« From the vault | Gene Expression Front Page | Islam-Peace or Submission? »
March 03, 2003

Harris, Kagan, Rushton, and the Power of Biology

Lately, due in large part to The Blank Slate, Iíve been interested in the developmental theories of Judith Rich Harris as laid out in her book The Nurture Assumption (book/website). Harrisí thesis, built mainly off of twin and adoption studies, is that parents have very little to do with shaping the intelligence, mental health, and personalities of their children, which can better be attributed to other factors such as shared genes, peer group socialization, and the mysterious wildcard of nonshared environment. Leaving praise and criticism of these ideas to Pinker and Sailer respectively, Iíd like to instead focus on a rather old Slate debate over The Nurture Assumption, that I came across while reading-up on the book. The dialogue comes from the time of the TNAís publication when the controversy was still flaring, and it is Between Mrs. Harris and, probably her most vocal (and high-profile) critic, Dr. Jerome Kagan, Harvard professor of psychology (he has called her book "an embarrassment to psychology"). Itís an ugly exchange, and both specialists seem driven by a certain degree of personal stakes that probably leads both into omission and over-statement. Particularly striking to me was the clear dishonesty of Dr. Kagan, who, in order to attack Mrs. Harris, seems to be omitting important ideas that have found formulation in his own scientific work. Kagan, who spends a curious amount of the dialogue attacking heritability, concludes:

A third flaw in your argument comes from studies of young children orphaned by war -- World War II and the Korean War. These children were adopted by nurturant families and years later their intense anxiety and retarded cognitive abilities were muted in a serious way. There is no other way to account for this result without attributing power to family practices.

No other way, to account for this result? Are you sure about that Dr. Kagan? Incidentally, I was familiar with (and actually excited about) Dr. Kagan, a particularly esteemed developmental psychologist, because I recently read about him in the "hereditarian"[1] book Born That Way by William Wright (Iíll read anything with an endorsement by E.O Wilson on the back). The book works to paint Kaganís journey as a staunch environmentalist of the Skinner age, to a modern-day Gene Expressor:

"For Kagan this faith [environmentalism] began to weaken in the early seventies, when he spent a year in Guatemala observing infants in a remote mountain village. The village children were of particular interest because of an unusually deprived first year of life that sprang from the Indiansí tradition. Local custom had the mothers isolating their children inside cramped dark huts for the first year, never allowing them outside, never playing with them, and almost never speaking to them. As a result, the children at the ages of one and two were observed to be unusually passive, quiet, and unresponsive. To Kagan, some appeared border-line retarded

Setting to work, Kagan and his colleagues set up a controlled study in which the village kids were measured on a number of cognitive functions. Those results were then compared with a Guatemalan group raised in a more normal fashion near a city and with a group of middle-class American children. The three groups might roughly be termed under-stimulated, normally stimulated, and over-stimulated.

The results showed that while the deprived village children measured lower in all tests in the first years, they tended to catch up as they grew older and their rearing setting became more similar to those of childhood everywhere- outdoor play, interaction with parents, with other children and so on. By eleven and twelve the Guatemalan children could be considered normally developed. This transformation was heartening to environmentalists like Kagan and to socially concerned individuals who placed hopes in intervention programs like Head Start which help underperforming children catch up with their age groups. Good behaviorist soldier that he was, Kagan wrote in an early paper on this study, "The data proves the potency of the environment." Without hesitating he credited the good environment with remedying the effects of the bad.

But the data he came to realize, could be seen in two ways. With the Guatemalan village children, there had been no intervention, no remedial program; they had merely been delivered from the negative environment of their first year, the dark hut. Something else seemed to be causing their improvement, and future evidence indicated it was their own normal development, in all likelihood their own genetic makeup triumphing over an adverse environment. Rather than the normal "outdoor" environment curing the negative effects of the grim first year, it may have merely allowed the childís personality to unfold as its genome, or as "nature", intended."

Now it's true that I am unfamiliar with the particulars of both the Guatemalan and the Korean study. Perhaps there was a control for the Korean study where some of the disturbed war-infants were put into these so-called "nuturant" families, while others were given to what Sandra Scarr would call "good-enough parents"[2]. Somehow, I doubt this. It appears more likely that Dr. Kagan was simply dishonest in the way he approached debating Mrs. Harris. How else can Dr. Kagan justify saying, "There is no other way to account for [the recovery of the Korean war orphans] without attributing power to family practices", when it was he, himself, who decided that it was genes (another "way"), instead of parents, that should be given credit for correcting the disturbed children in his Guatemalan study? Why is it that that "way" can't also apply to the Korean study? Couldn't the genome of the Korean children "unfold", and smooth out the wrinkles of their early years, just like it did with the children in the Guatemalan study? How much credit should these "nurturant" parents be given for the recovery of these Korean orphans? I mean, obviously some credit just for being parents, but the question is was there anything unique about their parenting to warrant a credit to the healing powers of the environment?

Of course its incoherent to say it's "just the genes" reviving these kids, It's just that the genes would have expressed themselves in a similar way in a broad range of unspectacular environments. If these Korean children were placed into homes where they were hardly fed and beaten and screamed at daily, it's likely they would retain some of that early trauma into adulthood. But it would seem that most garden-variety upbringings would have had the same amazing effect of healing these traumatized war orphans. It's because of this, the sufficiency of the "good-enough" parent, something Dr. Kagan himself was in part responsible for demonstrating, that makes his argument all the more insulting, fraudulent, and hypocritical.

But the inconsistency between Dr. Kagan's work and his manner of attack on Judith Harris grows ever wider. Dr. Kagan, Following the same ignoble formula of the radical environmentalists of the 70's, actually race baits Mrs. Harris in an attempt to stigmatize her hereditarian approach:

Perhaps the most serious source of vulnerability in your position comes from the fact that children from different cultures behave very differently even before peers have had a chance to have a serious effect. The work of the Whitings has proved this point to the satisfaction of most scientists. I trust that you will not claim that rural Mexican children are more nurturant and less aggressive than New England children because of heredity.

Most serious source of vulnerability? I'm not too sure what part of this is hazardous to Mrs. Harris theory. What I do know is that Dr. Kagan shamelessly used a racially charged comparison of differences to cast a shadow on Mrs. Harris. So what if "rural" [Is "rural" a codeword for more Amerindian?] Mexican children are more nurturant and less aggressive than "New England" [Ok, I know what this one is code for.] children? The more important question is are rural Mexican adults more nurturant and less aggressive than European adults? If not, then how is this damaging to Mrs. Harris theory? If the non-aggressive infants learn to be non-aggressive from their parents but grow up to be aggressive anyway, then it would seem, in perfect alignment with Mrs. Harris theory, that the early parental influence was negated by "something", be it peers, genes or what-have-you. But, even if the non-aggressive babies have non-aggressive parents and grow-up to be non-aggressive adults, then perhaps this is because, like Mrs. Harris argues, the child didn't learn to be non-aggressive, but simply picked up the same temperament genetically. But surely, as Dr. Kagan suggests, it would be irresponsible, or possibly even absurd, for Mrs. Harris to propose that this principle of inheritance could operate racially. . .wouldn't it?!! Continuing from Wright:

"Shaken, perhaps [by the results of the Guatemalan study], but Kagan, like the rest of the psychological profession , continued to believe in the environment as the most important molder of personality. His epiphany came fifteen years later, when he was working in Boston on a longitudinal study of infants, observing them from seven to twenty-nine months, with the aim of assessing the effectiveness of day-care. The group was made up of fifty-three Chinese-American infants and sixty-three Caucasian children. Part of the entire group had from the age of four months attended an experimental day-care center set-up for the study, part had attended other day-care centers, and part had been raised at home.

In the course of the experiment, Kagan noticed something unanticipated. The Chinese children, little more than babies, whether attending day-care or raised at home, were consistently more fearful and inhibited than the Caucasians. The differences were obvious. The Chinese children stayed close to their mothers and were quiet and generally apprehensive, while the Caucasians were talkative, active, and "prone to laughter". These characteristics were confirmed by the mothers as typical of their children's behavior at home as well. In addition, the researchers discovered that the Chinese tots had less variable heart rates than the Caucasians. Kagan could not avoid the clear evidence of an innate difference between the two groups of infants. It is ironic that this scientist's conversion to a biological-genetic view came along the lines of racial differences. Kagan was a political liberal who only three years earlier had been one of the most vociferous critics of Arthur Jensen's theories on the heritability of IQ, theories that he and most everyone else denounced as racist. Now he was publishing his observation of fundamental personality differences between racial groups. When we conversed in Harvard office many years later, I asked Kagan if there had been an uproar similar to the one Jensen provoked.

He smiled. "We got no flak on the Chinese paper All the reports of the book were about our day-care findings. Everyone ignored the fact that the Chinese children were different. I think it was because they were Asians, and Asians do well. If they would have been black we probably would have gotten flak."

Asked if it was dismaying for him, an unwavering liberal, to observe inherent racial differences, Kagan snapped, "Nature doesn't care what we want." More reflectively, he added, "I wasn't so much dismayed at my observations of the Chinese kids. . .I was a little bit saddened to see the power of biology."

Moving aside from the obvious hypocrisy of Dr. Kagan's underhanded race-baiting on Mrs. Harris (which she quickly calls him on), and his dishonest arguments which selectively omit evidence from his own scientific work, I must bask in the importance of this reluctant and confused, yet brilliant hereditarian bastard. I don't know if Kagan ever came around to acknowledging the heritability of intelligence, but his work demonstrating the heritability of personality is highly regarded and of amazing worth (almost as amazing as his cognitive dissonance :P). It was Jerome Kagan who discovered that personality characteristics which can be observed in early infancy remain relatively stable throughout life. Quoting from my Freshman year Psych textbook (Bernstein & Nash, Essentials of Psychology, Houghton Mifflin Company, 1999):

"Traces of early temperamental characteristics weave their way throughout childhood (Buss, 1995a). Easy infants usually stay easy, and difficult infants often remain difficult (Riese 1986). Timid toddlers tend to become shy pre-schoolers, restrained and inhibited eight-year-olds, and somewhat anxious teenagers (Shwartz, Kagan, & Snidman, 1995)"

The textbook then, to my amazement, goes on to talk about racial differences that start early and end up expressing themselves in the "extended phenotype" of the world's different human cultures:

"...Consider the temperament patterns of Chinese-American and European-American children (Kagan et. al., 1994, Smith & Freedman, 1983). At birth, Chinese-American babies are calmer, less changeable, less excitable, and more easily comforted when upset than European-American babies. . .These temperamental differences between children in different ethnic groups illustrate the combined contributions of nature and nurture. Mayan infants, for example, are relatively inactive from birth. The Zinacantecos, a Mayan group in southern Mexico reinforce this innate predisposition toward restrained motor activity [3] by tightly wrapping their infants and by nursing at the slightest sign of movement (Greenfield and Childs, 1991). This combination of genetic predisposition and cultural reinforcement is culturally adaptive. Quiet infants do not kick off their covers at night, which is important in cold highlands where they live. Inactive infants are able to spend long periods on their mother's back as she works. And infants who do not begin to walk until they can understand some language do not wander into the open fire at the center of the house. This adaptive interplay of innate and cultural factors in the development of temperament operates in all cultures."

How would this innate temperament that Dr. Kagan discovered in East-Asian children (which, with the help of a gene/culture feedback loop, remains relatively stable through-out adulthood) affect the subsequent cultures of East Asia where these genes are expressed? I would like to read more about what Kagan has to say about this, but from what I can find he stays away from the really important implications (lawfulness, social cohesion, etc.), but he does coyly (ctrl-F Kagan) acknowledge the influence that genes and race can have over culture:

"Following these children as they grew up, [Kagan] found that the Caucasian children were much more likely to flare up in toy disputes than the Asian toddlers. In an earlier book he observed that these same differences were reflected in the artistic works and tastes of the two cultures."

To anyone familiar with the school of Human Bio-Diversity the elephant left in the living room after all of this talk of the personality differences between Asians and Caucasians is, 'How does this relate to Rushton's Law of Three'. Rushton's amazingly reliable rule is that any important difference (be it physical, mental, developmental, temperamental,or behavioral) that can be found between Asians and Europeans, a similar relationship will most likely be found between Europeans and Africans. Regardless of the correctness of Rushton's theory for explaining this consistent relationship [4], the data he uses to verify the relationship itself is, to quote Frank Miele of Skeptic Magazine, "...replicable".

So what relationship of personality characteristics did Kagan observe between the Caucasian and Asian children, that we might use for this second comparison?: The Caucasian children were less fearful, less inhibited, more talkative, more active, more prone to laughter, had more variable heart rates, were less calm, were more changeable, more excitable, less easily comforted when upset, and more likely to flare up over toy disputes. With this data in mind, it would be predicted from a Rushtonian model that in a similar experiment using African-American and Caucasian infants, that the African-Americans, no matter how or where they were raised, would be even less fearful, even less inhibited, even more talkative, even more active, even more prone to laughter, have even more variable heart rates, would be even less calm, even more changeable, even more excitable, even less easily comforted when upset, and even more likely to flare up over toy disputes. If Kagan has demonstrated that Asians and Caucasians have different fundamental temperaments, Rushton feels that a proportional relationship can be drawn between Whites and Blacks as well. He writes in Race, Evolution, and Behavior (book/abridged):

"Across ages (24-hour-old infants, children, high-school students, university students, and adults), across traits (activity level, aggressiveness, cautiousness, dominance, excitability, impulsiveness, and sociability), and across methods (archival statistics, naturalistic observations, ratings, and self-reports), data shows that in terms of behavioral restraint Mongoloids average higher than do Caucasoids who, in turn, average higher than Negroids..."

That this pattern shows up in African-Americans by the time of kindergarten (and probably in pre-school), and continues throughout schooling and adulthood, should not be controversial from a sociological stand-point. The question is not whether the pattern of behavior exists (at least it shouldn't be), the question is, is it innate? I guess there is as of yet no smoking-gun to confirm this, just a lot of mixed clues and indirect evidence. I suppose, though, if the differences between Asians and Whites can find support using certain analytic methods, that similar methods might yield similarly satisfying answers to the question regarding Blacks. So how about it, do the familiar personality characteristics of Blacks show up in new-borns, like they do with Caucasians, Asians, and Amerindians?

Rushton tells us, flat-out, that these characteristics show up in "24-hour-old infants", but then fails to provide the infancy data for blacks to prove it. Without any back-up data , and with the first study for blacks that he cites starting at ages "4 to 6 years", I have to assume that Rushton was over-zealously "filling in the blanks" with his "24 hour" boast. Furthermore, I found a "lecture overview" primer on the Internet that provides data indicating that African newborns aren't in Rushtonian proportion to European ones. Outlining the work of Daniel G. Freedman and Nina Freedman (who, by the way, discovered the difference between Asian and Caucasian newborns before Jerome Kagan), the infant temperament check-list describes "African/Aboriginal" neonates as more temperamental than Brown, East Asian, and Amerindian new borns, but somewhat less temperamental than Caucasian ones.

The details of this data I am not familiar with, but right now it's the only data I know of, so I will have to consider that Rushton's particular claim regarding Black new-borns is false.

Despite that, I refuse to ignore the forest for the trees. Regardless of newborn behavior, it is highly unlikely that later Black temperament is, like Asians and Amerindians, less volatile than Whites (in a statistically innate sense), it is also highly unlikely that Black temperament is exactly the same, in statistical expression, as Whites [5]. I think the universality and ineluctability of black personality points to an innate temperament that is, on average, somewhat to substantially more volatile than that of whites. If anyone ever conducts a Kagan style experiment on the behavior of infants of European and African descent, I will gladly wager a large sum of money and a beer on the outcome.

[1] I know. . . no one actually considers themselves a "hereditarian" or an "environmentalist" for that matter; people with as disparate opinions as Rushton and Lewontin all consider themselves to be gene/environment interactionists. But since everyone considers themselves within the framework of the same reasonable position, itís probably more useful to use informative, if not rigidly literal, labels. Though I have noticed others in our camp, such as Bouchard, Jensen, Scarr, Eysenck, etc. refer to themselves as Galtonians, in an homage to the genius and his particular school of hereditary understanding, so I just might stick with that term.

[2] Good-enough parents = "...most parents are "good enough" at child rearing to support their children becoming the best they can be". (this quote comes from Sandra Scarr's tribute to Arthur Jensen , as published in the academic journal Intelligence)

[3] Wait a minute Dr. Kagan, did I read that right, or did my text-book just suggest that these "rural" Mexicans were low in aggression and highly nurturant due to their innate predispositions?! . . .More nurturant, I trust, than "New England" children. Like your Chinese children, this is, in fact, the case (Brazelton, T.B., Robey, J.S., & Collier, G.A. (1969). Infant development in the Zinacanteco Indians of Southern Mexico. Pediatrics, 44, 274-290.). Looks like somebody owes Mrs. Harris an apology.

[4] It should also be noted that both Kagan, and the authors of my quoted textbook use Rushton's same evolutionary theory (i.e. the coldness of their adaptive environments) to explain why Asians and Amerindians have more inhibited and less aggressive personalities. And as Razib has pointed out before, the distinguished neuro-psychologist William Calvin uses this Rushtonian explanation to account for racial differences as well. These differences are grounded in fact and there really doesn't appear to be a better or more comprehensive explanation for them.

[5] From an evolutionary perspective it makes little sense that relatively isolated populations will be completely identical in polymorphic frequency. That Africans, Asians, and Europeans are, recognizably, very dissimilar morphologically is the first common-sense clue that their relative isolation has resulted in separate paths of selection pressure. The modern consensus is that there is no trait which does not have some degree of heritability, which means that there is a lot of variation between individuals. But, it is undeniable that populations differ in polymorphic gene frequencies as well. These sub-populations are not all mutually dissimilar, but all share similar statistical relationships with other populations, and together can be viewed as super-clusters, or continental populations, or what I prefer to call by the old familiar term that always reliably described the intuitively discerned branches of this global family-tree: races. World sub-populations all differ from eachother in statistical variations of traits, and it might make sense to you that these super-clusters would be the most differentiated. So which populations are the most differentiated? You might think it is the isolated Amerindians, but Cavalli-Sforza's work shows the largest genetic split is between Sub-Saharan Africans and the rest of the world. This is why we would expect, as a null hypothesis, that the distributed differences between people of European and African descent should be greater than zero across most measurable traits. And, the thing is, they do, in fact, differ to one degree or another on most measurable traits. . .but the question is, are they different for genetic reasons? I think in many cases a genetic hypothesis of some sort, especially for traits apparent in a population that are universal, persistent, and have no clear environmental explanation, is a theoretically prudent one to consider.

Posted by Jason Malloy at 04:03 AM

I would characterize your excellent essay as showing that Leftist ideology destroys good science among psychologists.

I have a fully referenced article making a similar point below:


Posted by: John Ray at March 3, 2003 06:22 AM

A quick thanks for the excellent site, which I discovered recently.

Though I'm not an academic, my thinking has been influenced by Jerome Kagan's work on temperament. So it was something of a shock to learn, just a few days ago, of his opposition to Harris's book. I wonder if this is simply a case of a scientist not wanting to accept unanticipated implications of his work.

Posted by: C. McClellan at March 3, 2003 06:55 AM

I'll take your wager (the beer part, anyway). I believe that the differences Rushton sees in group behavior for his racial triad have more to do with hormone levels later in life than with the differences in infant behavior observed by Kagan. It's quite plausible that black infants would be less temperamental than white ones, and throwing out the Freedmans' data just because of the overall predictive power of Rushton's model seems foolish.

Posted by: bbartlog at March 3, 2003 08:24 AM

John Ray writes...
>I would characterize your excellent essay as showing that Leftist ideology destroys good science among psychologists

Quite unlike the salubrious effects of right-wing ideology on the sciences. Hah.
All ideology in is the nemesis of science, which is best conducted by those without axes to grind (left or right...)

Posted by: bbartlog at March 3, 2003 08:29 AM


I mostly agree with your hormone theory. Blacks have greater androgen receptivity and 3-19% more testosterone than Whites (who in turn have greater androgen receptivity and more testosterone than Asians). The link between libido, aggression, and testosterone is fairly clear, and its role in these differences could plausibly be substantial.
But, I see your beer, and I raise you a bottle of whiskey. Different hormone levels do not begin at puberty. A good place to start is the difference between little boys and little girls. Boys and girls aren't equally nurturant and aggressive until puberty, but are, like the Caucasian and Asian newborns, easily differentiated from the time of birth. Little boys "flare up in toy fights" and engage in "rough-and-tumble" play, not b/c they learned to do it, but b/c of perinatal exposure to higher levels of sex hormone. Young females, whether mouse, human or rhesus monkey, exposed to identical levels in their parinatal environment are born with essentially the same temperament as boys. Also interesting increasing a women's testosterone level will increase her spatial ability. But the relationship is not infinitely linear; spatial ability begins to drop after a certain level of increase:

Such studies suggest that although levels of androgen relate to spatial ability, it is not simply the case that the higher the levels, the better the spatial scores. Rather studies point to some optimal level of androgen (in the low male range) for maximal spatial ability. This finding may also hold for men and math reasoning; in one study, low-androgen men tested higher.

In other words, because there is a low "optimum" level of testosterone, it would be expected that Asian males would have the highest spatial ability.

I think the best explanation for why these racial differences begin so young, is that they are linked to different levels of sex hormones and different levels of exposure to sex hormones in the perinatal environment.

Posted by: Jason Malloy at March 3, 2003 11:48 AM

bbartlog writes...
>All ideology in is the nemesis of science...

I think that's too harsh. All hypotheses are shaded by preconceptions, some of which are hard to articulate (and thus potentially 'biases'), but the real importance is whether they 'work' in explaining more with less. Certainly a researcher would do well to not let biases override their appreciation of the facts (eg, Lysenko, Freud, 5-year plans...), because it is best to have them on your side, but I suspect a little zealotry is behind a lot of good scientific work: it can motivate and clarify better than ideological indifference.

Good science is objective, but often not neutral, and trying to be objective and neutral leads to anemic research.

Posted by: Eric at March 3, 2003 11:57 AM

on science, the lens is always a little fuzzy. who hasn't jigged a "line of best fit" by deleting one data point that seems wack :) but, if everyone is looking with a fuzzy lens, then you should be able to figure out the commonalities, since the fuzz will be different...reproducibility....

Posted by: razib at March 3, 2003 12:06 PM

Hm... I don't like whiskey, so I think we'll have to stick with beer. Anyway, yes, the evidence you list is interesting and I don't claim to be certain of my position. But I can certainly construct all sorts of theories to explain why black children would tend to be less temperamental than white ones; I am also influenced in part by my own experiences, which include a lot of blond-haired blue-eyed hellions and a handful of shy black children (sample size, of course, too small to be terribly relevant). Also, just so we're clear - I am talking about infants and toddlers (age

Posted by: bbartlog at March 3, 2003 12:46 PM