A Response to Professor Murray
By Martin Heilweil
Posted On October 18, 2005

(Ed. Note: Charles Murray published an article last week in the Wall Street Journal entitled, "The Inequality Taboo - It's time to start talking about differences between groups of people." Here Martin Heilweil offers his response.)

I got my PhD in 1973, working in a preschool intervention project, which focused on increasing IQ scores, and worked directly with IQ data, and with group differences. Even in 1969, to a lowly grad student, the question of racial differences in mean or group scores was contentious, and I have followed the debate for these same thirty years, through various career changes, with some of the same psychometric skills mentioned by Professor Murray and followed some of the same literature.

The New Inquisition's suppression of debate is well described, as we modern Galileos fear the auto da fe of our TIAA-CREF accounts. Odd how the Left has become the New Church of Ignorance, but in ignorance, the left and right circle, as Einstein said, the universe is infinite but circular, and converge into intolerance.

Group differences are indisputable, and inconvenient for social egalitarians, and susceptible to adverse politicization, but the scientist works in a more abstract domain, and so what does the egalitarian scientist (even a conservo Rightist, me) say, with great trepidation, speaking as a long-ago short-term worker-grunt, to one of the masters?

Sex differences are uninteresting; for starters, we seem to be on top of this on male mental performance as outliers on both ends of the spectrum; if visua-spatial brain functioning explains all, then by golly bully for our understanding. Whether this correlates with testosterone or not, is an empirical question, so far these are unrelated causal factors, and the linkage of testosterone and higher outlier fractions is also to be discussed empirically.

I have always enjoyed the science of non-human sex differences (called infra-human), our traditional socialization models do not quite explain monkey differences, and among the lions the females are the hunter gatherers, the males are the pimps and rentiers, living off the working classes, and perhaps among sharks there are no predatory-vegetarian differences at all; so now we have a new sex-related variable, hunter-gatherer vs farmer/ grubber vs social exploiter. Other infrahuman species are grazers and foragers male and female both.

Also in the animal domain we can alter sex hormones, we have this naturalistically in prenatal androgenization of female fetus, which produces male self-image and behavior. Voila! There is an enormous literature on this.

Racial differences are more interesting. Even in 1969 it was obvious that race was a continuum from racial mixing over the few centuries of Africans in this country, that 'African' was as much an artifact as "Negro" because 'racial' (tribal) distinct attributes were specifically known by slavers, the west African slavers of the Americas, and so "Africa' is as unitary, that is, imaginary, as "Caucasia." In Caucasia, we would hardly compare high performing Ashkenazi (European) Jews with lower performing hillbilly Caucasians. The critical experiment of 'racialness' genetically as correlated with G factor, apparently has not yet been done, for all skin colors.

The pro-difference camp has avoided the slavery controversy by looking at modern Africa's racial scores, and the same racial differences are found, to which the anti-difference group responds by pointing out the millennium of human despoilation by the east African slave trade, which was before and after our briefer America's enslavement of Africans.

I am fascinated also in these discussions by the exclusion of Asian superior scores, or perhaps "Asian" is better, by all measures they score higher, so black and white are only part of the story, the contentious but intellectually boring part. Of course in America Asians are the 'model minority' and the children hate this obligation upon them.

But back to the social egalitarian conservo (I also spend time with lawyers these days, including race-based questions, by coincidence), for whom governmental quotas are anathema: Unlike Harvard level math, or Nobel performance, most of what we do, in selling shoes, fixing cars, performing retinal surgery, or the margin brokering of selling Chrysler to Daimler, the overlap in racial intelligence scores is very large, so that most folks in one distribution overlap with most folks in the other distribution, so the question of G factor becomes an empirical one, what percentage of the distributions overlap, and what is the role of 'g' in life's success or political allocation of public resources.

I remain a scientist at heart and core, and enjoy our ever increasing understanding of the human creature. Science and falsifiable hypotheses though are a small part of how we live.

One other point has been made and lost; as we become a more and purer meritocratic society, or perhaps 'if' not 'as,' such racial or other between-group innate differences as exist will only become more significant, and these groups will diverge evermore. On the other side, many say that intelligence is a vector, not a scalar.

Martin Heilweil lives in New York City.

= = = =

Addendum (after reviewing Professor Murray’s original article at American Enterprise Institute, at
http://www.aei.org/publications/filter.all,pubID.23075/pub_detail.asp.)

I should like to offer some responses and comments to professor Murray’s discussion of intergroup differences, as presented in today's wall street journal's opinionjournal.com.

I see that these are a shorter version of his longer and more academic posting at AEI, which addresses some of the issues i thought were left out in the OJ piece.

I do not differ with his facts of intergroup differences, but discuss them in a different way.

As an aside, prof Murray discusses factor analysis in determining g (general intelligence) factor, discussing what is de facto principal components analysis (major variance) vs. forced equalization of factors (I used a version called Varimax); forced equalization is a more interesting analysis which has the effect of partitioning the variance, even while suppressing the power of the major factor or principal components, so it depends what question you are asking, I suppose, or avoiding.

In my own very long ago grad student and phd work in an intervention preschool oriented to raising IQ data (beyond test artifact of regression to the mean by reverse cherry picking the low-scorers for admission into the program), I worked on behavioral data, social interactions, rather than intelligence scores, but was immersed in the environment and have followed the debate.

I did respond to the OJ piece, as attached here, and then while doing homework found the 'long' version here at AEI which addresses some of the issues I raised as omitted from OJ.

I offer briefly in preview of my attached comments, that I think the tripartite black - white - asian discussion should be moved much more forward in our future discussions, this avoids or evades the political contentiousness of the US/ Western Civ black v. white discussions, and adds a greater scientific neutrality to the overall discussion, i also like infrahuman data and think that the hormonal influence, as shown by anomalies, is also important.

I offer other demurrals if not critiques also.

The huge discussion of identical twins, mixed genetic heritage, was editorially excised I see, for better or worse, but the heritability of IQ, which argues in favor or durable intergroup differences, was i think a critical and unfortunate omission, and I suspect that gender differences among the Asians are lesser than among the other two groups, although this is as yet (i) unknown and (ii) unallocated as to biology or culture.

I think also that dysgenic demographics does not account for very much variance in the end, as I conclude, equalization of opportunity only magnifies innate inequality, and so as a social policy our job, beyond intellectual honesty, is to 'teach the test' as we are doing to bring up the lower portion of the black distribution, as we are doing; the test we are to teach is the 'test of life' namely the educational and related social and psychological skills to bring the lower portions of each distribution up to its higher potential, as a social policy,

To use the athletic metaphor, the gifted athlete will always be better than the less gifted, but we train the less gifted as social policy; thus we 'teach the test'.

E-mail this article to a friend.