Genetics, Medicine, And Practical Eugenics

By J. A. Fraser Roberts and Eliot Slater. Eugenics Review 40: 62‑69 (1948).



     Inherited differences between human beings tend to fall sharply into two groups. First we have definite abnormalities and defects. Sometimes these are wholly deter­mined by heredity. Sometimes there is a genetic basis, though the co‑operation of other factors, often environmental, is needed if that genetic basis is to express itself as recognizable abnormality. These definite inherited departures from normality are practically always due to single genes. In the second place, however, when we turn to the hereditary differences which distinguish normal people it is rarely indeed that we find these to be due to the action of single genes. Multifactorial inheritance is the rule; that is, the effect we see is due to the combined action of many genes, each making a small contribution which is not individually distinguishable. So we get the familiar picture of continuous variation, the normal curve and certain average measures of resemblance between relatives.

    The distinction between these two modes of inheritance is a radical one, but it has not always been clearly drawn. In physical characters there was from the begin­ning no real confusion, but this was largely the result of accidental circumstances. Physical abnormalities were investigated by medical men by the pedigree method, and were found reconcilable with a single‑gene heredity. Quantitative variables, such as stature, were investigated by statistical methods by Karl Pearson and his school, and the apparently non‑Mendelian type of inheritance for which they were looking was found to apply.

    The inheritance of intelligence could not be adequately investigated until after the appearance of reliable quantitative tests. Up till that time it had to be judged on the basis of educational assessment or of social behaviour; thus some authorities of the early days thought in terms of qualitative distinctions between the mental defective and the normal, and of single‑factor inheritance. It is only fairly recently that we have been able to estimate the relative parts played by single genes and by total genetic equipment, and to separate clinically those forms of mental deficiency or backwardness attributable to the one and to the other. We now know that the hereditary component in the determination of the majority of instances of mental deficiency is multifactorial.

    When we come to the inheritance of temperamental qualities we are on less certain ground. We have no precise and reliable tests of temperament as we have of intelligence. Nevertheless, here too it seems probable that we have to look for single genes as the explanation of pathological events, and for a multiplicity of genes as the explanation of individual differences in the physiological range. Let us look first at the mental disorders of what is commonly called a "psychotic" kind, that is to say manic‑depressive illnesses, schizophrenia and paranoia, involutional melancholia, and the endogenous presenile dementias. In some of them a single‑factor inheritance has been demonstrated. This is so in Huntington's chorea and Pick's disease. Inheritance of the manic‑depressive psychoses is compatible with a theory of a single dominant factor, with the environment too playing a considerable role, although objections can be raised to this view. The findings with schizophrenia are even more difficult to interpret. But no serious worker has wished to depart from a single‑factor theory, although that single factor may be either dominant or recessive, or, most probably, schizophrenia may be not one but many diseases with different single genes responsi­ble from case to case. Generally speaking, the evidence suggests that specific single genes are essential for the appearance of the endogenous psychoses; but that multi­factorial inheritance also plays a part in determining constitutional resistance or susceptibility, just as it does in environmentally caused diseases such as tuberculosis.

    Neurosis and psychopathy have probably to be looked at in quite a different way. Almost certainly they depend on qualities of temperament and personality, which are determined by multifactorial inheritance. In clinical practice we are un­able to find any sharp dividing line between abnormal and normal personalities; the differences which exist must be of a graded and quantitative nature, whatever the particular temperamental trait we consider. Although our tests of temperament are nothing like as good as tests of intelligence, where they can be applied they show that the population is distributed in the normal curve. The work of Eysenck (1947) and his collaborators is particularly convincing on this point. From investigations on twins, we know that qualities of temperament and personality have a hereditary basis, and the sort of inheritance of those traits that are found running in families is readily compatible with a multifactorial theory, but very difficult to reconcile with any single‑factor hypothesis.

    In temperament and character, therefore, multifactorial inheritance holds, as it does in intelligence; but we must rate the importance of the environment consider­ably higher. Newman, Freeman and Hoizinger (1937), in their classic study of uni­ovular twins who had been brought up apart from one another from the earliest years, showed that there were larger differences between the twins in their response to tests of temperament than to tests of intelligence. Different environments had also imposed different patterns of development, and different life stories, on person­alities of the same constitutional and hereditary make‑up. In his balanced study of a large series of criminal twins, Kranz (1936) concluded that the fundamental struc­ture of the personality was largely determined by heredity, but that environment often exercised a decisive influence on behaviour. As pragmatists we are really more concerned with behaviour than with abstracted notions of the underlying person­ality; so we must make a larger allowance for environment in the problems raised by temperament, neurosis and psychopathy than we need do with those raised by intelligence and mental defect.


     From the point of view of practical eugenics, the kind of hereditary difference which is determined by multifactorial inheritance is far more important than that caused by single genes. Here are some of the reasons. First, those things determined, or partly determined, by multifactorial inheritance are more important socially. Intelligence, good physique, resistance to disease are more important to the com­munity than rare hereditary diseases, even including some of the low‑grade mental deficiencies, or those diseases which though rather commoner appear to have a definite and maybe simple genetic basis. In the same way, temperamental differences are more important than the insanities. All the hereditary mental illnesses together give a risk of only about 2 per cent in the general population, whereas the disabilities caused by neurosis and psychopathy are of fateful significance to us all.

    Secondly, and this is even more important, simply inherited abnormalities are individually rare or very rare. Seldom indeed does a hereditary defect attain a frequency of one in 5,000 in the population; it is usually much less than this. The sole exception is colour‑blindness, but this condition should perhaps be regarded as a more or less physiological variation rather than as an abnormality. Selection then, either natural selection based on natural differing rates of reproduction, or any conscious selection that it might be decided to apply, operates on a few human beings only. With recessive genes it is only very occasionally that we can identify genetically the outwardly normal carriers, so our general conclusion is little affected. Contrast with this multifactorial inheritance, determining in part some human quality such as general intelligence. Every human being exhibits the result of the action of many genes, some good, some bad. With the birth of every single human child the balance tips a little, up or down. Selection is operating, not on one person in five thousand, or fifty thousand, or five hundred thousand, but on every child born to every couple; for upon the sum total of their good and bad genes and the relative rates at which these are increasing or diminishing depends the future of the species, whether it is advancing, whether it is holding its own or whether it is retrogressing. Here is something about which the individual can have his say. Here is something not merely about which the State can legislate effectively if it wishes, but something which is automatically affected, for better or for worse, by social and economic measures of many kinds. We cannot confine our debate to the merits of action or inaction, for the actions of individuals, of groups and of governments are already exercising a continuous influence.

    There is a third reason why multifactorial inheritance is the more important practically. If the mute inglorious Milton and the village Hampden may be denied their opportunity by external circumstance, it is no less true that the potential genius, rich in good genes of many kinds, may be stultified by having the misfortune to possess just one single grossly deleterious gene. Too rigorous an attempt to eliminate such genes may prove harmful on occasion. To some extent this is so today, for it is the highly intelligent and conscientious who tend to limit the number of their children because of some harmful gene known to be present in the family group. Contrast with this the person whose unfortunate heredity is multifactorial and due to an accumulation of many bad genes. Selection in this instance is on balance practically devoid of risk.

    Finally, eugenic measures are much more hopeful in qualities caused by multi­factorial inheritance than they are with single genes. Even the strictest eugenic laws, such as the Draconian laws of Nazi Germany, would have remarkably little effect in reducing the incidence of recessive abnormalities below a certain level. But even a slight selective advantage will cause a growth in frequency of multifactorial genes. If we were conquered by an alien race from Mars, it would be quite a feasible project for our overlords to breed up humanity to the intellectual level of a world of first‑class honours men, or down to a world of morons, and in quite a few genera­tions. For the ease with which it lends itself to modification by almost invisible trends, multifactorial inheritance is of paramount significance in eugenics.


     In stressing the greater importance of multifactorial inheritance and the things with which it is concerned we are not trying to say that a consideration of inherited abnormalities is of no value, or, necessarily, that nothing can be done about them. But the field is limited and the useful application of our knowledge will often lie outside the field of eugenics. We must dismiss simple dominant and sex‑linked harmful genes as of minor practical importance. In the former every possessor is marked by the abnormality and so is fully exposed to the blast of natural selection. If the effect is at all serious the rate of elimination is so high that such genes cannot be other than very rare. Recessive sex‑linked genes are exposed to one‑third of this exceedingly severe selective force.

    Simple recessive genes present a much more intricate problem. We know that for every person we see with retinitis pigmentosa there are more than a hundred un­suspected heterozygous carriers; for every albino, two hundred and fifty; for every sufferer from alkaptonuria, two thousand outwardly normal people who carry the gene. Looking at it the other way round, while only one person in every five hundred carries the gene for such a rare condition as alkaptonuria, one person in every 70 carries the gene for albinism, and no less than one in every thirty or forty amongst us is a carrier of retinitis pigmentosa. Thus our population contains great numbers of these recessive genes; in all probability there are few of us who do not carry at least one. In this mankind corresponds to the rest of the animal and plant world, for wild populations too carry large numbers of harmful recessive genes, with little ill effect apparently on the race.

    The first lesson is a negative one. There is no eugenic reason why a sufferer from a recessive abnormality, or a known carrier, should not marry and have children, though, of course, he will be wise to avoid marrying a blood relative or anyone with a family history of the same defect. It is true that harmful genes are being handed on to future generations, but we are all of us doing the same thing, the only difference being that sufferers and their relatives know it and we do not. Here we must state our disagreement with the views of Professor Ruggles Gates as expressed in his Human Genetics (1946) and again in The Eugenics Review (1947). The eugenic effect of abstinence is negligible and if everyone adopted so conscientious an atti­tude towards posterity no one would feel justified in raising a family.

    Perhaps it is natural that in the past eugenic thought has concentrated on one particular line, namely influencing the next generation positively or negatively by increasing or decreasing the number of children of certain individuals. But the con­sideration of recessive defects leads us to another possibility, based on the fact that sometimes an individual can safely marry certain individuals but not others. Reces­sive defects crop up particularly when blood relatives marry because they tend to share the same genes in common, having received them from the same source. The incidence is greatly affected by changes even in the limited degrees of inbreeding allowed and practised in our community. The decline in the amount of inbreeding during, say, the past century must, as Haldane (1939) has shown, have greatly lowered the frequency of recessive defects. But of course mutation has been going on at its usual rate, and, as the rate of elimination has been reduced, the harmful recessive genes are slowly multiplying. Ultimately after many generations the old high incidence will once more be reached, but the proportion of carriers will be much higher. A sudden reversion to smaller, closed communities once again and relatively close inbreeding would result in a marked outbreak of recessive defects of all kinds.

    In the absence of a family history of a recessive abnormality the increased absolute risk of a consanguineous marriage is small. But the relative risk, which is what matters to the whole community, is greatly increased. Would the community be wise to discourage cousin marriage? If it were prohibited the immediate effect would be to lessen appreciably the amount of recessively determined abnormality, but in the long run this would be at the price of a greater accumulation of the genes in the population. The case for interference does not seem a strong one.

    Conscious control of recessive defects would be enormously easier if the hetero­zygous carriers could be identified. There is the promise of this in some current work. Provided that the heterozygotes could be identified with certainty the defect itself need never appear provided that marriage between the carriers were avoided. This is not merely a theoretical impossibility. It could be realized at the present time in erythroblastosis foetalis, a condition which is common compared with ordinary inherited defects, occuring say once in every 200 births. Confining ourselves to what is practically important, the abnormality need not arise if Rh negative women do not marry Rh positive men. But is it likely that 15 per cent of women will accept the proposition that they can only marry their like and that 85 per cent of the male population is barred to them? If such an attitude were, however, to become general we should see an interesting social development, for our population would split into two diverging strains, between which marriage rarely took place.

    Some reference has been made to an apparently simple genetic basis, or some­thing that looks like it, for certain not very uncommon conditions such as thyrotoxi­cosis, essential hypertension, or schizophrenia. But in all such instances the heredi­tary basis is only part of the story; the co‑operation of other factors, probably environmental, is necessary if genetic potentiality is to become transformed into the disease itself. If these single genes exist they must be relatively common and widely spread. Their elimination, or even substantial reduction, is probably impossible and in schizophrenia, for example, the tendency to self‑limitation is very strong. It would seem much more hopeful to concentrate on the environmental component and discover how to prevent the disease appearing in those whose genetic constitu­tion renders them susceptible.

    We are not going to claim that in all the foregoing instances there is no scope for eugenic action. The importance of such deformities and defects to the sufferers and their families may be very great. Much relief can be given by wise advice and by suitable legislation. The recent informative paper read to this Society by Professor Tage Kemp (1947) on fifteen years' experience in Denmark of eugenic legislation shows the importance of these humanitarian measures. But that experience also illustrates the limited scope and effect of such measures when we think of the community as a whole. In our country a few thousands of individual cases a year would be the utmost number affected, and even of those many would have, as we have seen, little eugenic significance. In those things determined by multifactorial inheritance lies the true and overwhelming importance of the application of genetics to practical eugenics.


     There is a large genetic component in human differences in certain physical traits, for example stature. This is also true of certain mental qualities, in particular intelligence. And there is doubtless a genetic component, also multifactorial, in much that goes to determine a sound physique and also degrees of resistance to certain diseases. As we saw earlier, variation is continuous, and selection is operating over the whole range of the human population. Either the good genes are multiply­ing at the expense of the bad, or vice versa; or, it may be, the population in regard to some of these things is in a state more or less of equilibrium. Now we have great sympathy with those who would claim that as environmental factors are also con­cerned we should concentrate on these first. Let us, it might be contended, exhaust the possibilities of reducing inefficiency and disease by improvements in environ­ment before thinking too much of the less tractable hereditary differences. And even when heredity is clearly important, the most useful practical applications may well lie not in attempting to influence relative birth‑rates but in taking special precau­tions to safeguard those who are known to be genetically susceptible. Let us take pulmonary tuberculosis as an example. The fine twin study of Kallmann and Reisner (1943) has shown most clearly that in the present state of our community heredity is of great moment in determining whether or not a person will develop the adult form of the disease. But is the lesson of this that those with a bad family history should not marry? In our opinion, no. There are so many unfavourable things that can be inherited and this is only one. It might well be that other good qualities would more than outweigh this particular weakness. The lesson is rather to protect those likely for genetic reasons to be highly susceptible: by periodical mass­fluorography for example, by supplements to the diet, by seeing that such people do not enter the tuberculosis service, by discouraging the nursing of a tuberculosis patient by a sister.

    When we turn to temperament we come up against objections of another kind. Few would deny the advantage of good build, fast and accurate neuromuscular co‑ordination, strength, physical health and a high natural resistance to disease. We can surely encourage those endowed with such qualities to regard themselves as proper parents for the next generation. But in the field of temperament, our aims are much less clear. There is no such thing as a good and desirable type of tempera­ment, and a bad and undesirable one. It takes all sorts to make a world. There is room for the quiet, the docile and the stable; but there is room also for the man of strong passions, for the natural born rebel, for the man whose moods take him freer phases of elation and enhanced activity into troughs of depression. It is true, indeed, that a man like Hitler can cause much greater harm to humanity than even a million mental defectives; but we cannot say, if Hitler had had children, that they would have been any more likely to have been bad citizens than anyone else's. It is particular combinations of temperamental qualities which are procluctivc ut sarge cl iea good or bad. Recombined in another way, the effect of the genes responsible wow have been different. In fact we are too uncertain of where we want to go, and t ignorant of the foundations on which we stand, for there to be at present a practical eugenic approach to the betterment of human personality.


     We come at last to intelligence. Even here, perhaps, the nature of our aims is not beyond all controversy. There are many jobs in society which do not demand a high level of intelligence; and we know that when people are set to jobs which provide no adequate outlet for their intelligence they are often unhappy and difficulties are likely to arise. But there would be general agreement that we are short of men and women of high intelligence, that the number of skilled jobs is increasing and that with the advent of mechanization unskilled jobs are becoming more and more skilled. The world in which we live is becoming more and more complex. If the average citizen is to understand it, and play his part in a democratic society, better education and better intelligence are both required. And a rise in the average level of intelligence would permit a fuller and richer mental life in the community.

    The attitude of concentrating on environmental improvement and of doubting the great urgency of positive eugenic measures seems to us to be justified only if one important condition is satisfied. This is that we have no reason to suppose that those of poorer genetic endowment are reproducing appreciably faster than those who are superior. Thus we can be philosophical about heredity and pulmonary tuberculosis. Hygienic measures are proving efficacious in lessening its incidence and there is no clear evidence that those who happen to carry an excess of genes making for poor resistance are reproducing any faster than the rest of the population. In all proba­bility we should be justified in assuming that owing to some elimination at a rela­tively early age the reverse is true. But suppose that we were told that those with the strongest hereditary predisposition to pulmonary tuberculosis were reproducing twice as fast as those who were most resistant, could we afford to take so detached an attitude? Surely not. And yet this is what is happening in regard to one quality ‑general intelligence ‑ a quality of the highest social importance and one determined in large measure by heredity.

    We do not propose to discuss the nature of what is measured by intelligence tests or to review the evidence which points to a large hereditary component. This has been done very recently by Burt (1946) with whose views we are in entire agree­ment. Whether we accept the estimate that three‑quarters of the variance in intelli­gence as measured is due to heredity or whether we are ultra‑cautious and simply say at least one‑half, it is evident that the problem is urgent. For here is no question of speculating on the composition of a more or less stable population and debating at leisure whether some change is desirable, whether it could be made, and if so how. We are confronted with what appears to us a ruinous rate of loss of good genes and so a corresponding rate of increase of bad ones. This is the eugenic problem that far outweighs all others. We are confronted with a relatively enormous dysgenic trend, one moreover which is readily amenable to selection, should selection be attempted either directly or, better and more practicable, indirectly through those things that affect our lives and are to greater or lesser degree within our control.

    There is no reason to suppose that the unfavourable differential birth‑rate in regard to intelligence exists in anything like the same measure in regard to anything else. Not only is it known solely with respect to intelligence, but is only likely to exist with respect to intelligence. Nevertheless we must not forget that good quali­ties tend in general to be positively correlated and similarly for bad qualities. This means that improvement in regard to the differential birth‑rate as it affects intelli­gence should automatically secure improvement in other directions also. But this in all probability would be small, important though it is to recognize that it would be in the right direction.

    If the average level of intelligence of the community continues to decline, we can expect to see the effect not only on the average but much more markedly at the two extremes. The average citizen will become less fitted by natural aptitudes to understand and exercise an effective voice in the society in which he lives. But the proportion of mental defectives will rise sharply. Most critically of all, the loss of good genes will be shown in a decrease in the number of the really talented. Civiliza­tion owes disproportionally much to men of the very highest intellectual powers, to such figures as Shaw and Einstein in the present age, Pasteur and Descartes in a past one. Even a relatively small fall in average intelligence implies a catastrophic fall in the number of such men. These are the men whose intellects place them in the one‑in‑a‑million class. With an average drop of half a dozen points on the Binet scale we can expect them to appear, not with a frequency of one in a million, but of less than one in ten million; for every ten such men today, tomorrow we should have but one.

    Eugenics in the past has had too much of an aristocratic air. Perhaps the brilliant pedigree of its founder has had too much fascination. Perhaps there has been a failure to distinguish between fruitful methods of investigation and the use that can be made of them. Studies of the extremes, on men of genius on the one hand, on mental defectives on the other, are both fascinating and illuminating, but the lessons that they teach apply to the whole community. Galton himself pointed out the significance of changes in the average for changes at the extremes, just as he pointed out that among the relatives of men of outstanding ability there are a remarkable number of the outstandingly able. These are two complementary aspects of the same phenomena, but we have tended to neglect the more important one. Galton's views about distinguished families were right enough; but we need to supplement them with the observation that, all the same, the great majority of intellectual giants come from families of moderate ability as the result of an exceptionally favourable chance. Taking them by and large, the children of university professors are no doubt more intelligent than the children of clerks. But there are so many more clerks in the world than university professors that any collection of brilliant men will include more clerks' sons than professors' sons.

    We need to look for a moment, not at the distribution of men of ability in the community, but at the distribution of the genes. No doubt the genes of hereditary ability are spread more thinly among those of average and slightly better than average intelligence than in the very able. But there are many more such genes in the former group than in the latter. Regarded as treasuries of potential ability, there is no comparison between the two. The principles of genetics inform us that our concern must be with the community as a whole, and not with a small and special section. The facts as we see them give rise to grave disquiet. Our task is to reverse the present trends, to change the sign of differential fertility, to encourage the more intelligent half of the community to have more, not fewer, children than the less intelligent.


     Inherited human differences tend to fall into two groups. First, there are definite departures from the normal, these being commonly due to the action of single genes. Secondly, there are the inherited differences that distinguish normal people, which tend strongly to show continuous variation. These are due to multifactorial inheritance. While there is some scope for eugenic action in differences of the first kind, this is small compared with the effect of even moderate degrees of selectioi, operating on all human beings, on differences of the second type. The outstanding eugenic problem, far outweighing all others in importance, is the present differential birth‑rate with respect to intelligence, for of this human quality only do we know that the less desirable genes are multiplying rapidly at the expense of the good.