The Study of Man: New Light on the Races of Man
Caught between nationalist and fascist affirmers of the significance of “race” on the one hand, and vocal liberal deniers of any meaningfulness to the concept on the other, the ordinary man scarcely knows what to believe. However, recent advances in the science of physical anthropology seem to offer a solution to his dilemma, and incidentally promise to light up the early history of man with a clarity unavailable to us up to now.
Until a few years ago, a book on race was more likely to deal with politics than biology. Even those whose sole concern was the scientific study of the varieties of mankind were drawn into the battle against the political misuse, by chauvinist and fascist forces, of real or presumed racial differences. Today, the danger is far less serious than it has been for the last seventy-five years. Which leaves us, still, with the scientific problem itself: what does it mean to speak of the races of man?
Of late, substantial progress has been made toward answering the question. We now understand, in large measure, why previous approaches to the problem of race classification resulted in confusion; and we are coming to see the way to a comprehensive and consistent understanding of the physical variations in man. The “races of man” now takes on a new meaning. For it has been possible to establish that there are reliable differences among groups, differences in the degree to which they possess certain genes, those basic hereditary constituents that biologically determine man’s characteristics. These are not hard and fast or absolute differences; it is a matter, rather, of relative frequencies of a few genes. And fortunately for those of us who are understandably anxious about the possible political misuse of race, these differences, distinct as they are, and, as we shall see, useful as they are as clues to the unraveling of the history of man, have only the most infinitesimal bearing, if any, on human nature in society.
For over a century anthropologists sought to classify man by establishing a series of indices that in their totality would sharply distinguish one racial group from another. Armed with calipers, measuring boards, metric tapes, goniometers, and an abiding faith in quantitative coefficients, they reduced the form of the human body to a series of measurements: cephalic index, nasal index, bigonial diameter, bizygomatic diameter, sitting height, pelvic breadth, face height, scapular index, and so on—one anatomist, it is said, took ten thousand measurements of the human skull!
The man in the street was untroubled by the fact that, while eye form may be used to distinguish the Mongoloid group from the Negroid and Caucasoid, it does not distinguish between Negroid and Caucasoid; that dark- and light-skinned people are found in all the major racial groups; that the tallest and shortest people in the world both belong to the same major racial group—the Negroid. To the anthropologist, however, these were serious problems. And there was no way of dodging the simple mathematical fact that the more you multiplied criteria, the more types you had to define. Before long, texts on race were littered with “primary races,” “primary sub-races,” “secondary races” (“mixed”), “composite primary sub-races,” and “residual mixed types” (an academic euphemism for the human types that defy classification).
The aim of all this was of course to arrive at a clear-cut classification of races. Such a classification might be the first step in tracing the paths of biological descent responsible for the present physical differences in mankind. But classification proved far more difficult in the case of man than in almost any other species.
At the root of the confusion lay the now outmoded and erroneous “blending theory” of biological inheritance. This “common sense” view of heredity, developed by the British biometrician, Francis Galton, was for long the dominant theory of human hereditary mechanisms. It still survives today, unfortunately, in many texts and treatises on the physical anthropology of man. His theory assumes that the two parents together contribute on the average one-half of each of the faculties inherited by their offspring—each parent contributes one-quarter, the four grandparents contribute among them one-quarter (or one-sixteenth each), and so on. The hypothetical sum of the infinite series 1/2, 1/4, 1/8, 1/16. . . is equal to one human being.
This theory had good scientific warrant in statistical studies, which revealed that the correlation between parents and offspring for many traits was about one-half, and between offspring and grandparents about one-quarter. It was concluded, therefore, that genetic traits appeared to “blend” in much the same way that a mixture of black and white pigment produces gray. These assumptions and studies supported popular usage in identifying a person’s racial membership as “pure-blooded,” “full-blooded,” “half-breed,” “one-eighth Negro,” “one-quarter Indian,” etc. The blending theory implied the presence—today or at some time in the past—of pure races, like primary colors, that had blended.
Since Galton, we have made enormous progress in our understanding of the mechanism of heredity. Just as in physics the originally hypothetical atom has revolutionized our understanding of reality, and become for the physicist—and for the rest of us—the reality underlying ordinary appearance, so too, in biology, an originally hypothetical particle, the gene, has become the key to unlock the reality behind biological appearance. We now know a great deal about genes, enough to be sure that the blending theory is only a very rough approximation of what actually occurs.
As every schoolboy knows, Gregor Mendel, an Austrian monk, discovered in the middle of the 19th century, by way of experiments on peas, the mechanism of heredity in sexual species, plant or animal. Apparently both the male and the female possess an agent for each physical characteristic, which agent was later called a gene. Thus, whether a pea is tall or short will be determined by two genes, one inherited from the female stamen and one from the male pistil. Genes may be either dominant or recessive; thus, in the case of peas, tallness is dominant, so that a cross between a tall and short plant will produce a tall one. Sometimes neither gene is dominant, and a cross produces an intermediate form, which is apparently a blend.
Mendel’s work was forgotten for close to fifty years, and only rediscovered at the turn of the century. By now, while no one has yet seen a gene, we know they are arranged in linear fashion along objects in the nucleus of the cell; these objects are called chromosomes. Each species has a distinctive number of chromosomes—man has forty-eight. The chromosomes—and of course the genes within them—are arranged in pairs, as Mendel’s work suggested, so that two genes, one inherited from the male and one from the female, determine each hereditary trait (it is believed that man possesses between 20,000 and 40,000 genes). Egg and sperm cells, in distinction to all other cells contain only half the chromosomes (selected at random) of the normal cell, and the full number is then restored in sexual union, giving a human being an assortment of twenty-four chromosomes from his mother and twenty-four from his father.
This bare recital conceals some remarkable possibilities. For example, since the twenty-four chromosomes that go into an egg or sperm cell are selected at random, each pair of parental chromosomes giving up one, it is possible for a single individual to produce no less than 16,777,216 types of egg or sperm cells differing in their hereditary characteristics!
How does all this affect our understanding of human race? For one thing, we know that genes are specific in effect and do not blend or mix; that is: interbreeding does not cause the genes to lose their relatively separate and distinct identities and properties. A single gene of the fruit fly has been traced through three hundred generations; at the end of these vicissitudes, its effect was the same as at the beginning. We also know that genes are not transmitted to biological descendants in set packages, but rather in the form of an enormous number of different possible “assortments.” So that the statement that some individual is “one-quarter Indian” can only mean that one of his grandparents was socially defined as “Indian,” but not that he has inherited exactly one-quarter of that grandparent’s Indian genes. He may have inherited more or less of them: it would depend on what characteristics we consider to define a person as Indian. And sheer chance would have decided which of these genes the subject inherited; for it is statistically possible that not a single one of his chromosomes should have come from a particular grandparent! (The chance, admittedly small in the case of a 48chromosome animal like man, is one in 16,777,216.)1
This problem of the “one-quarter Indian” warrants further examination. On the average, we could expect to have twelve chromosomes from each grandparent; but the exact number would depend on how the twenty-four chromosomes of the parents’ egg and sperm cells were divided. When we come to a person who is one-sixteenth something, we should expect to have three chromosomes from each great-grandparent—but the chance of having none rises greatly. Indeed, as soon as we get back to a level at which there are more ancestors than chromosomes (this is six generations back) the odds turn against the possibility that one possesses genes from any individual ancestor. Those persons who pride themselves on a single Mayflower ancestor some ten generations back have little chance of possessing anything biologically inherited from him. The genes of such a person are present in the population but so dispersed that none of his known descendants can claim with any high probability even one of his forty-eight chromosomes. (There is one exception: a male in the direct male line of descent possesses the chromosome for maleness—this chromosome however seems to be particularly deficient in other genetic material.)
With increasing knowledge about genes, almost all the criteria traditionally used for the classification of races has been thrown into question. Such characteristics as skin color, hair color and texture, shape of skull, stature, and other favorites, it became clear, are dependent on very complicated gene mechanisms. They had been selected in the first place because it was believed they were unchanging and stable from generation to generation—that they were, in other words, “non-adaptive,” and did not vary according to different environmental conditions. This is an important criterion for the classification of racial traits. For if some physical trait is “adaptive” and hence gives the organism a special advantage in survival in a specific environment, then this adaptive feature will soon spread through a population, and it will be impossible to tell whether two populations, one of which has this trait and the other not, are different in biological origin, or have simply experienced recent rapid changes. To take a simple example: if the leopard does not change his spots, we can always tell a leopard, no matter where he is. Suppose however some leopards wander into an environment in which they are very easily seen by their enemies because of their spots. We might then expect, if a few leopards are born by accident (mutations) without spots, to find the rise of a new group of leopards without spots. Now if we are interested in the racial history of the leopard, and believe the spots are non-adaptive, we may well mistakenly consider two closely related populations of spotted and unspotted leopards as being distinct.
So, then, the classic indexes of physical anthropology were assumed to be non-adaptive. Skulls in particular were highly favored—perhaps because their very hardness suggested their unchangingness, but, more important, because we could collect skulls from centuries back and reconstruct racial history. But skulls, the skeleton in general, hair color and texture, have all tumbled as useful classifying devices. There is the famous case of the head shapes of children of American immigrants which the late Franz Boas showed were quite different from those of their parents. These findings have been challenged; but another Columbia University anthropologist, H. L. Shapiro, has shown definitely that similar changes have occurred in the head shapes of children of Japanese immigrants to Hawaii, and it is now well established that the shapes of the head, which once seemed to be the key to racial history, can be affected in a single generation by environmental changes. Besides, there has been a long-range change in Europe from longheads to roundheads. This has been interpreted as a creeping insidious invasion of the roundheads by some anthropologists—but it is likely it is a response to some evolutionary advantage conferred by the roundhead, and tells us nothing about the racial history of an assumed round-headed race.
Skin color would not seem to be subject to this stricture. Negroes transported to America have probably lightened only because of mixture with whites. There are certain facts—for example, the fact that the original homes of dark-skinned races are in regions where there is the most and strongest sunlight which suggest the operation of an environmental factor of selection over a very long period of time. Yet skin color has stood up rather better than skull form as a criterion for race under modern criticism.
But aside from the question of whether such traits as skull shape and color are adaptive is the fact that their genetic constituents are actually quite complicated and as yet unraveled. This makes them hazardous as criteria for the setting up of races—if, that is, our concern with races is not simple classification but classification based on heredity. Without knowing their specific genetic mechanism, for example, we have little or no idea whether we should expect high, medium, or low rates of mutation.
Mutation—together with environmental selection—is one of the necessary conditions for biological variation and evolution. Under normal conditions, the gene is passed on through reproductive processes and duplicates itself in each generation. But occasionally this self-duplicating process fails; the failure of a gene to reproduce its facsimile is called a mutation. In general, mutation is a random process and takes place independently of the “needs” or “desires” of the organism. The rate of mutation varies both with the organism and with particular genes within the organism. By experimental and mathematical analysis it has been possible to calculate accurate mutation rates for a small number of genes in man. (J. B. S. Haldane, in 1935, found that the rate of mutation at which the gene for the more severe type of hemophilia appears in the London population is about 1 in 50,000 individuals—or one mutation in 100,000 genes—per generation.) Without knowing the rates of mutation for the genes determining the criteria used in race classification (skin color, hair texture, etc.), we cannot ascertain what kind of difference in a population is “normal” and what kind indicates the presence of two or more biologically differentiated groups.
The principal weaknesses of the classical approach to race classification can be summarized as follows: (1) an uncritical commitment to measurement for measurement’s sake; (2) the unfortunate tendency to change the criteria of racial membership as one goes from group to group, thereby compounding confusion; (3) neglect of the inevitable fact that any increase in the number of different criteria used to determine racial membership will automatically increase the number of races discovered by the investigator; (4) the acceptance of the “blending theory” of inheritance, which set the classical investigator off on the search for “pure” and “ideal” racial types; and (5) a general unawareness of the genetic complexities underlying the transmission of conventional criteria of racial membership, e.g., cephalic index, skin color, hair and eye form, stature, and the like.
These criticisms reflect the gradual shift in biological analysis from what we may call taxonomic-descriptive studies to studies of the hereditary process and the genetic determinants of biological variation. It became clear that men could be divided into biological classes according to the presence, absence, or frequency of certain genes.
Professor William C. Boyd has recently written an important work that is representative of the new approach to the problem in the field of physical anthropology, Genetics and the Races of Man: An Introduction to Modern Physical Anthropology (Little, Brown, 1950). Professor Boyd stresses the central idea that every human being is a member of a biological community, and that community constitutes a genetic reservoir or pool from which genes are drawn by mating and reproduction, to be returned to it in the form of offspring. As a result of various “isolating” mechanisms—the social rejection of certain possible mates or geographical isolation—groups develop within the biological community that differ from each other in the frequency of certain genes. According to Boyd, such groups may be called races: “A human race is a population which differs significantly from other populations in regard to the frequency of one or more of the genes it possesses. It is an arbitrary matter which, and how many, gene loci we choose to consider as a significant [racial] ‘constellation.’”
How arbitrary it actually is we may see from an extreme example: Suppose we consider a group composed of three brothers and one sister: Tom, Dick, Harry, and Martha. Tom and Martha have blue eyes, dark hair, and belong to blood group A, blood type M, with Rhesus factor Rh plus. Dick and Harry have brown eyes, light hair, and belong to blood group O, blood type N, with Rhesus factor Rh minus. Now consider a chimpanzee with brown eyes, blood group A, blood type M, and Rhesus factor Rh plus. If we take into consideration only the blood and pigmentation factors named above, it follows that Tom and Martha (homo sapiens) are more closely related biologically to the chimpanzee (pan) than to their brothers Dick and Harry.
Thus if we try to assign a single individual to a race on the basis of any fixed feature then we must run into the absurdity of having members of the same family fall into two different races. This absurdity arises whether we use skull measurements, skin color, or certain specific genes to define race—it is only overcome when we accept the fact that the human genetic pool has been so thoroughly stirred that we cannot find single genes that will characterize all members of a given population, no matter how isolated, and when we define race, as Boyd does, as a population differing in the frequency of the presence of a certain gene.
But what characteristics are then to be specified in genes if we are to use them to define “race”? The genes defining a genetic race must have a high or complete degree of “penetrance”: that is, a dominant gene must always produce a specific physical characteristic or effect, and no other. Thus the genes that are expressed in skin color, if usable as criteria, have to express themselves in human beings in all parts of the world and in all climatic conditions. In other words, the genes in question should always be present when the physical characteristic they determine is present, and absent when those characteristics are absent. They should not be influenced in any way by the growth of the individual organism; otherwise, we might have a gene that defines a race present at one age and absent at another. The gene considered should also have “constant expressivity,” that is, there should be no variation in the effect, or characteristic, produced by the gene in different persons. The gene should also be found in both males and females. And the genetic characteristic must be identifiable in more than one population group so that comparisons can be made between different groups.
A number of human genes have been discovered that meet these specifications, and suggest the possibility of an unambiguous classification of race, clearly established on the basis of heredity mechanisms. The best known of these genes are those that manifest themselves in the blood groups A, B, AB, and O, and in the blood types M, N, and MN, the eight Rhesus blood types, and in taste reactions to a chemical substance called phenylthiocarbamide (PTC).
How this mysterious entity, a single gene manifestation, is discovered, is a fascinating subject in itself. In his book Professor Boyd tells the interesting story of how the gene for sensitivity to PTC was discovered. PTC is a chemical substance that most people (about a little under three out of four) declare has an unpleasant or extremely bitter taste, while the remainder claim that it is tasteless. Quite by accident, a debate arose in a chemical laboratory over this difference in reaction to PTC, one person claiming he could taste it, another that it had no taste at all. Subsequently, by testing people in the same biological family, it was discovered that the capacity to taste or not taste this substance is inherited as a rule (though not always) along simple dominant and recessive lines. “Certain groups of people,” Professor Boyd reports, “such as the American Indians, seem to consist of nearly 100 per cent tasters, while in other ethnic groups, as in Wales, large numbers of non-tasters, amounting in some cases to more than 50 per cent of the total, may be found.”
In the same way, it was possible to isolate the specific genes that determine the blood groups and types. And these, too, show great variation from one human group to another. Let us consider one of the simplest of the genetically determined blood constituents—the blood types A, B, AB, and o, which were discovered by Landsteiner in 1900-1902. People with A type blood react to anti-A serum; with B type to anti-B serum; with AB type to both; with O type to neither. There are three genes operative to determine blood type; A, B, and O, and both A and B are dominant over O. We can determine the proportion of each gene in various populations, and plot the results on a map. For example, the distribution of B shows a very interesting pattern. The highest proportion is to be found in Central Asia—Mongolia, Sinkiang, Russian Turkestan, Persia, and Northern India. The proportion falls off as we move out from this center, falling to small proportions in the British Isles, Scandinavia, and Spain; and it is not to be found at all in the Americas and most of Australia. Blood type A shows a converse distribution, low in the Asiatic center and high in peripheral regions. The consideration of these and other distributions suggests the specific history by which at some point in the past, human beings had become differentiated into groups which we may (if we wish) call races.
On the basis of gene frequencies for blood groups and types, Professor Boyd suggests tentatively the following early racial classification: (1) Hypothetical Early American; (2) European (Caucasoid); (3) African or Negroid; (4) Asiatic or Mongoloid; (5) American Indian; (6) Australoid.
Professor Boyd points out what the reader may have surmised, that these races do not differ in any startling way from the older classifications based on skin color, hair form, eye form, cephalic index, lip structure, and so on. He argues, however, that unlike traditional investigators, he began without any concern for culture or geography, and at no time found it necessary to change his criteria from group to group. And yet, applying these uniformly, he arrived at a racial distribution that conforms with what we know at present of the history of mankind. Thus, marked genetic differences divide the natives of Greenland and the aboriginal Americans from the Western Europeans, while it is impossible to draw a sharp geographical dividing line in Eastern Europe between the European (Caucasoid) and the Asiatic (Mongoloid) races. Furthermore, it is demonstrated that the varieties of human beings can be classified by genes more simply and without the many inconsistencies of method and interpretation that encumbered the older methods. And last but not least, we can be confident we know something of the real biological connections between one group and another.
Professor Boyd’s approach opens new vistas in the study of race. For example, the calculation of gene frequencies “throws some light on the probable origin (evolutionary) of the racial differences which we observe at present, and allows us to attempt a historical reconstruction of gene distributions which may have existed in the past. . . .” Newly devised techniques for determining blood groups can, in some instances, be applied to bones and tissues (some work has been done on Egyptian mummies that are over three thousand years old) and we may yet solve some of the mysteries of racial history.
It must be added, however, that no complete theory of racial history on the basis of gene frequencies has yet been developed. Boyd admits, for example, that his scheme of six genetic races makes no provision for the Pacific peoples because they “do not agree in exhibiting any distinctive combination of frequencies of the genes which we have thus far identified.” They show affinities with the Australians, the Asiatics, and the Early Americans—in this last, offering some incidental support for the hypothesis of the author of Kon-Tiki. However, the discovery of new genes may yet demonstrate a sharp genetic frontier that would make it unlikely that the Polynesians are related to the American Indians. A discovery such as this is likely to be one of the most interesting results of the genetic approach to race.
The notion of a “genetic frontier” demands a few more words. As a matter of fact, we more often find genetic gradients than genetic frontiers—thus, as we advance from region to region we find the proportion of some gene slowly rising or declining. We must arbitrarily decide at which point we set up a frontier and consider that here one race ends and another begins. Indeed, since genes are quite independent of one another, and have different distributions, quite anomalous situations develop. The “Asiatic” or “Mongoloid” blood type can be found far West in Europe—the equally Asiatic or Mongoloid eye-fold (whose genetic mechanism is still unknown) has a more restricted distribution; besides all this, there is no tendency for a person with Asiatic blood to have an Asiatic eye fold—the genes determining either, once introduced into a population by the same group, will have quite distinct histories, for one child may inherit a gene for one character that his sibling will not. The Indians, on the basis of blood type, form part of the Asiatic race, which may disturb those who have assumed they are the parent Aryans. However, the discovery of new genes may show them linked to European groups, too.2
Another possibility opened up by genetic analysis is that of determining the actual degree of biological relatedness between groups of Jews throughout the world. One researcher, David C. Rife of the Institute of Genetics at Ohio State University, has discovered that Jewish students at Ohio State University have a different frequency of blood group A from Protestant students—but show the same frequency as Jews in Odessa. Other Jewish groups, on the other hand, show very different frequencies of the various blood types from those typical for Jews in Russia and America. Further work along these lines will permit us to determine more accurately than before the biological relationship between different groups of Jews.3
Although Professor Boyd uses blood groups and types to illustrate his argument, he is quick to point out that blood-group analysis alone will not necessarily offer any unique advantage in racial classification. The advantage,
as such, lies in the fact that we know “the way in which they [blood groups] are inherited, and we possess the reagents which detect the results of the action of nearly every individual gene involved.” In time, it is certain that more genes determining human hereditary characteristics will be isolated.
He suggests as one of the advantages of his scheme of racial classification that “in no part of the world does the possession of a blood group A gene, or even an Rh negative gene, exclude a person from the best of society.” This may be so, but the history of doctrines of racial and biological superiority gives us little grounds for security on this point. Physical characteristics in themselves play a neutral part in racism; one trait may well be substituted for another, and the prejudices remain.
Indeed, it is fortunate that political considerations play so small a part in Professor Boyd’s discussion. Man’s cultural achievement, man’s beliefs and values, obviously have nothing to do with his race—or rather, we know nothing about the hereditary transmission of genes in human groups as yet that leads us to think there is any connection. The effort to make race and culture coincide has been disastrous, both to the science of physical anthropology and to human welfare. It is to be hoped that in the future the problem of the transmission of genes, and of the relations between different human groups—a problem that has its own intrinsic interest—will remain as carefully delimited as it is now, and not become involved with the task of explaining human character and culture.
1 A full discussion of all the complications and fascinations in human heredity can be found in The New You and Heredity, by Amram Scheinfeld, New York, Lippincott, 1950.
2 Sizeable differences in the frequencies of different blood types have also been shown to exist in two castes in Bombay. Again this is valuable evidence on the question of the origin of castes through racial conquest.
3 For a full account of the work along this line to date, see J. D. Brutzkus, “The Anthropology of the Jewish People,” in The Jewish People: Past and Present, Volume I, New York, 1946.