RICHARD C. LEWONTIN RACE AND INTELLIGENCE
In the Spring of 1653 Pope In programmatic conclusion is that nocent X condemned a pernicious A controversial doctrine that there is no use in trying to re heresy which espoused the doc genetic factors outueigh envi move the difference in I.Q. by trines of "total depravity, irre- ronment in producing I.Q. dif education since it arises chiefly sistable grace, lack of free will, ferences belts ecn racial groups from genetic causes and the beat predestination and limited atone has created a tempest in the thing that can be done for black ment." That heresy was Jansen educational and psychological children is to capitalize on those ism and its author was Cornelius groves of academe. The doc skills for which they are biologi Jansen, Bishop of Ypres. trine has been labeled “ jen• cally adapted. Such a conclusion In the winter of 1968 the same senitm," after its proponent. is so clearly at variance with the doctrine appeared in the "Har Arthur R. Jensen, professor of present egalitarian consensus and vard Educational Review.” That educational psychology at the so clearly smacks of a racist elit doctrine is now called "jensen- University of California. It has ism, whatever its merit or moti ism” by the "New York Times also attracted sttde attention in vation, that a very careful analy Magazine” and its author is Ar the general media because it sis of the argument is in order. thur R. Jensen, professor of edu appears to challenge the envi The article begins with the cational psychology at the Uni ronmental deprivation theory- pronouncement: "Compensatory versity of California at Berkeley. on uhich federal compensatory education ha* been tried and it It is a doctrine as erroneous in education programs are based. apparently has failed.” A docu the twentieth century as it was In this article. Richard C. mentation of that failure and a in the seventeenth. I shall try l.euonlin, professor of biology definition of compensatory edu to play the Innocent. at the University of Chicago, cation arc left to the end of the Jensen's article, “How Much dissects the Jensen paper uhich article for good logical and peda Can We Boost I.Q. and Scholastic precipitated the grossing con gogical reason*. Having caught Achievement?" created such a fu troversy last year. Professor our attention by whacking us over ror that the "Review" reprinted l.euonlin s conclusion: Jensen the head with a two-by-four, like it along with critiques by psy is strong! that famous trainer of mules, chologists, theorists of education Jensen then asks: and a population geneticist under "What has gone wrong? In the title "Environment. Heredity neous conclusions. I shall say other fields, when bridges do not and Intelligence." The article little or nothing about the cri xtand, when nircraft do not fly. first came to my attention tiques of Jensen's article, which when machine* do not work, when, at no little expense, it was would require even more space when treatments do not euro, de sent to every member of the Na to criticize than the original ar spite all the conscientious effort* tional Academy of Sciences by ticle itself. on the part of many persons to the eminent white Anglo-Saxon m u p o s i t i o n make them do so. one begins to inventor, William Shockley, as question the basic assumptions, part of his continuing campaign Jensen’s argument consists es to have the Academy study the sentially of an elaboration on two principles, theories, and hypothe ses that guide one’s efforts. Is it effects of inter-racial mating. It incontrovertible facts, a causative time to follow suit in education?*’ i* little wonder that the "New explanation and a programmatic York Times” found the matter conclusion. The two fact* are Who can help but answer that newsworthy, and that Professor that black people perform, on the last rhetorical question with a re Jensen has surely become the average, more poorly than whites sounding “Yes”? What thought most discussed and least read on standard I.Q. tests, and that ful and intelligent person can essayist since Karl Marx. I shall special programs of compensatory avoid being struck by the intel try, in this article, to display education so far tried have not lectual and empirical bankruptcy Professor Jensen's argument, to had much success in removing of educational psychology as it show how the structure of hi* ar this difference. His causative ex is practiced in our mass educa gument is designed to make his planation for these facts is that tional systems? The innocent point and to reveal what appear I.Q. is highly heritable, with most reader will immediately fall into to be deeply embedded assump of the variation among individ close sympathy with Professor tions derived from a particular uals arising from genetic rather Jensen, who, it seems, is about world view, leading him to erro- than environmental sources. His to dissect educational psychology 2 and show it up as a prc-scicntific jumble without theoretic coher ence or prescriptive competence. But the innocent reader will be wrong. For the rest of Jensen’s article puts the blame for the failure of his science not on the scientists but on the children. According to him, it is not that his science and its practitioners have failed utterly to understand human motivation, behavior and development but simply that the damn kids arc ineducable. The unconscious irony of his metaphor of bridges, airplanes and machines has apparently been lost on him. The fact is that in the twentieth century bridges do stand, machines do work and airplanes do fly, be cause they arc built on clearly understood mechanical and hy drodynamic principles which even moderately careful and intelligent engineers can put into practice. In the seventeenth century that of Jensen’s thesis which can only testing for so-called “deprived” was not the case, and the general be appreciated when seen in this children, since it is supposed that opinion was that men would never light. they have developed in a sub succeed in their attempts to fly Having begun with the assump culture that does not prepare because flying was impossible. tion that I.Q. cannot be equal them for such tests. What role Jensen proposes that we take the ized. Jensen now goes on to why does this “environmentalist" ar same view of education and that, not. He begins hi* investigation gument play in Jensen’s thesis? in the terms of his metaphor, fal with a discussion of the “nature Is it simply evidence of his total len bridges be taken as evidence of intelligence," by which he fairness and objectivity? No. of the unbridgeability of rivers. means the way in which intelli Jensen has seen, more clearly The alternative explanation, that gence is defined by testing and than most, that the argument of educational psychology is still in the correlation of intelligence test the specific cultural origins of the seventeenth century, is ap scores with scholastic and occupa I.Q. testing and especially the parently not part of his philos tional performance. A very strong high correlation of these tests ophy. point is made that I.Q. testing with occupational status cuts This view of technological fail was developed in a western indus both ways. For if the poorer per ure as arising from ontological trialized society specifically as a formance of blacks on I.Q. tests rather than epistemological sourc prognostication of success in that has largely genetic rather than es is a common form of apology society by the generally accepted environmental causes, then it fol at many levels of practice. Any criteria. He makes a special point lows that blacks are also gene one who has dealt with plumbers of noting that psychologists' no tically handicapped for other high will appreciate how many things tions of status and success have status components of Western "can't be fixed" or “weren’t a high correlation with those of culture. That is. what Jensen is meant to be used like that.” the society at large, so that it is arguing is that differences be Physicists tell me that their fail entirely reasonable that testa tween cultures are in large part ure to formulate an elegant gen created by psychologists will cor genetically determined and that eral theory of fundamental par relate highly with conventional I.Q. testing is simply one mani ticles is a result of there not be measures of success. One might festation of those differences. ing any underlying regularity to think that this argument, that In this light we can also under be discerned. How often men, in I.Q. testing is “culture bound,” stand his argument concerning their overweening pride, blame would militate against Jensen’s the existence of "general intelli nature for their own failures. This general thesis of the biological gence” as measured by I.Q. tests. profcssionalist bias, that if a and specifically genetical basis of Jensen is at some pains to con problem were soluble it would I.Q. differences. Indeed, it is an vince his readers that there is a have been solved, lies at the basis argument often used against I.Q. single factor, g, which, in factor Marti, 1970 Bulletin of the Atomic Scientutt 3 analysis of various intelligence is that true merit, given equality ally caused deaf-mutism is han testa, accounts for a large frac of opportunity, act as a basis for dicapped to different extents in tion of the variance of scores. the natural assorting process.” cultures requiring different de The existence of such a factor, What a world view is there re grees of verbal performance, yet while not critical to the argument, vealed! The most rewarding his disorder did not have an en obviously simplifies it, for then places in society shall go to those vironmental origin. I.Q. testa would really be testing with "true merit” and that is Jensen first dispenses with the for '‘something" rather than just the best we can hope for. Of question of developmental stabil being correlated with scholastic course, Professor Jensen is safe ity of I.Q. Citing Benjamin and occupational performance. since, despite the abject failure of Bloom's survey of the literature, While Jensen denies that intelli educational psychology to solve he concludes that the correlation gence should be reified, he comes the problems it has set itself, that between test scores of an individ perilously close to doing so in his failure does not arise from lack ual at different ages is close to discussion of g. of "true merit" on the part of unity after the age of eight. The Without going into factor ana psychologists but from the natu inference to be drawn from this lysis at any length, I will point ral intransigence of their human fact is, I suppose, that it is not out only that factor analysis docs subjects. worth trying to change I.Q. by not give a unique result for any Having established that there training after that age. But such given set of data. Rather, it gives are differences among men in the an inference cannot be made. All an infinity of possible results degree to which they are adapted that can be said is that, given among which the investigator to higher status and high satis the usual progression of educa chooses according to his tastes faction roles in Western society, tional experience to which moat and preconceptions of the models and having stated that education children are exposed, there is suf he is fitting. One strategy in fac has not succeeded in removing ficient consistency not to cause tor analysis is to pack as much these differences, Jensen now any remarkable changes in I.Q. weight as possible into one fac moves on to their cause. He raises That is, a child whose education tor, while another is to distribute the question of “fixed” intelli al experience (in the broad sense) the weights over as many factors gence and quite rightly dismisses may have ruined his capacity to as possible as equally as possible. it as misleading. He introduces perform by the age of eight is Whether one chooses one of these us here to what he regards as not likely to experience an en or some other depends upon one’s the two real issues. "The first vironment in his later years that model, the numerical analysis issue concerns the genetic basis will do much to alter those ca only providing the weights ap of individual differences in in- pacities. Indeed, given the pres propriate for each model. Thus, telligence; the second concerns ent state of educational theory the impression left by Jensen that the stability or constancy of the and practice, there is likely to be factor analysis somehow natural I.Q. through the individual’s life a considerable reinforcement of ly or ineluctably isolates one fac time." Jensen devotes some early performance. To say that tor with high weight is wrong. three-quarters of his essay to on children do not change their I.Q. is not the same as saying they - TRUE MERIT”! attempt to demonstrate that I.Q. is developmental!)’ rather stable, cannot. Moreover, Jensen is In the welter of psychological being to all intents and purposes curiously silent on the lower cor metaphysics involving concepts fixed after the age of eight, and relation and apparent plasticity of "crystallized" as against that most of the variation in I.Q. of I.Q. at younger ages, which is “fluid" intelligence, “generalized” among individuals in the popula after all the chief point of intelligence, "intelligence” as op tion has a genetic rather than en Bloom's work. posed to "mental ability," there vironmental basis. Before looking THE GENETIC ARGUMENT is some danger of losing sight of in detail at some of these argu Jensen's main point: I.Q. tests ments. we must again ask where The heart of Jensen’s paper is are culture bound and there is he is headed. While Jensen ar contained in his long discussion good reason that they should be, gues strongly that I.Q. is “cul of the distribution and inheri because they are predictors of ture bound,” he wishes to argue tance of intelligence. Clearly he culture bound activities and val that it is not environmentally de feels that here his main point is ues. What is further implied, of termined. This is a vital distinc to be established. The failure of course, is that those who do not tion. I.Q. is "culture bound" in compensatory education, the de velopmental stability of I.Q., the perform well on these tests are the sense that it is related to per less well suited for high status obvious difference between the and must paint barns rather than formance in a Western industrial performance of blacks and whites pictures. We read that "We have society. But the determination can be best understood, he be to face it: the assortment of per of the ability to perform cultural lieves. when the full impact of sons into occupational roles sim ly defined tasks might itself be the findings of genetics is felt. ply is not ‘fair' in any absolute entirely genetic. For example, a In his view, insufficient attention sense. The best we can hope for person suffering from a genetic has been given by social scientists 4 to the findings of geneticists, and After establishing the near quence of environments through I must agree with him. Al normality of the curve of I.Q. which the organism has passed though there arc exceptions, there scores, Jensen goes directly to a during its development. For some has been a strong professional discussion of the genetics of con traits the variations in environ bias toward the assumption that tinuously varying characters. Ho ment have little effect, so that human behavior is infinitely begins by quoting with approba once the genotype is known, the plastic, a bias natural enough in tion E. L. Thorndike's maxim: eventual form of the organism men whose professional commit “In the actual race of life, which is pretty well specified. For other ment is to changing behavior. It is not to get ahead, but to get traits, specification of the gene is as a reaction to this tradition, ahead of somebody, the chief de tic makeup may be a very poor and as a natural outcome of his termining factor is heredity.” predictor of the eventual pheno confrontation with the failure of This quotation along with many type because even the smallest educational psychology, that Jen others used by Jensen shows a environmental effects may affect sen’s own opposite bias flows, as style of argument that is not con the trait greatly. But for all I have already claimed. genial to natural scientists, how traits there is a many-many rela The first step in his genetical ever it may be a part of other tionship between gene and char argument is the demonstration disciplines. There is a great deal acter and between environment that I.Q. scores are normally dis of appeal to authority and the ac ond character. Only by a spe tributed or nearly so. I am un ceptance of the empirically un cification of both the genotype able to find in his paper any substantiated opinions of eminent and the environmental sequence explicit statement of why he re authorities as a kind of relevant can the character be predicted. gards this point os so important. evidence. We hear of "three emi Nevertheless, traits do vary in From repeated references to Sir nent geneticists,’' or “the most the degree of their genetic deter Francis Galton, filial regression, distinguished exponent [of gene mination and this degree can be mutant genes, a few major genes tical methods]. Sir Cyril Burt." expressed, among other ways, by for exceptional ability and assor- The irrelevance of this kind of their heritabilities. tative mating, it gradually emerg argument is illustrated precisely The distribution of character es that an underlying normality by the appeal to E. L. Thorndike, values, say I.Q. scores, in a popu of the distribution appears to who, despite his eminence in the lation arises from a mixture of a Jensen as an important conse history of psychology, made the large number of genotypes. Each quence of genetic control of I.Q. statement quoted by Jensen in genotype in the population does He asks: . is intelligence 1905, when nothing was known not have a unique phenotype cor itself — not just our measure about genetics outside of at responding to it because the dif- ments of it — really normally tempts to confirm Mendel's pa ferent individuals of that geno distributed?” Apparently he be per. Whatever the eventual truth type have undergone somewhat lieves that if intelligence, quite of his statement turns out to be. different environmental sequenc aside from measurement, were Thorndike made it out of his ut es in their development. Thus, really normally distributed, this ter ignorance of the genetics of each genotype has a distribution would demonstrate its biological human behavior, and it can only of I.Q. scores associated with it. and genetical status. Aside from be ascribed to the sheer preju Some genotypes are more com a serious epistemological error in dice of a Methodist Yankee. mon in the population so their distributions contribute heavily volved in the question, the basis IIERITABILITY for his concern is itself erroneous. to determining the over-all dis There is nothing in genetic theory' To understand the main gene tribution, while others are rare that requires or even suggests tical argument of Jensen, we and make little contribution. The that a phenotypic character must dwell, as he does, on the total variation in the population, should be normally distributed, concept of heritability. We can as measured by the variance, re even when it is completely deter not speak of a trait being molded sults from the variation between mined genetically. Depending by heredity, as opposed to en the mean I.Q. scores of the dif upon the degree of dominance of vironment. Every character of an ferent genotypes and the varia genes, interaction between them, organism is the result of a unique tion around each genotypic mean. frequencies of alternative alleles interaction between the inherited The heritability of a measure at the various gene lod in the genetic information and the se ment is defined as the ratio of population and allometric growth relations between various parts of the organism transforming pri mary gene effects, a character may have almost any uni-modal distribution and under some cir cumstances even a multi-modal one. March 1970 Bulletin o f the Atomic Scientist* 6 the variance due to the differ The estimation of heritability much more numerous cases of to ences between the genotypes to of a trait in a population depends tally unrelated children raised in the total variance in the popula on measuring individuals of the same family. Neither of these tion. If this heritability were 1.0, known degrees of relationship to cases is completely reliable, how it would mean that all the varia each other and comparing the ever. since twins separated from tion in the population resulted observed correlation in the trait birth are nevertheless likely to be from differences between geno between relatives with the theo raised in families belonging to types but that there was no en retical correlation from genetic the same socio-economic, racial, vironmentally caused variation theory. There are two difficul religious and ethnic categories, around each genotype mean. On ties that arise in such a proced while unrelated children raised the other hand, a heritability of ure. First, the exact theoretical in the same family may easily be 0.0 would mean that there was correlation between relatives, ex treated rather more differently no genetic variation because all cept for identical twins, cannot than biological sibs. Despite these individuals were effectively iden be specified unless there is de difficulties, the weight of evidence tical in their genes, and that all tailed knowledge of the mode of from a variety of correlations be the variation in the population inheritance of the character. A tween relatives puts the herita- arose from environmental differ first order approximation is pos bility of I. Q. in various human ences in the development of the sible, however, based upon some populations between .6 and A. different individuals. simplifying assumptions, and it For reasons of his argument, Jen Defined in this way, hcritabili- is unusual for this approximation sen prefers the higher value but ty is not a concept that can be to be badly off. it is not worth quibbling over. applied to a trait in general, but A much more serious difficulty Volumes could be written on the only to a trait in a particular pop arises because relatives are corre evaluation of heritability esti ulation, in a particular set of en lated not only in their heredities mates for I.Q. and one can find vironments. Thus, different pop but also in their environments. a number of faults with Jensen’s ulations may have more or less Two sibs are much more alike treatment of the published data. genetic variation for the same in the sequence of environments However, it is irrelevant to ques character. Moreover, a charac in which they developed than are tions of race and intelligence, and ter may be relatively insensitive two cousins or two unrelated per to questions of the failure of com to environment in a particular sons. As a result, there will be pensatory education, whether the environmental range, but I* ex an overestimate of the heritabil heritability of I.Q. is A or .8, so tremely sensitive outside this ity of a character, arising from I shall accept Jensen's rather range. Many such characters are the added correlation between high estimate without serious ar known, and it is the commonest relatives from environmental si gument. kind of relation between charac milarities. There is no easy way The description I have given ter and environment. Finally, to get around this bias in general of heritability, its application to some genotypes are more sensi so that great weight must be put a specific population in a specific tive to environmental fluctuation on peculiar situations in which set of environments and the dif than others so that two popula the ordinary environmental cor ficulties in its accurate estima tions with the same genetic vari relations arc disturbed. That is tion arc all discussed by Jensen. ance but different genotypes, why so much emphasis is placed, While the emphasis he gives to and living in the same environ in human genetics, on the hand various point* differs from mine, ments. may still have different ful of cases of identical twins and his estimate of heritability is heritabilities for a trait. raised apart from birth, and the on the high side, he appears to have said in one way or another just about everything that a ju dicious man can say. The very judiciousness of his argument has been disarming to geneticist* es pecially, and they have failed to note the extraordinary conclu sions that are drawn from these reasonable premises. Indeed, the logical and empirical hiatus be tween the conclusions and the premises is especially striking and thought-provoking in view of Jen sen's apparent understanding of the technical issue*. The first conclusion concerns the cause of the difference bc- 6 tween the I.Q. distributions of ality. If that is the only con performance are. Thus, Jensen blacks and whites. On the aver clusion he means to come to, Jen compares blacks with American age, over a number of studies, sen has just wasted a great deal Indians whom he regards as far blacks have a distribution of I.Q. of space in the "Harvard Educa more environmentally disadvan scores whose mean is about 15 tional Review." But of course, taged. But a priori judgments of points—about 1 standard devi like all cant, the special langu the importance of different as ation—below whites. TakinR in age of the social scientist needs pects of the environment are val to account the lower variance of to be translated into common ueless, as every ecologist and scores among blacks than among English. What Jensen is saying plant physiologist knows. My ex whites, this difference means that is: “It is pretty clear, although ample will speak to that point os about 11 per cent of blacks have not absolutely proved, that most well. I.Q. scores above the mean white of the difference in I.Q. between Let us take two completely score (as compared with 50 per blacks and white* is genetic*!.” inbred lines of com. Because cent of whites) while 18 per cent This, at least, is not a trivial con they are completely inbred by of whites score below the mean clusion. Indeed, it may even be self-fertilization, there is no gen black score (again, as compared true. However, the evidence of etic variation in cither line, but to 50 per cent of blacks). If. ac fered by Jensen is irrelevant. the two lines will be genetically cording to Jensen, “gross socio IS IT LIKELYf different from each other. Let economic factors” are equalized us now plant seeds of these two between the tested groups, the How can that be? We have inbred lines in flower pots with difference in means is reduced admitted the high heritability of ordinary potting soil, one seed to 11 points. It is hard to know I.Q. and the reality of the dif of each line to a pot. After they what to say about overlap be ference between the black and the have germinated and grown for a tween the groups after this cor white distributions. Moreover, few weeks we will measure the rection, since the standard de we have seen that adjustment for height of each plant. We will viations of such equalized popu gross socio-economic level still discover variation in height from lations will be lower. From these leaves a large difference. Is it plant to plant. Because each line and related observations, and the not then likely that the difference is completely inbred, the varia estimate of .8 for the heritabil- is genetic? No. It is neither like tion in height within lines must ity of I.Q. (in white populations, ly nor unlikely. There is no evi be entirely environmental, a re no reliable estimate existing for dence. The fundamental error of sult of variation in potting con blacks), Jensen concludes that: Jensen's argument is to confuse ditions from pot to pot. Then heritability of a character within the heritability of plant height . . all we are left with are vari a population with heritability of in both lines is 0.0. But there ous lines of evidence, no one of the difference between two pop will be an average difference in which is definitive alone, but ulations. Indeed, between two plant height between lines that which, viewed altogether, make populations, the concept of herit arises entirely from the fact that it a not unreasonable hypothesis ability of their difference is mean the two lines are genetically dif that genetic factor* are strongly ingless. This is because a vari ferent. Thus the difference be implicated in the average Negro- ance hased upon two measure tween lines b entirely gcnctical white intelligence difference. The ments has only one degree of even though the heritability of preponderance of evidence is. in freedom and so cannot be parti height b 0! my opinion, less consistent with tioned into genetic and environ Now let us do the opposite ex a strictly environmental hypoth mental components. The genetic periment. We will take two hands- esis than with a genetic hypoth basis of the difference between ful from a sack containing seed esis. which, of course, does not two population* l>car* no logical of an open-pollinated variety of exclude the influence of environ or empirical relation to the herit com. Such a variety has lots of ment on it* interaction with gene ability within populations and genetic variation in it. Instead of tic factors." cannot be inferred from it, as I using potting soil, however, we Anyone not familiar with the will show in a simple but realis will grow the seed in vermiculite standard litany of academic dis tic example. In addition, the no watered with a carefully made claimers ("not unreasonable hy tion that eliminating what appear up nutrient, Knop's solution, pothesis.” “doe* not exclude.” a priori to be major environment used by plant physiologists for ”in my opinion”) will, taking this al variables will serve to eliminate controlled growth experiments. statement at face value, find a large part of the environmental One batch of seed will be grown nothing to disagree with since it ly caused difference between the on complete Knop’s solution, but say* nothing. To contrast a populations is biologically naive. the other will hare the concentra "strictly environmental hypoth In the context of I.Q. testing, it tion of nitrates cut in half and, esis” with "a genetic hypothesis assume* thnt educational psy in addition, we will leave out the which . . . doc* not exclude the chologist* know what the major minute trace of zinc salt that b influence of the environment” is sources of environmental differ part of the necessary trace ele to be guilty of the utmost trivi ence between black and white ments (30 parts per billion). Af- March 1970 Bullrtm oj the Atomic Seicntiitt 7 ter several weeks we will measure education for the disadvantaged century the infant mortality rates the plants. Now we will find vari (blacks, chiefly) has failed. The were many times their present ation within seed lota which is explanation offered for the fail level at all socio-economic levels. entirely genetica] since no envi ure is that I.Q. has a high herit Using what was then the normal ronmental variation within lota ability and that therefore the range of environments, the infant was allowed. Thus heritability difference between the races is al mortality rate of the highest so will be 1.0. However, there will so mostly genetical. Given that cio-economic class would haw be a radical difference between the racial difference is genetical, been regarded as the limit below seed lots which is ascribablc en then environmental change and which one could not reasonably tirely to the difference in nutri educational effort cannot make expect to reduce the death rate. ent lewis. Thus, we have a case much difference and cannot close But changes in sanitation, public where heritability within popula the gap very much between health and disease control— tions is complete, yet the differ blacks and whites. I have alrea changes which are commonplace ence between populations is en dy argued that there is no evi to us now but would haw seemed tirely environmental! dence one way or the other about incredible to a man of the seven But let us carry our experiment the genetics of inter-racial I.Q. teenth century—haw reduced to the end. Suppose we do not differences. To understand Jen- the infant mortality rates of "dis know about the difference in the sen’s second error, however, we advantaged” urban Americans nutrient solutions because it was will suppose that the difference well below those of even the rich really the carelessness of our as is indeed genetical. Let it be en est members of seventeenth cen sistant that was involved. We tirely genetical. Docs this mean tury society. The argument that call in a friend who is a very that compensatory education, compensatory education is hope careful chemist and ask him to having failed, must fail? The less is equivalent to saying that look into the matter for us. He supposition that it must arises changing the form of the seven analyzes the nutrient solutions from a misapprehension about teenth century gutter would not and discovers the obvious—only the fixity of genetically deter haw a pronounced effect on pub half as much nitrates in the case mined traits. It was thought at lic sanitation. What compensa of the stunted plants. So we one time that genetic disorders, tory education will be able to ac add the missing nitrates and do because they were genetic, were complish when the study of hu the experiment again. This time incurable. Yet we now know that man behavior finally emerges our second batch of plants will inborn errors of metabolism are from its pre-scientific era is any grow a little larger but not much, indeed curable if their biochemis one’s guess. It will be moat ex and we will conclude that the try is sufficiently well under traordinary if it stands as the difference between the lots is stood and if deficient metabolic sole exception to the rule that genetic since equalizing the large products can be supplied exogen- technological progress exceeds by difference in nitrate level had so uously. Yet in the normal range manyfold what even the most little effect. But. of course, we of environments, these inborn er optimistic might have imagined. would be wrong for it is the miss rors manifest themselves irrespec The real issue in compensatory ing trace of zinc that is the real tive of the usual environmental education does not lie in the her culprit. Finally, it should be variables. That is, even though itability of I.Q. or in the possible pointed out that it took many no environment in the normal limits of educational technology. years before the importance of range has an effect on the char On the reasonable assumption minute trace elements in plant acter, there may be special en that ways of significantly alter physiology was worked out be vironments. created in response ing mental capacities can be de cause ordinary laboratory glass to our knowledge of the underly veloped if it is important enough ware will leach out enough of ing biology of a character, which to do so, the real issue is what many trace elements to let plants are effective in altering it. the goals of our society will be. grow normally. Should educa But we do not need recourse Do we want to foster a society in tional psychologists study plant to abnormalities of development which the “race of life” is "to physiology? to see this point. Jensen says get ahead of somebody” and in Having disposed, I hope, of that “there is no reason to be which "true merit,” be it gen Jensen’s conclusion that the high lieve that the I.Q.’s of deprived etically or environmentally deter heritability of I.Q. and the lack children, given an environment mined. will be the criterion of of effect of correction for gross so of abundance, would rise to a men’s earthly reward? Or do we cio-economic class arc presump higher level than the already want a society in which every tive evidence for the genetic basis privileged children’s I.Q.’s.” It is man can aspire to the fullest mea of the difference between blacks empirically wrong to argue that if sure of psychic and material ful and whites, I will tum to his the richest environment experi fillment that social activity can second erroneous conclusion. The ence we can conceive does not produce? Professor Jensen has article under discussion began raise I.Q. substantially, that we made it fairly clear to me what with the observation, which he have exhausted the environment sort of society he wants. documents, that compensatory al possibilities. In the seventeenth I oppose him. a The Jensen Thesis: Three Comments
1. RACE AND THE GENETICS OF INTELLIGENCE: A REPLY TO LEWONTIN ARTHUR R. JENSEN
Professor Lewontin (Bulletin, which is to see the Academy openly March 1970) has likened my article. With the “Harvard Educational encourage scientific inquiry into the "How Much Can We Boon IQ and Review's" publication of Pro genetics of human abilities and pro Scholastic Achievement?" ("Harvard fessor Jensen's article last win clivities. including tbeir racial as Educational Review," Winter, 1969) ter, “jenseniim" came into the pects. Lewontin's approach makes to the "pernicious heresy . . . of language. Whether it is a term it appear to me that he views the total depravity, irresistible grace. lack of opprobrium or approbation problems of criticizing my article as of free will predestination and lim depends on which educational that of making a case for the ' good ited atonement" attributed to Bishop psychologist you talk to. The guys" versus the "bad guys," and he Jansen in the seventeenth century. ripples of "fensenism" have ex wants there to be no doubt in the Lewontin goes on to claim that the tended far beyond the usually reader's mind that he is very much same doctrine is now called "jensen- placid pond of the schoolmen's one of the "good guys." Thus he ism" (a term coined by the "Wall publications, having received finally makes it perfectly clear in Street Journal"), and that "jensen- attention from the “ Ynr York the last few sentences of his article ism" is "as erroneous in the twentieth Times Magatine" and the “Wall that he opposes me mainly for ideo century as it was in the seventeenth." Street Journal." The Jensen the logical reasons and not on scientific Lewontin proposes to play the role of sis holds that differences in or technical grounds. Pope Innocent X (who denounced learning ability are largely ge Bishop Jansen in Id5>) by holding netic rather than environmental. A PERSISTENT QUESTION up and condemning his own version, In March the Bulletin published incomplete and distorted, of “jensen- a criticism of Jensen's position Lewontin's statement that "Jensen ism." by Richard lewontin. professor has made it fairly clear to me what Thus Lewontin sets the stage for of biology at the University of sort of society he wants" is not based the ad homincm flavor of the rest Chicago. This month ice pre on knowledge that Lewontin has of of his paper. His role may resemble sent the rejoinder. Arthur Jen my social or political philosophy. It that of Pope Innocent's in trying to sen is professor of educational is a subjective surmise reflecting put down what he perceives as a psychology at the University of Lewontin's antipathy for anyone who heresy, but readers of Lcwontin's California. Berkeley. would raise the question of genetic piece may be reminded of a closer racial intelligence differences in an ecclesiastical parallel in Bishop Wil- obviously non-political, scholarly con berforce. who. in debating evolution text. The question of whether the with T. H. Huxley, resorted to com scientific atticles. and winner of nu observed racial differences in mental menting that Darwin's physiognomy merous scientific awards and distinc abilities and scholastic performance bore a simian resemblance; and he tions) as "the eminent white Anglo- involve genetic as well as environ begged to know of Huxley, "was it Saxon inventor." (True, Shockley has mental factors is indeed tabooed. through his grandfather or grand, 85 patented inventions, including the Nevertheless, it is a persistent ques mother that he claimed his descent junction transistor.) If Lewontin is tion. My belief is that scientists in from a monkey?" Thus we see trying to be uncomplimentary, it is the appropriate disciplines must face Lewontin, albeit in a milder s-ein, interesting to sec the labels he picks the question and not repeatedly sweep referring to Edward L. Thorndike for this. it back under the rug. In the long (probably America's greatest psy In connection with Lewontin's ref run. the safest and sanest thing we chologist and a pioneer in twin stud erence to Shockley, an error of fact can urge is intensive, no-holds-barrcd ies of the heritahility of intelligence) calls for correction. Shockley has inquiry in the best tradition of sci as a "Methodist Yankee" and to Wil not urged the Academy to study "the ence. liam Shockley (a Nobel Laureate in effects of interracial mating." This Before proceeding with comments physics, author of some three hundred is a distortion of Shockley's aim. on specific technical points in Lewon- May 1970 Bulletin of the Atomic Scientists 17 tin's paper. it would be well to put turally disadvantaged with some of harmful, just as it would be harmful them in ptopcc perspective by giving the cultural amenities enjoyed by for a physician to give all his pa a capsule summary of what my article middle class children for a period of tients the same medicine. was about. a year or two before they enter I know personally of many in Khool. we arc told, the large differ SURVEY FINDINGS stances in which children with edu ences in Kholastk aptitude would be cational problems were dented the Fir»t, I reviewed some of the evi minimized and the Khools could go school's special facilities for dealing dence and the conclusions of a nation on thereafter treating all children with such problems (small classes, wide survey and evaluation of the very much alike and expect nearly all specialist teachers, tutorial help, di large, federally-funded compensatory to perform as "average children" for agnostic Ktviccs. etc.), not because education programs made by the their grade in Khool. the children did not need this special U.S. Commission on Civil Rights, It hasn't worked. And educators attention or because the services were which concluded that these special arc now beginning to say: "Let's not available to the Khool, but sim programs had produced no signifi really look at individual differences ply because the children were black cant improvement in the measured and try to find a satiety of instruc and no one wanted to single them intelligence or scholastic perfor tional methods and differentiated out as being different or in need of mance of the disadvantaged children programs that will accommodate special attention. So instead, white whose educational achievements these these differences." Whatever their middle-class children with similar programs were specifically intended causes may be. it now seems certain educational problems get nearly all to improve. The massive evidence that they arc not so superficial as to the attention and special treatment, presented by the Civil Rights Com be erased by a few months of "cul and most of them benefit from it. mission suggests to roe that merely tural enrichment." "verbal stimula No one objects, became this is not applying more of the same approach tion," and the like. I have pointed viewed by anyone as "dberimina- to compensatory education on a still out that some small-Kale experi lion." But some Khool districts have larger Kale is not at all likely to lead mental intervention programs, which been dragged into court for trying to the desired results, namely, in gear specific instructional methods to provide similar facilities for mi creasing the benefits of public educa to developmental differences, have nority children with educational tion to the disadvantaged. The well- shown more promise of beneficial problems. In these actions the well- documented fruitlessncss of these results than the largc-Kalc programs intentioned plaintiffs undoubtedly well-intentioned compensatory pro based on a philosophy of general cul viewed themselves as the "good grams indicates the importance of tural enrichment and a multiplication guys." Many children. I fear, by be now questioning the assumptions, of the resources in already existing ing forced into the educational mold theories and practices on which they programs for the “average child.” of the "average child" from Grade 1 were based. on. are soon "turned off" on Khool I agree with Lewontin that these THE OPPORTUNITIES learning and have to pay the con sequences in frustration and defeat, assumptions, theories and practices— One of the chief obstacles to pro espoused over the past decade by the viding differentiated educational both in Khool and in the world of work for which their Khooling has majority of educators, social and be programs for children with different not prepared them. havioral scientists—are bankrupt. 1 patterns of abilities, aside from the do not blame the children who fail lack of any detailed technical knowl I do not advocate abandoning ef to benefit from these programs, as edge as to how to go about this most forts to improve the education of the Lewontin would have his readers effectively, is the fact that children disadvantaged. I urge increased em think. A large part of the failure. in different visibly identifiable sub phasis on these efforts, in the spirit I believe, has resulted from the fail populations probably will be dispro of experimentation, expanding the ure and reluctance of the vast ma portionately rcprcKntcd in different diversity of approaches and improv jority of the educational establish instructional programs. This highly ing the rigor of evaluation in order ment. aided and abetted by social probable consequence of taking in to boost our chances of diKOvcring Kicntists, to take seriously the prob dividual differences really seriously the methods that will work best. lems of individual differences in de is miKonstrued by some critics as velopmental rates, patterns of ability, inequality of opportunity. But ac LEARNING AND IQ and learning styles. The prevailing tually. one child's oportunity can be My article also dealt with my philosophy hat been that all children another's defeat. To me. equality of theory of two btoad categories of are basically very much alike—they opportunity does not mean uniform mental abilities, which I call intelli arc all "average children"—in their treatment of all children, but equal gence (or abstract reasoning ability) mental development and capabilities, ity of opportunity for a diversity of and associative learning ability. and that the only causes of the vast educational experiences and services. These types of ability appear to be differences that show up as they go If we fail to take account either of in distributed differently in various so through Khool arc due to cultural nate or acquired differences in abili cial classes and racial groups. While factors and home influences that ties and traits, the ideal of equality of large racial and social class differ mold the child even before he enters educational opportunity can be in ences arc found for intelligence, kindergarten. By providing the cul terpreted so literally as to be actually there are practically negligible dif- 18 ferences among these groups in as sociative learning abilities such as memory span and serial and paired- associate rote learning. Research should be directed at de lineating still other types of abilities and at discovering how the particular strengths of each individual's pat tern of abilities can be most effec tively brought to bear on school learning and on the attainment of oc cupational skills- By pursuing this path. I believe we can discover the means by which the reality of indi vidual differences need not mean educational rewards for some chil dren and utter frustration and defeat mental factors contributing a minor least three sociologists who are stu for others. portion of the variance among indi dents of this problem — Pitirim Soro viduals. The heritability of the IQ— kin, Bruce Eckland and Otis Dudley INTELLIGENCE that is. the percentage of individ Duncan—all agree that selective I pointed out that IQ tests evolved ual differences variance attributable factors in social mobility and associa to predict scholastic performance in to genetic factors—comes out to tive mating have resulted in a genetic largely European and North Ameri about 80 per cent, the average value component in social class intelligence can middle class populations around obtained from all relevant studies differences. As Eckland points out, the turn of the century. They evolved now reported. this conclusion holds within socially to measure those abilities most rele These estimates of heritability are defined racial groups but cannot vant to the curriculum and type of based on tests administered to properly be generalized among ra instruction, which in turn were European and North American pop cial groups, since barriers to upward shaped by the pattern of abilities of ulations and cannot properly be gen mobility have undoubtedly been the children the schools were then eralized to other populations. I be quite different for various racial intended to serve. lieve we need similar heritability groups. IQ or abstract reasoning ability is studies in minority populations if we thus a selection of just one portion of are to increase our understanding of RACE DIFFERENCES the total spectrum of human mental what our tests measure in these pop I have always advocated dealing abilities. This aspect of mental abil ulations and how these abilities can with persons as individuals, each in ities measured by IQ tests is im be most effectively used in the edu terms of his own merits and charac portant to our society, but is obvious cational process. teristics and I am opposed to accord ly not the only set of educationally ing treatment to persons solely on or occupationally relevant abilities. CLASS DIFFERENCES the basis of their race, color, national Other mental abilities have not yet Although the full range of IQ and origin or social class background. But been adequately measured; their dis other abilities is found among chil I am also opposed to ignoring or tributions in various segments of the dren in every socioeconomic stratum refusing to investigate the causes population have not been adequately in our population, it is well estab of the well-established differences determined; and their educational lished that IQ differs, on the average, among racial groups in the distri relevance has not been fully ex among children from different social bution of educationally relevant plored. class backgrounds. The evidence, traits, particularly IQ. I believe a much broader assess some of which I referred to in my I believe that the causes of ob ment of the spectrum of abilities and article, indicates to me that some of served differences in IQ and scholas potentials, and the investigation of this IQ difference is attributable to tic performance among different their utilization for educational environmental differences and some ethnic groups is. scientifically, still achievement, will be an essential of it is attributable to genetic differ an open question, an important ques aspect of improving the education of ences among social classes—largely tion and a rcsearchablc one. I be children regarded as disadvantaged. as a result of differential selection of lieve that official statements such as: the parent generations for different "It is a demonstrable fact that the INHERITANCE patterns of ability. talent pool in any one ethnic group Much of my paper was a review of I have not yet met or read a mod is substantially the same as in any the methods and evidence that led ern geneticist who disputes this in other ethnic group" (U S. Office of me to the conclusion that individual terpretation of the evidence. In the Education. 19 Chine*, of any other group" (U S. There has been only one acceptable natal. and cultural, and my article in Dept, of Labor. 1965) are without hypothesis — the environmentalists' cludes sections on all these factors. scientific merit. They la RICHARD C. LEWONTIN 2. Further Remarks on Race and the Genetics of Intelligence “. . . and I found a few faults." scientific paper, is attacked on non- completely inadequate theoretical Rickard C. Isvonlin, icko con scientific, ideological grounds. But and experimental basis. The paper tinues here to take issue Kith Jensen is wrong in two respects. in question appeared in 1905. 1) "jensenism" is professor of bi- There is no ad homincm argument year* before Fisher's paper establish ology at the University of Chi- in my article. I confess to one epi ing the statistical theory on which OflgO. sode of self-caricature when I com hcritabilitics arc estimated, 10 years pared my role to that of a Pope de before Fisher worked out the sampl Professor Jensen has, understand nouncing a heresy, and to a rather ing distribution of the correlation co ably. responded at some length to my vulgar attempt to have some fun efficient, and 5 jears before Mor analysis of his article "How Much with Dr. Shockley by describing him gan's chromosome theory of inherit Can We Boost IQ and Scholastic in reverse racist terms, but in no case ance. In an attempt to explain how Achievement?” In large part, his re is my argument about Dr. Jensen's "America's greatest psychologist" sponse only reinforce* many of the paper made on any grounds but its could have made such an obviously points I made about his original ar merit and logic. Indeed, my remarks unscientific statement. I postulated ticle, but he does raise some new and about E. I.. Thorndike are the best that it was a prejudice that might be very interesting issues. I shall try demonstration of that. In what Jen expected from the son of a New to deal with his reply as briefly as I sen informs us is "an empirical paper England Methodist clergyman I did can. based on twin correlations," Thorn not attempt by that hypothesis to dis Jensen's overall objection is that dike makes the remarkable statement credit Thorndike's statement. It dis my article makes liberal use of ad that "In the actual race of life, which credits itself. homincm argumentation in an at is not to get ahead, but to get ahead But more important. Jensen's ar tempt to establish myself as a "good of somebody, the chief determining ticle is not an objective empirical guy" attacking a "bad guy.” Thus. factor is heredity." That maxim is a scientific paper which stands or falls Professor Jensen establishes himself conjunction of a socio-economic prej on the correctness of his calculation as a dispassionate scientist who. hav udice about the nature of human re of heritability. It is, rather, a close ing written an objective empirical lations and a scientific statement with ly reasoned ideological document May 1970 Bulletin of the Atomic Scientists 23 springing, as ! have shown, from a vote, but I do not think Jensen tions arc genetically much alike but deep-sealed professional bias and would fas-or that technique any more live in environments that differ from permeated, like Thorndike's work, than I would, especially in view of each other in some critical limiting with an elitist and competitive world the fact that the membership at last factor. All of these occur in nature, view. While Jensen's original article year's meeting of the National Acad and again no a priori likelihoods can gave many instances of this world emy voted almost unanimously not be fairly assigned to them. view, of which 1 quoted a few, his to consider the question of race and For the race problem, however, we reply to my analysis provides yet the inheritance of IQ. (They have can say something because of Other fresh evidence. Thus, we read of the since reversed themselves.) information. The first possibility is attempt to equalize children’s school Jensen has parried my major scien quite unlikely because the result of performance as being analogous to tific thrust at his thesis by saying selection would be the elimination the attempt to "transmute base met that I have demanded an unrealistic of additive genetic variance, leaving als into gold." Jensen speaks of level of proof. I share with Jensen only dominance and inter-action vari "particular strengths in each individ an impatience with that smart-alec ance. But Burt's data, quoted by ual's pattern of abilities” as if he re who is always telling us our evidence Jensen, show that -18 per cent of the garded those differences in a value- is only circumstantial and we haven't variance in IQ is additive genetic free way. objectively. Yet a little really proved our point. But that variance. This is a high figure for a later he discusses dysgcnic trends" is not my case at all. I think there quantitative trait in general, and ab among blacks. How revealing is is an honest misunderstanding here, surdly high for any trait that has rhetoric. not simply a polemical question. If long been under natural selection. It two populations have very low herit appears that IQ has been selectively WEIGHT OF AUTHORITY ability for a trait but differ from neutral, at least over much of our each other on the average, there arc species history. The second and third As in his original article. Jensen three possibilities. Each population possibilities are more or less equally in his reply relics heavily on the may have been highly inbred, in likely explanations of the situation weight of authority as relevant evi which case the genetic component of in man and I would not care to bet dence. We hear of a "Nobel Lau the differences between them may be the educational future of any chil reate in physics." or "three sociolo very high. Each population may have dren on one or the other. gists who arc students of this prob been subject to a different force of lem " and who “all agree.” "gene natural selection, again causing them FUTURE IQi ticists such as K. Mather, C. D. Dar- each to he nearly homozygous, so Jensen remarks that I have said lington. R. A. Fisher and Francis that again the difference between Crick, to nime a few" and finally, “a nothing about "dysgcnic" trends populations might be chiefly gene among blacks which he regards as number of highly qualified geneti tic. Finally, both populations might cists" who have reviewed his "treat "socially more important than the be highly variable genetically, in question of racial differences per sc." ment of quantitative genetics and which case the populations almost Apparently Jensen believes that low have found no fault with it." Well, certainly owe their observed differ I am a very highly qualified geneti er IQ blacks are out-breeding higher ence almost entirely to environment. IQ blacks so that the average differ cist whose field is the study of gene One cannot assign a priori probabili tic variation in natural populations, ence between blacks and whites will ties (or likelihoods, better) to these become even greater than it is. The and I found a few faults. For exam three situations. In any common ple. the estimate of heritability by- evidence for this is indirect and is sense meaning of the word, they have of the form: lower socio-economic ratios of correlation differences are equal likelihoods, since all three cir classes have more children than high upwardly biased by environmental cumstances occur quite frequently in er ones, lower socio-economic classes correlations which may be consider the history of species. What about have lower IQ Mores. IQ score is able. (One of those Other "highly- the reverse situation, the one applic highly heritable, therefore IQ will qualified geneticists” also points this able to our problem? If two popula Out in his comments on Jensen's arti decrease. Such eminent geneticists as tions have high heritabilities for a C I). Darlington and R. A. Fisher cle in the "Harvard Educational Re character, and there is an average used to make this argument about view.” so Jensen's phalanx of au difference between them, is that dif social classes among whites too. but thorities is not quite unbroken.) ference mostly gcnetical? One pos Moreover, heritabilities, being ratios, they were proved wrong by a then- sibility is that the populations differ unknown human geneticist who should not be averaged in the usual genetically because of a previous his teaches at Grand Valley State Col way. No standard error is given for tory of differential selection of a lege. Carl Bajcma (whose work is Jensen's estimate of heritability. No type that causes genetic variation to examination of the sensitivity of quoted by Jensen) showed that the be stabilized. Another possibility is heritability estimates to different old story that lower IQ classes out- that the populations may differ gene genetic models is given. This is im breed higher IQ classes was the er portant if there is a lot of dominance tically because of historical accidents roneous result of an egregious sta variance. And so on. But if author of genetic sampling (genetic drift) tistical blunder: They forgot to ity is evidence, what do we do when without differential selection. A count women who had no children! authorities disagree? We might take third possibility is that the popula In fact, women with low IQs have 24 much bigger families when they have deeper flaw in Jensen's scheme. Put and what rewards he will receive, is a family, but many fewer of them ting questions of race quite aside, we social. At present our society is tru have families. The result is that the must expose the fallacy that, because ly one in which "the race is to get reproductive rate of the highest IQ human behavior is chiefly genetical ahead of somebody" and nothing classes is actually the highest. This ly determined at present, it must al- suits that dog-cat-dog philosophy information docs not exist for blacks w ap be so and ought alwap to be so. better than the notion that winners, and all the information quoted by Children arc different. They are dif like heroes, are born, not made. But Jensen about blacks is of the pre- ferent at birth and different when that is a social attitude, not an ine Bajrma biased variety. they reach school. But what Jensen luctable biological result. In answer I would like to end my contribu continues to misunderstand is that to Prof. Jensen's rhetorical question tion to this controversy by returning whether those differences are gene- "How much Can We Boost IQ and to my original point. Jensen has tical, maternal, obstetrical or Oedi- Scholastic Achievement?" I say "As spent a great deal of energy on the pal. the decision about what role each much or as little as our social values question of whether there is a genetic child is to play eventually in society may eventually demand." difference between blacks and whites in IQ. He believes this to be an im portant social question and not sim ply a matter of vulgar curiosity. But suppose the difference between the black and white IQ distributions were completely genetic: What pro EUGENE RA BINOWITCH gram for social action flows from that fact? Should all black children be given a different education from all white children, even the II per 3. Jensen vs. Lewontin cent who are better than the average white child? Should all black men (A Comment) be unskilled laborers and all black women clean other women's houses? Jensen sap he believes in the pri There is no doubt that intelligence showing that there is no general as macy of the individual, yet he is —particularly as defined by some sociation between the two traits. deeply concerned with the genetic number derived from standard tests To explore the IQ of dark-skinned causation of group differences. Why? —is affected by hereditary factors, children in comparison with that of Because, he says, "Since much of the similarly to other mental traits, such light-skinned ones has, it seems to current thinking behind civil rights, as musicahty, artistic talent, verbalixa- me, as much scientific significance fair employment, and equality of tion capacity and all other character as exploring the correlation of any educational opportunity appeals to istics of an individual human being. two probably independent characters: the fact that there is a dispropor It is, however. not very likely that for example, comparing the IQ of tionate representation of different the inheritance of one such trait is red-haired children with that of racial groups in various levels of the correlated with that of some other black-haired ones, or of children educational, occupational, and socio- unrelated, trait—be it body height, with hereditary inclination to obesity economic hierarchy, we are forced to skin color, color of the eyes or of with that of congenitally skinny ones. examine all possible reasons for this the hair, unless we deal with a closed, It is not impossible that statistically inequality. . . isolated population. In a small ge significant associations will be found netically inbred group, one may find in any such case—the more likely X O T T R U E the preponderance of, say. red hair the smaller and genetically more ho coupled with musicality. or gray eyes mogeneous the groups used for com Nonsense. Does Jensen really be coupled with low intelligence. parison. But what of it? The ex lieve that all the fuss about civil In large populations which have istence of such correlations would rights has occurred because someone gone through much inter-breeding, not be considered as an argument noticed that blacks were under-rep as all major human races have (par against common education for all of resented in college classes? It is sim ticularly in America, both before and them. Would anybody suggest a ply not true that "we are forced to after migration to this continent), separate education for redheads, if examine all possible reasons for this there is little probability of predom it be proved that they have a lesser inequality." What we arc morally inant association of two entirely in (or greater) scholastic aptitude than obliged to do is to eliminate black dependent characteristics. Even such black-haired children? Or that of ness per sc as a cause of unequal physical characteristics as (hick lips, children of Mediterranean extraction treatment and for that program we often found associated with dark if it be found that they have a greater have no need of genetics. skin in America, is not typical of (or lesser) verbal ability than chil But that program cannot be accom other dark-skinned populations in dren of Anglo-Saxon extraction? plished unless we challenge a yet parts of Africa or in the Caribbean. We do not consider the possible M a y 1970 H utlrtin o/ the Atomic Scientist* 2 5 existence of qualitative or quintiitive respects, cither below or above the unrelated physical, physiological, be sex-rclzted intellectual charzctcristics average. Perhaps, the greatest prob havioral and intellectual character a sufficient reason for not sending lem of universal education in a demo istics. boys and girls through common cratic society is how to treat each By all means, let us explore sys educational channels. We make no child as an individual, providing him tematically the hereditary and en attempt to provide separate educa with remedial attention if he needs vironmental factors affecting—or at tional channels for children with it. and with special stimulation and least found to be associated with— different IQs. as England docs, sepa adequate challenges to his abilities intelligence as well as other mental rating at age 11 children showing if they are substantially above the abilities. But let the results of such high scholarly aptitudes from those average. studies not interfere with funda with lower ones. Such separation But before any group recognizable mental principles of democratic edu would be contrary to the prevailing by some visible trait—be it skin cation: that every child is entitled American attitude to national educa color, hair color, body height, or any to the same education and that the tion. This attitude is based on belief other racial or familial characteristic profit to him of common schooling that whatever the IQ or the early —is summarily tossed into a separate outweighs the disadvantages which school achievement of a child, he or educational bag, whether superior or a child with certain less-than-average she will profit more from being edu inferior to that of the others, demo intellectual traits (as well as a child cated together with children of high cratic society must fulfil its promise with higher-than-average ones) can er (and lower) IQs or school grades, to give everybody the same chance, be expected to derive from such com than by being separated and attend by equalizing the essential environ mon schooling, even if it is by ne ing an educational institution attuned mental conditions under which chil cessity adapted to "average" intel to a particular type of intellectual dren of different groups grow up. lectual capacities endowment. We believe that com Our society is just beginning to ful Let us provide all possible remedial mon education can best prepare chil fill this long unrecognized, and still attention to the former, and all pos dren for life in a society where equxl neglected, obligation to its black sible stimulation to the latter! And rights and obligations, and equxl minority. above all, let us not slacken in the voice in public affairs, arc available If voices are now raised suggesting drive to equalize the economic and —or at least are supposed to be avail that this belated attempt be declared educational opportunities open to able—to all. Only extreme cases of futile, because early experiments such all—black or white, tall or squat, mental deficiency or abnormality as Project Head Start have not been gray-eyed or black-eyed. Catholic, justify, in our society, exclusion from a universal success, a grave suspicion Protestant or Jewish—whether or not participation in a common education arises that behind such suggestions we have reasons to suspect that some al system. there lies a racist prejudice—meaning of these groups have certain intel This system of common education by this term not racial hatred or lectual, aesthetic or other mental en for all does not exclude special at racial contempt, but unthinking be dowments differing, qualitatively or tention for those children whose in lief that one genetic trait—such as quantitatively, from those of some tellectual capabilities are. in certain skin color—may predetermine other others. Next Month: A Bulletin Special Issue Alamogordo + 25 Years A review by many of its principal figures of the Nuclear Energy Revolution in peace and war in the quarter century since the first atomic bomb rose like the sun out of New Mexico, July 16, 1945. This double-sized issue is another unique contribution of The Bulletin of the Atomic Scientists to the contemporary scene and the historical record of the twentieth century. Articles and commentaries by: Glenn T. Seaborg, Alice Kimball Smith, Robert Marshak. Alvin Weinberg, Eugene Rabinowitch. Sir Rudolph Pcierl*. Hans Bcthc, Edward Creutz, Gerald Johnson. Robert R. Wilson. Ryukichi Imai. Ralph Lapp, Gen. Leslie R. Grove* and others! 26