2 The Consequences of Eugenic Sterilization in Alberta

Douglas Wahlsten

It is widely, although perhaps not universally, held today that eugenic sterilization in Alberta was morally, legally, and politically wrong. The judge in the case of Muir v. Alberta strongly condemned what had been done in the name of eugenics, and Premier Ralph Klein (1942–2013) delivered an apology to Leilani Muir in person.¹ The Muir trial explored the illegality of many things done by the Alberta Eugenics Board (AEB), but it did not delve into the scientific basis for eugenics itself. No experts in population genetics testified at the trial.² This chapter examines the science available to the AEB in the 1950s and 1960s, when hundreds of Alberta children, among others, were sterilized, and estimates the likely effects of all those sterilizations.³ This estimate was not made at that time, so let us do it now. For historical and contemporary reasons, it is also informative to compare the likely genetic consequences to enhancements in children’s intelligence that could have been achieved by improving their environments, according to knowledge available at that time.⁴

The author specializes in the study of genetic influences on behaviour and has published several scientific articles on heredity and human intelligence.⁵ He attended much of the trial, became familiar with the Muir case in detail, and later served as an editor of her book, A Whisper Past.⁶ After her stunning victory in court, many others who had been sterilized by order of the AEB initiated their own lawsuits. The author was then asked to serve as an expert witness in the resulting trials, in the capacity of which he reviewed many complete case files and prepared a scientific assessment of the actions of the AEB.⁷ These cases were later settled out of court, and the identities of most of the plaintiffs are now protected.

Figure 2.1 Timeline of genetic advances. The Sexual Sterilization Act of Alberta was passed in 1928 and repealed in 1972. The scientific principles of genetic inheritance were understood before the act was passed but were not taken into account. Figure compiled and designed by Douglas Wahlsten, 2013.

EUGENICS PREDATED GENETICS

History reveals that the practice of eugenic sterilization, before Alberta’s Sexual Sterilization Act was passed in 1928, was conceived before the scientific principles of genetics and heredity were understood. It is quite a telling fact to see that the Mendelian laws of genetics and early scientific insights from the field of synthetic evolution had not been integrated with eugenics legislation until the act was repealed in 1972 (see Figure 2.1).⁸ By the 1920s, when eugenic policies were adopted in Alberta and several US states, scientific knowledge of heredity had advanced greatly,⁹ but there are serious doubts as to whether those responsible for implementing eugenic policies in Alberta had even a basic understanding of the scientific basis for what they wanted to accomplish.

British psychologist Francis Galton (1822–1911),¹⁰ cousin of evolutionary biologist Charles Darwin (1809–82), published Heredity Genius in 1869 on the basis of his observations of the achievements of eminent British men and their male offspring.¹¹ High achievement tended to follow the male line in those days, and Galton interpreted his observations as support for a “like begets like” view of heredity. He concluded, “I find that talent is transmitted by inheritance in a very remarkable degree.”¹² In 1883 he fathered the doctrine of eugenics, which he defined as “the study of the agencies under social control, that improve or impair the racial qualities of future generations either physically or mentally.”¹³ Galton advocated that government policies should be applied to foster breeding by the best specimens and restrain the breeding of supposedly inferior individuals. Although the Bohemian Augustinian monk and naturalist Gregor Mendel (1822–84) presented his brilliant experiments with the garden pea to the Bruenn Natural History Society in 1865, and the work was published in 1866, Galton took no account of Mendel’s work at all—nor did biological science appreciate its significance until 1900, when it was rediscovered and Mendel’s results were confirmed in a wide variety of species.¹⁴ From that time, the science of genetics developed rapidly.¹⁵ The mathematical aspects of genetics and risks of genetic disease in different kinds of relatives were well understood in the period when eugenic policies were active in Alberta, and mathematical methods for applying them to animal breeding in farming had become widely accepted.

LEVEL OF KNOWLEDGE

The AEB’s decisions involved primarily the disciplines of genetics, the psychology of child development, and intelligence testing. The board included people with higher education but not specific expertise in those fields. It would be reasonable to expect the AEB to have had at least a level of knowledge that students in undergraduate (bachelor’s degree) and graduate (master’s degree) programs in those fields at the University of Alberta would have acquired during that period. The University of Alberta calendar as well as the syllabi or course outlines deposited in the university archives for many pertinent courses were therefore consulted.¹⁶ The principles of genetics were well understood in the 1920s,¹⁷ and undergraduate courses at the university presented Mendelian genetics to students before the Sexual Sterilization Act was passed in 1928. In 1917, the course Animal Husbandry 52(a): Animal Breeding taught “the principles of animal breeding, variation, selection, heredity and modern methods of practice.” A new course, Biology 51: Comparative Anatomy and Genetics, was offered in 1918, and, in 1920, Biology 1: General Elementary Biology included “Mendelism, Genetics.” In 1922, Animal Husbandry 53: Animal Breeding was described as an “application of the principles of genetics to the improvement of domesticated livestock” and its text was Genetics in Relation to Agriculture.¹⁸ The first University of Alberta course devoted entirely to genetics was Field Husbandry 61: Genetics in 1923.

Selective breeding to improve the genetic quality of Alberta’s population was the principal rationale for eugenic sterilization at this time.¹⁹ The University of Alberta calendar for 1950–51 lists the course Animal Husbandry 65: Animal Breeding, which applied the principles of genetics to the improvement of farm animals, and the more general course Genetics 51, which introduced the basic principles of heredity. The texts for those courses, as listed in the calendar, were Animal Breeding Plans by the pioneering Kansas animal geneticist Jay Lawrence Lush (1896–1982) and Genetics (1945) by Texas-based Drosophila geneticist Edgar Altenburg (1888–1976), respectively.²⁰ These books indicate what ought to have been known by anyone using a scientific approach to selective breeding. After all, if contemporary genetic science was used to choose parents for propagating farm animals, surely the level of knowledge applied to the selective breeding of humans should have been at least as high as for work with domestic fowl and cattle.

The AEB relied heavily on data from intelligence testing.²¹ During the academic year of 1950–51, and for many years thereafter, the undergraduate course Psychology 63: Elementary Psychometrics taught students about “the nature of psychological tests, with practical work in administration, scoring and interpretation of tests,” and the graduate course Psychology 107: Clinical Psychometrics taught of “diagnostic clinical tests: Wechsler-Bellevue, Binet . . . with practical work in administration, scoring and interpretation.” Education 172–176 was a course to be taken by all students in education and included instruction about “the intelligence testing program, and the wise employment of test results.” The text for the course was Educational Psychology by the Washington-based child psychologists Lester D. Crow (1897–1983) and Alice Crow (life dates unknown), and supplementary references used as required reading included Differential Psychology by New York-based test psychologist Anne Anastasi (1908–2001) and Educational Psychology by the Columbia psychologist Arthur I. Gates (1898–1989) et al.²² The 1950–51 reading list for Education 476: Educational Psychology included The Meaning of Intelligence by University of Illinois education professor George D. Stoddard (1897–1981), Measuring Intelligence by the American psychologists Lewis M. Terman (1878–1956) and Maud A. Merrill (1888–1978), and Measurement of Adult Intelligence by the Romanian American psychologist David Wechsler (1896–1981).²³ Students in the course Education 476 were later assigned Intelligence and Experience by educational psychologist Joseph McVicker-Hunt (1906–91) as a reading, as indicated by annotations in copies of the book in the University of Alberta Library.²⁴ The library also held multiple copies of Anastasi’s Psychological Testing, for use as a required reading in the Edmonton university’s psychology courses.²⁵

ALBERTA LAW

Admission of a mentally deficient person to a provincial mental institution was governed by the Mental Defectives Act, whereas operations of the AEB and requirements for sterilization were governed by the Sexual Sterilization Act.²⁶ The Mental Defectives Act defined a “mentally defective person” as a “person in whom there is a condition of arrested or incomplete development of mind existing before the age of eighteen years, whether arising from inherent causes or induced by disease or injury.”²⁷

Section 14(4) of the Mental Defectives Act anticipated eugenic sterilization where it stated, in part, “The Superintendent (a) may discharge a mentally defective person from an institution in any case in which the Superintendent considers . . . that the power of procreation of such person no longer exists.”

The Sexual Sterilization Act used the same definition of a “mentally defective person” and stipulated,

The medical superintendent or other person in charge of a mental hospital may cause a patient of a mental hospital whom it is proposed to discharge therefrom, to be examined by or in the presence of the [Alberta Eugenics] Board.²⁸

Section 6(1) further stated, with respect to the examination process of so-called mentally defective persons,

If . . . the Board is unanimously of the opinion that the exercise of the power of procreation by that person (a) would result in the transmission of any mental disability or deficiency to his progeny, or (b) involves the risk of mental injury either to such person or his progeny, the Board may in writing direct such surgical operation for the sexual sterilization of that person as may be specified in the written direction.²⁹

Thus, the law clearly required that a person be judged mentally defective both at admission to an institution and when considered by the AEB. Surgical sterilization could be directed if the board believed the person’s mental disability or deficiency would otherwise be transmitted to progeny.

INTELLIGENCE

There have long been two general approaches to establishing mental deficiency: the psychometric and the social.³⁰ The psychometric method assesses mental ability using a standardized psychological test, generally known as an intelligence quotient (IQ) test; the resulting IQ score does not indicate why a child may score exceptionally low. The social method assesses the child’s progress in school and whether he or she is able to carry out the basic requirements of daily living, but it also explores the possible environmental sources of mental deficiency. The AEB considered only the psychometric evidence, primarily the IQ score.³¹

There was no general agreement among psychologists in the 1950s on the precise nature of intelligence, and much the same situation prevails today.³² The French psychologist Alfred Binet (1857–1911) and physician Théodore Simon (1873–1961) devised the first formal tests of intelligence, in 1908.³³ Binet’s original objective was to detect children doing poorly in school, who he believed would be aided by a program of special education.³⁴ Having explained how to measure intelligence, he stated, “After the evil comes the remedy. After identifying all types of intellectual defects, let us pass on to their treatment.” He presented a new course of “mental orthopaedics exercises” that “make the child an active individual instead of reducing his role to that of a listener.”³⁵

When Binet’s test was imported to the United States, however, it was initially adapted and applied by psychologists who placed a much greater emphasis on biological determinants.³⁶ The test was translated and adapted by Lewis M. Terman at Stanford University.³⁷ The Stanford-Binet test was intended by Terman and Merrill to provide an “all-round clinical appraisal of a subject’s intellectual ability” using a “method of standardized interview which is highly interesting to the subject.”³⁸ The test obtained a general estimate of intelligence by tapping a wide range of capacities and then using the total score on all items. According to the author of a competing intelligence test, the Wechsler Intelligence Scale for Children (WISC), “Intelligence is the aggregate or global capacity of the individual to act purposively, to think rationally and to deal effectively with his environment [emphasis in the original].”³⁹ Wechsler’s conceptualization went beyond intellectual or reasoning ability per se. He argued that non-intellective and personality factors were essential parts of intelligence. For him, intelligence depended strongly on the individual’s drive or incentive to perform well on several tasks, the “persistence in attacking them” and “zest and desire to succeed.”⁴⁰

While differing to some extent on what intelligence is, leading experts from the mid- to the late twentieth century generally agreed on what it is not. Specifically, an intelligence test is not a measure of some innate or biologically fixed entity. According to George D. Stoddard, “A definition of intelligence may be expressed independently of the way in which abilities are developed in the organism. . . . But to define intelligence as a composite of inherited factors, or as a derivative of environmental pressures, would beg the question.” He criticized earlier theorists such as Cyril Burt who said intelligence is “inborn.” Stoddard reviewed numerous authorities on intelligence and discerned a change since the 1920s and 1930s: “a noticeable departure from flat statements about the constancy of the IQ and the inheritance of intelligence.”⁴¹ Similarly, McVicker-Hunt observed that “most of the general textbooks written before World War II tended to present the view that the IQ is essentially constant because intelligence is fixed.” He also noted that “a transformation has been taking place. . . . Evidence from various sources has been forcing a recognition of central processes in intelligence and of the crucial role of life experience in the development of those central processes.”⁴² According to David Gibson, all of the texts on mental testing used in the 1950s taught that cultural impoverishment could reduce scores on IQ tests.⁴³ Writers who asserted that intelligence is a hereditary trait were considered to be out of the mainstream of contemporary expertise. Altenburg criticized “extremists” on the political right who said heredity was everything as well as those on the political left who said heredity was irrelevant to intelligence.⁴⁴ Gates and colleagues also abjured “dogmatic statements of extremist viewpoints.”⁴⁵ It is evident that, according to the opinions of leading experts in intelligence testing in the 1950s, the AEB would unquestionably have qualified as a group of extremists on the political right.

Psychometric IQ

In the psychometric approach to measuring intelligence, a series of test items is first devised that is appropriate for children across a certain range of ages. Some items are quite easy for all but the youngest children, whereas others are beyond the capabilities of all but the brightest among the oldest children. The number of items a child can answer correctly is the raw score. On a well-designed test, the average raw score increases rapidly with age as mental growth occurs. Next, the test is administered to a large standardization sample of children that is supposed to be representative of the population to which the test will be applied in the future. This sample is used to establish age-specific norms for average, high, and low scores.

The WISC used a scaled deviation score to obtain the IQ. At each age, the average or mean score and the variability of scores among the different children in the standardization sample are computed. The statistical measure of variability is the standard deviation, which on the Wechsler tests is set at 15 IQ points. For any one child, the IQ score is based on the number of standard deviations by which the child exceeds or falls short of the mean. The Wechsler tests are scaled so that the formula for IQ is IQ = 100 + 15 (scaled deviation score). Tables in the test manual convert raw test scores to scaled scores at each age. For a child who remains consistently at the average score for her age over several years, the deviation score will be 0 and the IQ score will remain near 100. For one who scores two standard deviations above the mean, the deviation score is 2.0 and the IQ = 130, whereas a child scoring two standard deviations below the mean has IQ = 70. The theoretical distribution of IQ scores in a large population generally resembles a normal bell-shaped curve (see Figure 2.2a), with most people scoring near 100 and relatively few scoring very high or low. Because of the way the IQ score is scaled, it is possible to know how a child did on the test relative to his or her peers simply by knowing the IQ, even if the child’s age is not known. When another intelligence test, the California Test of Mental Maturity, was given to all Edmonton schoolchildren in 1956, the actual distribution of scores on an IQ test was very similar to the theoretical bell-shaped curve (see Figure 2.2b).⁴⁶

Figure 2.2 (a). Theoretical distribution of IQ scores when they occur as a normal bell-shaped curve. (b). Actual distribution of scores on an IQ test given in 1956 is close to a bell-shaped curve (Clarke, et al., 1978).

MENTAL DEFICIENCY

Psychometric Criterion for Mental Deficiency

The psychometric definition of mental deficiency is generally based on the IQ score. Several leading authorities classified mental deficiency as an IQ less than 70, and this criterion was widely adopted by professional organizations, including the World Health Organization in 1954.⁴⁷ The terms used to describe different levels of mental deficiency differed between countries and underwent revision to reflect social attitudes toward the mentally deficient in degrees of “mental deficiency” and “subnormality” (see table 2.1). In practice, Alberta also adhered to this criterion for clinically diagnosed mental deficiency; it was applied by the AEB. Dr. R. Kenneth Thomson (d. 1999), an internist, psychiatrist, and board member for more than twenty years and board chair from 1965 to 1972, stated explicitly that the cut-off score for mental deficiency was an IQ of 70.⁴⁸

Table 2.1 Categories of mental deficiency in relation to IQ score range

Source: Table adapted from Penrose (1949), p. 9, and updated as indicated.

Terman and others also recognized the range from IQ 70 to 80 as “borderline,” but it is clear that authorities on intelligence testing generally did not consider this range to constitute psychometric grounds for diagnosing mental deficiency. In 1955, the total patient population in each of the three diagnostic categories plus “borderline” mental deficiency in Alberta mental institutions (shown in table 2.2) revealed a scarcity of “borderline” cases.

Table 2.2 Diagnoses of 1,042 patients in Alberta mental institutions on December 31, 1955

Source: Annual Report of the Department of Public Health (1955), Province of Alberta, p. 87, p. 95, and p. 113. ICD is the International Classification of Diseases established by the World Health Organization in 1948 under auspices of the United Nations.

After several years of experience with IQ testing and school performance measures, psychologists began to define mentally deficiency using a statistical criterion: the lower 2 percent or 2.5 percent of the population.⁴⁹ To some extent the figure of 2.5 percent was based on social rather than strictly psychological criteria. As Wechsler pointed out, “Arbitrary and callous as it may seem, the definition of mental deficiency in terms of social criteria is a function of the number of individuals which a given community can afford to call so, that is, is able to institutionalize, or believes ought to be institutionalized if it had the resources to do so.”⁵⁰ In Alberta in 1955, a rather small proportion of the school-age population of children was institutionalized for diagnosed mental deficiency. The Department of Public Health’s annual report for that year reported that 193 patients in the Provincial Training School for Mental Defectives (PTS) at Red Deer were in the age range of ten to fourteen years.⁵¹ The Provincial Mental Hospital in Ponoka housed 80 patients with a primary diagnosis of mental deficiency in 1955, but only 2 of 482 admissions that year involved persons less than fifteen years of age. The Oliver Mental Hospital housed 191 patients with a primary diagnosis of mental deficiency, and comments in the annual report made it clear that some of these were “in the school age group.”⁵² If the age distribution was approximately the same as in the PTS (27.6 percent in the age range of ten to fourteen years), then there might have been another 53 mentally deficient children housed in Edmonton. Thus, about 246 Alberta children between the ages of ten and fourteen were institutionalized for mental deficiency in 1955. The annual report of the Department of Education reveals that 90,393 children in this same age range were enrolled in school at the time.⁵³

Undoubtedly there were other children in this age range who were neither in school nor in a mental institution. Thus, if roughly 246 out of more than 91,000 children aged ten to fourteen years were institutionalized for mental deficiency in 1955, this amounts to about 0.3 percent or 1 per 300 Alberta children, a figure somewhat higher than in most industrialized countries in the 1930s.⁵⁴ Nonetheless, the patient population in that age range was much smaller than the expected 1,800 mentally deficient children in Alberta at the time 2 percent of 91,000 children.

That is, no more than 14 percent of the children expected to have an IQ score of less than 70 were actually institutionalized for mental deficiency. It was also expected that far more Alberta children would be in the IQ range of 50 to 69 than in the range of 20 to 49 (see Figure 2.2), yet the latter category was much more common in Alberta mental institutions (see table 2.2). Hence, most children in the 50 to 69 IQ range were not in a mental institution at the time. Some of them were attending special schools but children in these schools were never presented to the AEB for sterilization.⁵⁵ Many others were simply kept at home and did menial chores.

All leading authorities at the time stressed that mental deficiency in the clinical or legal sense should never be diagnosed solely on the basis of an intelligence test.⁵⁶ Wechsler remarked, “Intelligence, like personality, is too complicated an entity to be defined by a single number. . . . [I]ndividuals having the same IQ’s may differ considerably in either their actual or potential capacity for intelligent behaviour.”⁵⁷ American psychologist Florence L. Goodenough (1886–1959) referred to one of the “unfortunate errors in thinking . . . [of] those who, on the basis of a single test, even when given at a tender age, are ready to ‘diagnose’ the child’s present mental level, make predictions as to his future, perhaps even take action with respect to matters of vital importance for his future.”⁵⁸ The problem of test interpretation was especially apparent for children living in rural areas, who tended to score lower than urban children because of their poorer schooling. The average IQ score in an Alberta survey was fully ten points lower for children from ungraded (one-room) rural schools than those from large urban centres, and it was also ten points lower for children living in homes where a language other than English (mostly French or Ukrainian) was spoken.⁵⁹ Goodenough observed that IQ tests “deal with the results of learning, from which capacity to learn is inferred. When opportunity and incentives have been reasonably similar, the inference is sound, but its validity may be questioned when a comparison is to be made between two or more groups for whom these factors have been markedly different.”⁶⁰

CONSEQUENCES OF STERILIZATION FOR THE GENETIC QUALITY OF THE POPULATION

The original purpose of the 1928 Sexual Sterilization Act was to reduce the frequency of social failure and misconduct in Alberta by preventing the reproduction of individuals who were believed to possess defective genes.⁶¹ The crusaders for passage of the act in the 1920s included women’s rights activist Emily Murphy (1868–1933), a police magistrate, and the United Farm Women of Alberta (UFWA). The minister responsible for the act in 1928 was George Hoadley (1867–1955), a farmer. There was no indication that any of these people understood genetic principles. Later, in 1950, a government committee said the object of the AEB was “to reduce the level of hereditary mental defects,”⁶² yet no person competent in genetics was appointed to the AEB or advised the board on technical matters until May 27, 1960, when Margaret Thompson (1920–2014), assistant professor of human genetics at the University of Alberta, was appointed.⁶³

Heredity

When considering mental deficiency, two basic kinds of hereditary defect may be involved: discrete or multi-factor. These kinds of defects are diagnosed in fundamentally different ways, and they differ in the likelihood of transmission to progeny. The consequences of sterilization can be computed for both kinds of defects using appropriate methods readily available in the 1950s and 1960s. Discrete defects of heredity are changes (mutations) in a single gene (DNA) and are inherited according to distinctive patterns known as the Mendelian laws.⁶⁴ The DNA contains a code for the structure of a protein molecule, and a mutation can lead to an altered protein that does not function correctly. This in turn can lead to mental deficiency. Mental deficiency itself is not transmitted from parent to offspring. Clearly, those who decided on the wording of Alberta’s Sexual Sterilization Act did not possess an adequate understanding of heredity. Multi-factor defects involve the combined effects of several defective genes as well as adverse environmental factors. More elaborate statistical methods are used to study and predict them.⁶⁵ Discrete defects were discussed at length by Canadian biology professor Roy Fraser (d. 1956), British psychiatrist Lionel S. Penrose (1898–1972), American geneticist and eugenicist Shelden C. Reed (1910–2003), and others, and those works on human genetics were widely available at the time.⁶⁶ Table 2.3 lists disorders that in the mid-1950s were believed to be caused by a genetic defect and were also known to cause mental deficiency in most cases. Most involved relatively severe abnormalities and were therefore relatively rare in the population. These disorders were also accompanied by characteristic physical or physiological symptoms and were not purely mental. A genetic disorder would ideally have been diagnosed by someone with expertise in both human genetics and pediatric medicine, although in the 1950s the profession of medical genetics and genetic counselling was in its infancy in Canada.⁶⁷ In that decade, three kinds of specific genetic transmission were recognized: recessive, dominant, and sex-linked inheritance. The sex-linked form was not encountered by the AEB and is therefore not discussed here.

Table 2.3 Genetic defects known in the 1950s to cause mental deficiency

Source: Fraser (1949), Penrose (1949), Reed (1955). Population frequencies from Fraser (1949) and Reed (1955) may be higher for some ethnic groups and even zero for others.

Recessive disorders

The effects of sterilization depend strongly on the mode of inheritance, whether recessive or dominant. In the 1950s several recessive genetic disorders, such as phenylketonuria, were well documented (see table 2.3) and must have appeared in Alberta institutions for the mentally deficient, although several recessive genetic disorders may not have been diagnosed as mental deficiencies. Most of them were relatively rare in the population because afflicted individuals were usually unable to reproduce because of infertility or simply lack of viability. That is, natural selection had over the centuries reduced the defective form of the gene to a rather low frequency in the population. The idea behind eugenic sterilization was that deliberate prevention of reproduction would reduce the frequency of the defective gene even further and thereby prevent many future cases of the disorder. Yet how great would that reduction actually be? This can be calculated with the aid of principles well established in the 1950s.

A recessive disorder is one in which the child must inherit two defective copies of the gene, one from each parent, in order to have the disorder. A child who inherits one defective and one normal form of the gene will be normal, although he or she will be a carrier of the defect. Cystic fibrosis is a well-known example of such a disorder.⁶⁸ The typical pattern for a rare recessive disorder is that neither parent has the disorder (although both are carriers of one defective gene) and only 25 percent of their children have it. This kind of disorder often occurs in an otherwise normal, healthy family. Thus, recessive genetic disorders do not appear to run in families. The afflicted child does not resemble the parents’ mental abilities. The defective gene is hereditary, but the disorder itself is not.

Let us consider a numerical example. Suppose a recessive genetic disorder afflicts one child in 2,500 in the population. This would be more common than any serious genetic disorder known to cause mental deficiency in the 1950s (see table 2.3) and overstates the likely effects of sterilization. Knowing the frequency of the disorder, the gene frequency can be calculated. Suppose there are only two forms of the gene present in the population, one being quite normal (+) and the other being a harmful recessive mutation (k). Genes occur in pairs in each person—one inherited from the father and one from the mother—and the specific pair constitutes the person’s genotype. If there are only two kinds of genes (+, k), there can be only three possible genotypes (+/+, +/k, k/k); only genotype k/k will have the disorder, whereas persons +/k will be carriers of the mutant gene but not manifest the condition.⁶⁹ Table 2.4 presents the most likely frequency of each kind of individual in a hypothetical population of 100,000 people, each with two genes for a total gene “pool” of 200,000 genes. The table reveals a well-established fact about rare recessive disorders: the vast majority of abnormal genes are carried by individuals who show no clinical signs of the condition itself. In this example, 98 percent of the k genes are possessed by carriers. In the 1950s there was absolutely no way to determine who was a carrier except by observing their offspring. If a person had an afflicted child, that person must have been a carrier.⁷⁰

Table 2.4 Population frequency of a rare recessive disorder

The second step in this exercise is to consider reproduction. For purposes of explanation, let us presume that each person, even one afflicted with the disorder, is fully fertile and that couples have an average of four children, a reasonable number for the 1950s. Because there is no way to detect a carrier prior to mating, we may fairly expect that mating relationships are random pairings with respect to the three genotypes, although mating relationships will of course not be random with regard to other valued and visible qualities. In the population of 100,000 people, half males and half females, there will be 50,000 pairings that yield a total of 200,000 children.⁷¹ Table 2.5 depicts the expected frequency of each kind of pairing and the composition of the offspring generation. Two salient points are evident from the table. First, almost all of the afflicted children are produced by couples where neither father nor mother showed any indication of the specific mental disability, because they were carriers, not genotype k/k. Second, the vast majority of mutant k genes are contributed to the gene pool of the next generation by couples who appear normal and have no afflicted children.

Table 2.5 Numbers of matings and children from a population of 100,000 parents

Note: Numbers in bold are those who might be eliminated by eugenic sterilization.

Knowing these figures, we can take the crucial third step to see what difference eugenic sterilization would make. The Sexual Sterilization Act of Alberta was supposedly intended to curtail reproduction of those actually afflicted with mental deficiency and would have had no impact whatsoever on the breeding of carriers, who would have been mentally normal.⁷² The process would be highly inefficient because only 164 of the 8,002 abnormal k genes would be possessed by people showing the disorder. Suppose that all afflicted children with genotype k/k were institutionalized and then sterilized. This would result in no reproduction by two categories in table 2.5: afflicted x normal and afflicted x carrier. Of course, the afflicted by non-carrier mating would never yield any afflicted offspring, even with no sterilization, but there would be about 152 carriers among their offspring. The consequence for the next generation is that there would be 160 fewer children, 4 fewer afflicted children, and 164 fewer mutant genes in the gene pool. Thus, the frequency of the disorder in the population would decline from 1 per 2,500 to 1 per 2,595, and the frequency of the mutant gene would decline from 2.0 percent to 1.96 percent. In order to prevent the birth of 4 afflicted children, the birth of 156 children not afflicted with the disorder would also be prevented by eugenic sterilization of afflicted persons. From a purely genetic standpoint, then, sterilization of those afflicted with a rare recessive genetic disorder would have a trivially small impact on the future generation of children in that population. As British geneticist Lancelot Hogben (1895–1975) pointed out, to have a major impact on the frequency of mentally defective children in a future generation, it would be necessary to sterilize the prospective parents who were definitely carriers, plus the siblings, most of whom were carriers, not just individuals with the disorder.⁷³ In practice, the process would be even less efficient because (a) it is unlikely that every child with the disorder would be institutionalized unless it was a severely disabling defect, and (b) if it did cause severe disability, reproduction would most likely be curtailed without any order by a eugenics board. Furthermore, there would be diminishing returns: the change in gene frequency would become even smaller each generation as the defective gene becomes rarer. Indeed, as shown by Lush, to reduce the frequency of a harmful recessive gene from 2 percent to 1 percent in a population by sterilizing all afflicted children would require fifty-one generations of selective breeding, or more than one thousand years.⁷⁴

Dominant disorders

A dominant disorder occurs when every person carrying just one copy of the mutant gene shows signs of the defect. A classic example is the genetic neurodegenerative disease identified by the American physician George Huntington (1850–1916), a neurological syndrome of mental and motor deterioration beginning in middle age. In such a case, there is almost always one parent who died from the same disease, and about half of the children are afflicted if they live long enough. This defect does indeed occur in families and the risk of it occurring in the offspring of a carrier is quite high, 50 percent.⁷⁵ If all children harbouring a defective dominant gene were institutionalized because of the disorder and later sterilized, the scourge could be purged from the population in one generation and never reappear, except in the extremely rare eventuality of a new mutation caused by radiation or a dangerous chemical. State-mandated eugenic sterilization could have a major impact on a dominant genetic disorder if the symptoms are typically obvious and manifest in childhood.⁷⁶ It would have virtually no impact on a neurological disorder such as Huntington’s disease, where symptoms typically appear after the age of reproduction and carriers cannot be detected until symptoms appear.⁷⁷

Non-specific, multifactorial disorders

For characteristics such as IQ, instances where a single genetic mutation has major effects are very rare; instead, development is influenced substantially by both heredity and environment. The hereditary factor likely involves defective variants of several genes.⁷⁸ The only recourse here is statistical. The method involves the coefficient of “heritability” (h²), which estimates the percentage of variation in IQ in the population that is attributable to genetic variation. If the value of heritability of a characteristic is known, the response to selective breeding can be estimated from the formula R = h²S.⁷⁹ The selection differential S is the difference between the average score of the original population and the average score of those individuals chosen to be parents of the next generation. In agricultural or laboraory applications with animals, usually a small fraction of the population is chosen for breeding, perhaps the top-scoring 5 or 10 percent, and the selection differential is quite large.⁸⁰ The response to selective breeding (R) is the difference between the average of the original population and the mean score of the offspring of the selected parents. The h² coefficient is very difficult to estimate accurately with humans and its interpretation is controversial.⁸¹ Nevertheless, a numerical exercise can be done to get a very rough idea of the prospects for eugenic improvement of human intelligence. Let us assume h² = .5, being neither particularly high nor low. When the American psychologists John L. Fuller (1910–92) and W. Robert Thompson (d. 1960?) first reviewed the literature on heritability of intelligence in 1960, they made an educated guess that the weighting for heredity should be a little heavier than for environment.⁸² More recently, Devlin et al. arrived at a smaller value of about h² = .4.⁸³

In order to calculate the expected response to selective breeding, we first need to know the mean score and variation about the mean for the original population before selection begins. The precise distribution of IQ scores of all Alberta children in the 1950s was probably was not too far away from a bell-shaped curve with a mean IQ of 100 for Alberta school children (see also Figure 2.2).⁸⁴ Thus, for the purpose of this exercise, the number of children in each IQ score range in the population is computed from a bell-shaped curve. The most challenging task is to find the average IQ of those allowed to breed. This is done for two situations: the first uses data on Alberta children actually sterilized to make an educated guess; the second presumes that every child in Alberta with an IQ score less than 70 was sterilized, regardless of whether he or she was in an institution. This involves far more children than were actually sterilized, and consequently it yields an estimate of what the AEB might have achieved with a maximally efficient program based on IQ alone.

Table 2.6 Sterilization of 14-year old children in a population of 18,000 children

For the first calculation, the numbers of Alberta children of a particular age and IQ range who were sterilized in 1955 are estimated. In 1955, in Alberta, there were about 18,000 fourteen-year-old children.⁸⁵ The AEB knew exactly how many of which age and diagnosis were considered for sterilization, but we must assemble the picture from shards of evidence. Seven of the cases I reviewed were seen by the board at the age of fourteen, and this is also an age when most normal children were still enrolled in school, so let us use the numbers for fourteen-year-old children. Available evidence indicates about 37 such children in the IQ range from 50 to 79 would eventually be sterilized and then discharged, as shown in table 2.6, amounting to about half of the 75 institutionalized children at that age. Because those in the IQ range below 50 would likely not have been considered for discharge in any event, or would almost certainly have been infertile,⁸⁶ with or without the Sexual Sterilization Act, the infertility of only the 37 individuals in the IQ range 50 to 79 should be ascribed to the eugenic sterilization policy. Removing those 37 from the population of 18,000 children born in 1941 would yield a mean IQ of the remaining 17,963 children equal to 100.086. Thus, the selection differential would be S = 100.086 – 100.0 = 0.086, the selection response would be R = 0.5(.086) = 0.043, and the expected mean IQ of the next generation would be 100.043. An increase in average IQ score in the population of 0.043 points is trivially small.

Next, suppose the act required the AEB to seek out and sterilize every child in Alberta with an IQ score of less than 70 while not tampering with the reproductive organs of those with an IQ of 70 or more. There would have been about 49 of those children with an IQ of less than 70 but more than 372 among a total of 18,000 children at that age. The mean IQ of those allowed to breed under this scenario would have been 100.758. The selection differential would have been S = 100.758 – 100.0 = 0.758. Thus, the improvement in average IQ of the Alberta population would have been about R = 0.5(0.758) = 0.379 or about one-third of an IQ point. This would have been difficult to support politically, because it would have included every child below IQ 70, including those dearly loved, nurtured, and sheltered by their parents. The point of this second calculation is to show that even a tenfold increase in the number of children sterilized would have had very little impact because it would still entail a small minority of all children.

To have an appreciable impact on the next generation of Alberta children, unrealistically large numbers would need to be subjected to the eugenic surgeon’s knife.⁸⁷ For example, if only children in the top 10 percent of the IQ distribution were approved for parenthood, the average IQ of this elite group would be 126.3 and the selection differential would be 26.3 IQ points. The response to selective breeding in the first generation would then be about 10.5 IQ points and the new population average score might rise to 110 points. The parents would need to contribute at least twenty children per couple to the next generation to compensate for all the forced infertility.

The conclusion from this exercise is that, according to well-known genetic principles that were well understood in the 1950s and taught at the University of Alberta at that time, the likely effect on the average IQ of Albertans from a program of eugenic sterilization similar to what was implemented under the Sexual Sterilization Act would have been trivially small. This is one of the major reasons why many reputable geneticists deserted the eugenics movement in the 1920s and 1930s after the genetics of human populations became better understood. Stoddard, in 1945, observed wryly that “there has grown up a certain disjunction between the sober writing of geneticists and the expectation of eugenicists.”⁸⁸

SOCIAL AND EDUCATIONAL ENVIRONMENT

Having found that the genetic consequences of eugenic sterilization in Alberta in the 1950s would have been small indeed, it is informative to consider what degree of change in IQ was to be expected from improved environments. Thorough discussions of social and educational factors related to childhood intelligence were provided by Stoddard, Harold E. Jones, Anastasi, and McVicker-Hunt.⁸⁹ There was considerable dispute in the 1950s and 1960s about the potency of the psychological environment in shaping intelligence. Authors such as Jones tended to minimize environmental effects while others, including McVicker-Hunt, believed intelligence to be quite malleable, especially in early childhood. However, there was little dispute as to whether environment had any effect. Virtually all authorities acknowledged the importance of upbringing, especially extreme departures from the normal mode of life. The big unknown was the precise strength of the influence of specific kinds of experience in terms of points on the IQ scale.⁹⁰ In many examples, the change in IQ amounted to five to fifteen points. Table 2.7 summarizes results of several studies of environmental effects that were well known in the 1950s. The difference in IQ score attributed to each effect represents the difference between two groups of people: one with and the other without the environmental treatment in question. The IQ point differences provide an approximation of the kinds of effects that were to be expected in the 1950s.

The cohort effect was a direct test of the predictions of many leading voices in the early eugenics movement who claimed that people of lower intelligence usually had larger families and, therefore, intelligence would decline over the years. Birth cohorts are entire populations of children born in different years. If national intelligence is changing, those born many years apart should differ substantially on an IQ test. Of course, improved medical care and education over a period of several decades could increase intelligence if environment plays a major role. There was clear evidence of a general increase in the intelligence test performance of Americans. For instance, soldiers taking an IQ test during World War II scored almost twenty points higher on average than those in World War I.⁹¹ A portion of this improvement was evidently related to education; the average years of schooling of American men in World War I was eight and in World War II was ten. When the Stanford-Binet test was re-standardized in 1960, children scored about five IQ points higher on the older 1937 version of the test that had become culturally out of date than did the 1937 standardization sample.⁹² In 1936 and 1937, Raymond Bernard Cattell (1905–98), personality psychologist and eugenicist from the University of Illinois, compared IQ scores of parents and their children and reported data showing the children scored ten to fifteen IQ points higher than their parents.⁹³ Cattell predicted a subsequent decline in national intelligence but thirteen years later was surprised to find evidence of an increase in IQ.⁹⁴

Table 2.7 Environmental effects on children’s IQ scores

Rearing in a rural environment was associated with lower IQ scores, and the longer a child remained in a rural setting, the greater was the disadvantage.⁹⁵ Medical care and education were generally inferior in rural settings from the 1930s to the 1950s, and much of the IQ difference from urban children might have been caused by the environmental difference. Some academics speculated that the pattern might have resulted from selective migration, whereby the genetically brightest people moved to the cities, leaving the biological dolts and laggards behind on the farm. Evidence in support of this view was generally lacking. In more recent times in Alberta, the government has done much to improve rural health and education, and the urban-rural difference has now largely disappeared.⁹⁶

Institutional rearing, especially in orphanages where staff were overworked and had little time to give their wards individual attention, was widely known to impair intellectual growth.⁹⁷ The longer the residence in the institution, the lower the IQ score compared with the cohort of the same age living at home and attending regular schools. This finding was based on IQ testing from the Iowa Adoption Study of 1949; these biases from the poor staffing conditions and low adherence to observational and testing standards of course influenced the numerical distributions of IQ scores when they occur as a normal and the actual distribution of scores on an IQ test given in 1956 being close to a bell-shaped curve (see Figure 2.3). Poor development was not an inevitable consequence of being in an institution, however. An enriched training program could avoid most of the retardation usually occasioned by neglect.

Foster home rearing and adoption were known to facilitate mental growth when an infant from an impoverished background was placed into a prosperous home with well-educated parents. One of the most widely cited studies involved adopted children in the state of Iowa.⁹⁸ Each child chosen for study had been adopted prior to the age of six months and IQ test scores were available for both the birth mother and, later, her biological child during school years. One hundred children were given three IQ tests over a period of several years. As shown in Figure 2.3, the average IQ score of the sixty-three birth mothers was 85.7, whereas the average score for their one hundred children was 114.8, a difference of about 30 IQ points. Fourteen of the mothers had IQ scores in the range of mental deficiency, but none of their children was below an IQ of 80. A major portion of the improvement must have been environmental.

A variety of enriched early education programs, especially those involving special language training for normal children, were known to enhance IQ test results by a few points.⁹⁹ Since the studies cited above were done, many further research projects have been conducted with better controls and better methods of testing. As summarized by Douglas Wahlsten, major improvements in environment are now well established as causing an upward change of as much as fifteen points on IQ tests.¹⁰⁰ The cohort effect has become very widely known.¹⁰¹ The early eugenics movement raised an alarm about a pending decline in national intelligence if people with low IQ scores were allowed to reproduce. Now we are witness to perhaps the ultimate refutation of this raison d’être for the eugenics movement. Since World War II, intelligence, as measured by IQ tests, has in fact increased very substantially in every industrialized country where large-scale IQ testing has been done.¹⁰² The precise magnitudes of environmental effects were not known in the 1950s and 1960s, but there was sufficient evidence at the time to show that the effects were at the very least increases of several IQ points, far larger than any paltry boost from eugenic sterilization on the scale that was actually practised in Alberta.

Figure 2.3 Illustration of the IQ testing from the Iowa Adoption Study, 1949. Top: IQ scores of birth mothers having little formal education who gave up at least one child for adoption. Bottom: IQ scores of 100 adopted-away children of those mothers who were reared in good homes. Based on data provided by Skodak and Skeels, 1949, compiled and designed by Douglas Wahlsten, 2013.

BLINDED BY THE EUGENICS CREED

It is a chilling and distressing experience to read patient files from the PTS and realize the utterly horrid conditions in which many had been living prior to admission to the institution. The AEB ignored this information completely. For those devoted adherents of the eugenics creed, intelligence was fixed by heredity. Consequently, no additional influences needed to be considered, and they were not. The board simply accepted the recommendations of the superintendent of the PTS, presented in a brief case summary, which affirms the contention of Hansen and King that the superintendents of mental institutions played crucial roles in decisions for sterilization.¹⁰³ Cases were heard by the AEB at a rate of about one case every five minutes.¹⁰⁴ For many of the children confined in the PTS, life there was actually better than at home. Leilani Muir (b. 1944), for example, had been starved and beaten at home, but in the PTS she began to gain weight and made some lifelong friends. Her IQ score in 1957, shortly before she was sterilized, was 71, which is in the normal range, but she was sterilized anyway as the result of an error in totalling her scores that placed her at an IQ of 63.¹⁰⁵ Several years after leaving the PTS, as an adult, she scored in the range of 85 to 90, which was a clearly normal performance. In another example, a boy in the PTS was tested four times from 1957 to 1961 and showed a clear improved pattern of IQ each time he was tested; his score increased from 42 to 76—until he was above the cut-off for mental deficiency—yet he too was sterilized (see Figure 2.4).

Figure 2.4 IQ test comparisons from the Leilani Muir Case, 1957 to 1989. Left: A boy in the Provincial Training School in Red Deer was given an IQ test four times and improved each time until he was above the cut-off for mental deficiency, but the Alberta Eugenics Board ordered him sterilized. Right: Leilani Muir also scored above 70, but an error in totalling her scores placed her at an IQ of 63. As an adult, she scored higher than 85, a clearly normal performance. Reproduced by permission from Douglas Wahlsten.

In one case reviewed by the author, police and social worker reports of the home situation provided vivid testimony about the circumstances of Ilsa Anderchuk (a pseudonym):

[Report of July 8, 1949:] There are about 5 or 6 small children . . . and the parents living in a small one room shack. The shack is filthy and is practically without furniture. The mother and children are filthy and are practically without clothing. . . . The mother is a mental defective who cannot speak English. The father can talk English but does not appear to be overly bright. . . . I have seen plenty of pretty poor homes in this and other districts but this is one of the very worst and conditions can only be described as deplorable. . . .

August 24, 1949 (W. Filewych, Welfare Inspector): I don’t believe I have ever seen anything like it before. The mother is a mental case. She has been in the Ponoka Mental Institution for quite a while. The husband took her out in 1944 knowing that her condition was not in any way bettered. I spoke to her for a few minutes and she is definitely not in her right mind. . . . She was only scantily dressed. . . . She sometimes wanders away from home. . . . She does not wash or feed the children. . . . There have been four children since the mental condition set in. . . . These children should either be placed in a children’s home or perhaps be looked after by relatives. It is hard to believe that in the twentieth century conditions such as these exist. There is evidence of severe neglect on part of both parents. . . .

[Nov. 10, 1949 (RCMP): The home] was found to be in the filthiest condition imaginable. The children were only partially clothed. . . . Two basins of slops sat on the stove which had been there for seven days at least. . . . Mrs. ANDERCHUK’s major problem has undoubtedly been the raising of her children, and their ages will give an indication of the strain the woman has undergone, and which no one can gainsay is the reason for her present mental condition.¹⁰⁶

This report then identifies the children, aged one, two, four, five, eight, and nine years. It is apparent that young Ilsa, the nine-year-old, was the eldest sister, who must have actually borne the principal responsibility for raising her younger siblings under extremely difficult circumstances. There would have been no time to devote to schoolwork. Ilsa was later interviewed briefly by the AEB, at the age of thirteen, one month after scoring 63 on an IQ test. The report of the 1954 Stanford-Binet IQ test described the girl as a “mid-grade moron” and included comments—“Rather tense—short of breath. ‘sob breathing.’ Poor in attention and memory”—indicating she was very anxious during testing. Ilsa was ordered to be sterilized, for the following reason: “Danger of the transmission to the progeny of mental disability or deficiency, also incapable of intelligent parenthood.”¹⁰⁷ Evidently the eugenics creed blinded both the superintendent of the PTS and the eugenics board to what was happening right before their eyes. They held fast to their belief in hereditary intelligence despite abundant evidence to the contrary. They boldly asserted genetic explanations without having an adequate understanding of the genetic or psychological science existing at the time.

CONCLUSION

Many important human traits such as intelligence, personality, and emotions are very complex, cannot be traced to simple genetic origins, and are influenced in a substantial way by experience. There are only a few well-defined disorders of the nervous system that are caused by a defect in a specific gene (as Mendelian inheritance) that is inherited from the parents. The more severe defects are generally very rare. A eugenics program to reduce the frequency of these defects in the population can be founded on good scientific principles that were fully elucidated by the 1930s. Whether such a program should be voluntary and implemented through education and counselling or forcible by edict of the government is a question of ethics and politics, not biological science.¹⁰⁸ A eugenics program to alter these traits in the population was proposed in Alberta based on principles used to selectively breed improved strains of farm animals. Those principles were well understood by about 1940 and were widely taught in schools of agriculture during the 1950s. There is, however, no evidence—as this chapter has argued—that the AEB (since its functional existence from 1929 onward) was well aware of or cared about those scientific principles. The practices of the board were apparently based on social prejudice, not genetic knowledge, and membership on the board evidently was determined by allegiance to a pseudo-scientific creed. That creed allowed the government of the time and the AEB to deny the fundamental humanity of their victims and authorize actions against them that were without a doubt crimes under Canadian law.