high and low fearful members in all mammalian species suggests that both styles must have adaptive features in order to survive over generations. The animal -- or child -- who is cautious to discrepancy is less likely to risk harm by impulsively approaching an unfamiliar object. On the other hand, the bold individual is more likely to gain resources that are limited. Steven Suomi, of the National Institute of Child Health and Human Development, reports that some fearful rhesus living on the small island of Cayo Santiago are liable to starvation because they wait until the other animals have had their turn at the food stores placed daily by research staff; on some occasions all the food is gone when the bolder animals leave. On the other hand, some of the fearless animals die of wounds incurred in attacking a stronger animal. Thus, the advantages associated with each temperament are balanced by disadvantages in a different context. There are no free lunches.As for your daughters, I wouldn't worry too much, as there is no evidence that they will live any less happy or productive lives. :)
p. 264
The evidence presented in The Long Shadow of Temperament suggests that infant temperament is a poor predictor of temperament in puberty. In part, these findings may be due to the authors' choice to focus on discrete types, which are statistically and conceptually problematic. Stronger results have come from researchers such as Avshalom Caspi and colleagues (2), who conceive of temperament in terms of continuously distributed personality traits. They see moderate continuity from age 3 to age 26, and extensive research shows that personality traits are impressively stable from adulthood through extreme old age (3). (italics mine)And (skipping back up to the meatier part of that review):
Kagan and Snidman note that with one exception (the enhanced startle response) the biological data are "in modest accord with the expected outcomes for children who had been high- or low-reactive infants." But there are several troubling inconsistencies in their account. For example, in summarizing their behavioral assessments they comment that "about 1 in 3 high-reactives (22 of 67 children) and 1 in 2 low-reactives (46 of 92 children) had developed social behaviors that were predictable from their infant temperaments. Only 8 high-reactives and 13 low-reactives developed a profile seriously inconsistent with expectations." Seen another way, these data show that two-thirds of the high-reactives do not develop into the inhibited type, and half of the low-reactives do not remain uninhibited. In another example, the authors speculate that the uninhibited profile was better preserved because "family and friends encourage sociability rather than shyness, and American children would rather be sociable than shy." Overgeneralization aside, such environmental sculpting of the sociable disposition is hard to reconcile with their finding that for children evaluated at 4, 7, and 11 years, 61 percent of the high-reactives were always subdued and 33 percent of the low-reactives were always spontaneous.Furthermore, it's important to recognize that even if, unlike the reviewer for Science, you accept Kagan's arguments, his conclusion is that early temperament is an influence on later character, not that it decides everything in advance. What Kagan says he shows is that a small proportion of people who start out shy remain shy, while the vast majority stop being shy; and that those who start out really shy generally don't join the category of the extremely extroverted later on.
Is it really?
posted by meehawl at 8:55 AM on October 23, 2005