Hen’s Teeth and Horse’s Toes
If a group generally regarded as intellectually able could be ranked as inferior, then the basic argument for restriction would be greatly enhanced, for who would then defend the groups that everyone considered as stupid? Pearson, however, loudly decried any attempt to attribute motive or prior prejudice to his study. One can only recall Shakespeare’s line, “The lady doth protest too much, methinks.”
There is only one solution to a problem of this kind, and it lies in the cold light of statistical inquiry…. We have no axes to grind, we have no governing body to propitiate by well advertised discoveries; we are paid by nobody to reach results of a given bias. We have no electors, no subscribers to encounter in the market place. We firmly believe that we have no political, no religious and no social prejudices…. We rejoice in numbers and figures for their own sake and, subject to human fallibility, collect our data—as all scientists must do—to find out the truth that is in them.
Pearson had invented a statistic so commonly used today that many people probably think it has been available since the dawn of mathematics—the correlation coefficient. This statistic measures the degree of relationship between two features of a set of objects: height versus weight or head circumference versus leg length in a group of humans, for example. Correlation coefficients can range as high as 1.0 (if taller people are invariably heavier to the same degree) or as low as 0.0 for no correlation (if an increase in height provides no information about weight—a taller person may weigh more, the same, or less, and no prediction can be made from the increase in height alone). Correlation coefficients can also be negative if increase in one variable leads to decrease in the other (if taller people generally weigh less, for example). Pearson’s study of Jewish immigration involved the measurement of correlations between a large and motley array of physical and mental characters for children of Jewish immigrants living in London.
Pearson measured everything he imagined might be important in assessing “worthiness.” He established four categories for cleanliness of hair: very clean and tidy, clean on the whole, dirty and untidy, and matted or verminous. He assessed both inner and outer clothing on a similar scale: clean, a little dirty, dirty, and filthy. He then computed correlation coefficients between all measures and was generally disappointed by the low values obtained. He could not understand, for example, why cleanliness of body and hair correlated only .2615 in boys and .2119 in girls, and mused:
We should naturally have supposed that cleanliness of body and tidiness of hair would be products of maternal environment and so highly correlated. It is singular that they are not. There may be mothers who consider chiefly externals, and so press for tidiness of hair, but it is hard to imagine that those who emphasize cleanliness of body overlook cleanliness of hair.
Pearson concluded his study of physical measures by proclaiming Jewish children inferior to the native stock in height, weight, susceptibility to disease, nutrition, visual acuity, and cleanliness:
Jewish alien children are not superior to the native Gentile. Indeed, taken all round we should not be exaggerating if we asserted that they were inferior in the great bulk of the categories dealt with.
The only possible justification for admitting them lay in a potentially superior intelligence to overbalance their physical shortcomings.
Pearson therefore studied intelligence by the same type of short and subjective scale that had characterized his measures of physical traits. For intelligence, he relied upon teachers’ judgments rated from A to G. Computing the raw averages, he found that Jewish children were not superior to native Gentiles. Jewish boys ranked a bit higher, but the girls scored notably lower than their English classmates. Pearson concluded, with a striking analogy:
Taken on the average, and regarding both sexes, this alien Jewish population is somewhat inferior physically and mentally to the native population…. We know and admit that some of the children of these alien Jews from the academic standpoint have done brilliantly; whether they have the staying power of the native race is another question. No breeder of cattle, however, would purchase an entire herd because he anticipated finding one or two fine specimens included in it; still less would he do it, if his byres and pastures were already full.
But Pearson realized that he was missing one crucial argument. He had already admitted that Jews lived in relative poverty. Suppose intelligence is more a product of environment than inborn worth? Might not the average scores of Jews reflect their disadvantaged lives? Would they not be superior after all if they lived as well as the native English? Pearson recognized that he had to demonstrate the innateness of intelligence to carry his argument for restricted immigration based on irremediable mediocrity.
He turned again to his correlation coefficients. If low intelligence correlated with measures of misery (disease, squalor, and low income, for example), then an environmental basis might be claimed. But if few or no correlations could be found, then intelligence is not affected by environment and must be innate. Pearson computed his correlation coefficients and, as with the physical measures, found very few high values. But this time he was pleased. The correlations produced little beyond the discovery that intelligent children sleep less and tend to breathe more through their nose! He concluded triumphantly:
There does not exist in the present material any correlation of the slightest consequence between the intelligence of the child and its physique, its health, its parents’ care or the economic and sanitary conditions of its home…. Intelligence as distinct from mere knowledge stands out as a congenital character. Let us admit finally that the mind of man is for the most part a congenital product, and the factors which determine it are racial and familial…. Our material provides no evidence that a lessening of the aliens’ poverty, an improvement in their food, or an advance in their cleanliness will substantially alter their average grade of intelligence…. It is proper to judge the immigrant by what he is as he arrives, and reject or accept him then.
But conclusions based upon negative evidence are always suspect. Pearson’s failure to record correlations between “intelligence” and environment might suggest the true absence of any relationship. But it might also simply mean that his measures were as lousy as the hair in his category 4. Maybe a teacher’s assessment doesn’t record anything accurately, and its failure to correlate with measures of environment only demonstrates its inadequacy as an index of intelligence. After all, Pearson had already admitted that correlations between physical measures had been disappointingly small. He was too good a statistician to ignore this possibility. So he faced it and dismissed it with one of the worst arguments I have ever read.
Pearson gave three reasons for sticking to his claim that intelligence is innate. The first two are irrelevant: teachers’ assessments correlate with Binet test scores, and high correlations between siblings and between parents and children also prove the innateness of intelligence. But Pearson had not given Binet tests to the Jewish children and had not measured their parents’ intelligence in any way. These two claims referred to other studies and could not be transferred to the present case. Pearson appreciated this weakness and therefore advanced a third argument based upon internal evidence: intelligence (teachers’ assessment) failed to correlate with environment but it did correlate with other “independent” measures of mental worth.
But what were these other independent measures? Believe it or not, Pearson chose “conscientiousness” (also based on teachers’ assessments and scored as keen, medium, and dull), and rank in class. How else does a teacher assess “intelligence” if not (in large part) by conscientiousness and rank in class? Pearson’s three measures—intelligence, conscientiousness, and rank in class—were redundant assessments of the same thing: the teachers’ opinion of their students’ worth. But we cannot tell whether these opinions record inborn capacities, environmental advantages, or teachers’ prejudices. In any case, Pearson concluded with an appeal to bar all but the most intelligent of foreign Jews:
For men wi
th no special ability—above all for such men as religion, social habits, or language keep as a caste apart, there should be no place. They will not be absorbed by, and at the same time strengthen the existing population; they will develop into a parasitic race.
Goddard’s and Pearson’s studies shared the property of internal contradictions and evident prejudice sufficient to dismiss all claims. But they differed in one important respect: social impact. Britain did not enact laws to restrict immigration by racial or national origin. But in America, Goddard and his colleagues won. Goddard’s work on Ellis Island had already encouraged immigration officials to reject people for supposed moronity. Five years later, the army tested 1.75 million World War I recruits with a set of examinations that Goddard helped write and that were composed by a committee meeting at his Vineland Training School. The tabulations did not identify Jews per se but calculated “innate intelligence” by national averages. These absurd tests, which measured linguistic and cultural familiarity with American ways, (see my book, The Mismeasure of Man, W. W. Norton, 1981), ranked recent immigrants from southern and eastern Europe well below the English, Germans, and Scandinavians who had arrived long before. The average soldier of most southern and eastern European nations scored as a moron on the army tests. Since most Jewish immigrants arrived from eastern European nations, quotas based on country of origin eliminated Jews as surely as collegiate quotas based on geographical distribution once barred them from elite campuses.
When quotas were set for the Immigration Restriction Act of 1924, they were initially calculated at 2 percent of people from each nation present in America at the census of 1890, not at the most recent count of 1920. Since few southern and eastern Europeans had arrived by 1890, these quotas effectively reduced the influx of Slavs, Italians, and Jews to a trickle. Restriction was in the air and would have occurred anyway. But the peculiar character and intent of the 1924 quotas were largely a result of propaganda issued by Goddard and his eugenical colleagues.
What effect did the quotas have in retrospect? Allan Chase, author of The Legacy of Malthus, the finest book on the history of scientific racism in America, has estimated that the quotas barred up to six million southern, central, and eastern Europeans between 1924 and the outbreak of World War II (assuming that immigration had continued at its pre-1924 rate). We know what happened to many who wanted to leave but had no place to go. The pathways to destruction are often indirect, but ideas can be agents as surely as guns and bombs.
23 | The Politics of Census
IN THE CONSTITUTION of the United States, the same passage that prescribes a census every ten years also includes the infamous statement that slaves shall be counted as three-fifths of a person. Ironically, and however different the setting and motives, black people are still undercounted in the American census because poor people in inner cities are systematically missed.
The census has always been controversial because it was established as a political device, not as an expensive frill to satisfy curiosity and feed academic mills. The constitutional passage that ordained the census begins: “Representatives and direct taxes shall be apportioned among the several states which may be included within this union, according to their respective numbers.”
Political use of the census has often extended beyond the allocation of taxation and representation. The sixth census of 1840 engendered a heated controversy based upon the correct contention that certain black people had, for once, been falsely over counted. This curious tale illustrates the principle that copious numbers do not guarantee objectivity and that even the most careful and rigorous surveys are only as good as their methods and assumptions. (William Stanton tells the story in The Leopard’s Spots, his excellent book on the history of scientific attitudes toward race in America during the first half of the nineteenth century. I have also read the original papers of the major protagonist, Edward Jarvis.)
The 1840 census was the first to include counts of the mentally ill and deficient, enumerated by race and by state. Dr. Edward Jarvis, then a young physician but later to become a national authority on medical statistics, rejoiced that the frustrations of inadequate data would soon be overcome. He wrote in 1844:
The statistics of insanity are becoming more and more an object of interest to philanthropists, to political economists, and to men of science. But all investigations, conducted by individuals or by associations, have been partial, incomplete, and far from satisfactory…. They could not tell the numbers of any class or people, among whom they found a definite number of the insane. And therefore, as a ground of comparison of the prevalence of insanity in one country with that of another, or in one class or race of people with that in another, their reports did not answer their intended purpose.
Jarvis then praised the marshals of the 1840 census as apostles of the new, quantitative order:
As these functionaries were ordered to inquire from house to house, and leave no dwelling—neither mansion nor cabin—neither tent nor ship unvisited and unexamined, it was reasonably supposed that there would be a complete and accurate account of the prevalence of insanity among 17 millions of people. A wider field than this had never been surveyed for this purpose in any part of the earth, since the world began…. Never had the philanthropist a better promise of truth hitherto undiscovered…. Many proceeded at once to analyze the tables, in order to show the proportion of lunacy in the various states, and among the two races, which constitute our population.
As scholars and ideologues of varying stripes scrutinized the tables, one apparent fact rose to obvious prominence in those troubled times. Among blacks, insanity struck free people in northern states far more often than it afflicted slaves in the South. In fact, one in 162 blacks was insane in free states, but only one in 1,558 in slave states. But freedom and the North posed no mental terror for whites, since their relative sanity did not differ in the North and South.
Moreover, insanity among blacks seemed to decrease in even gradation from the harsh North to the congenial South. One in 14 of Maine’s black population was either insane or idiotic; in New Hampshire, one in 28; in Massachusetts, one in 43; in New Jersey, one in 279. In Delaware, however, the frequency of insanity among blacks suddenly nose-dived. As Stanton writes: “It appeared that Mason and Dixon had surveyed a line not only between Maryland and Pennsylvania but also—surely all unwitting—between Sanity and Bedlam.”
In his first publication on the 1840 census, Jarvis drew the same conclusion that so many other whites would advance: slavery, if not the natural state of black people, must have a remarkably beneficent effect upon them. It must exert “a wonderful influence upon the development of moral faculties and the intellectual powers.” A slave gains equanimity by “refusing many of the hopes and responsibilities which the free, self-thinking and self-acting enjoy and sustain,” for bondage “saves him from some of the liabilities and dangers of active self-direction.”
The basic “fact” of ten times more insanity in freedom than in slavery was widely bruited about in the contemporary press, often in lurid fashion. Stanton quotes a contributor to the Southern Literary Messenger (1843) who, concluding that blacks grow “more vicious in a state of freedom,” painted a frightful picture of Virginia should it ever become a free state, with “all sympathy on the part of the master to the slave ended.” He inquired:
Where should we find penitentiaries for the thousands of felons? Where, lunatic asylums for the tens of thousands of maniacs? Would it be possible to live in a country where maniacs and felons meet the traveler at every crossroad.
But Jarvis was troubled. The disparity between North and South made sense to him, but its extent was puzzling. Could slavery possibly make such an enormous difference? If the information had not been stamped with a governmental imprimatur, who could have believed it? Jarvis wrote:
This was so improbable, so contrary to common experience, there was in it such a strong prima facie evidence of error, that nothing but a document, coming forth with all t
he authority of the national government, and “corrected in the department of state,” could have gained for it the least credence among the inhabitants of the free states, where insanity was stated to abound so plentifully.
Jarvis therefore began to examine the tables and was shocked by what he discovered. Somehow, and in a fashion that could scarcely represent a set of random accidents, the number of insane blacks had been absurdly inflated in reported figures for northern states. Jarvis discovered that twenty-five towns in the twelve free states contained not a single black person of sound mind. The figure for “all blacks” had obviously been recopied or misplaced in the column for “insane blacks.” But data for 135 additional towns (including thirty-nine in Ohio and twenty in New York) could not be explained so easily, for these towns actually reported a population of insane blacks greater than the total number of blacks, both sane and unhinged!
In a few cases, Jarvis was able to track down the source of error. Worcester, Massachusetts, for example, reported 133 insane in a total black population of 151. Jarvis inquired and discovered that these 133 people were white patients living in the state mental hospital located there. With this single correction, the first among many, black insanity in Massachusetts dropped from one in 43 to one in 129. Jarvis, demoralized and angry, began a decade of unsuccessful campaigning to win an official retraction or correction of the 1840 census. He began:
Such a document as we have described, heavy with its errors and its misstatements, instead of being a messenger of truth to the world to enlighten its knowledge and guide its opinions, it is, in respect to human ailment, a bearer of falsehood to confuse and mislead.