Nonconcordant Variation

views updated

Nonconcordant Variation

Nonconcordant variation and discordant variation are the phrases that have historically been used to describe the commonly found noncorrelative and nonassociated nature of variation between pairs of genetically controlled traits. Said more positively, the phrases mean that traits tend to vary independently from each other.

A common example of a pair of nonconcordant traits is skin color and height. On both an individual and group level, skin color is independent of height (and height is reciprocally independent of skin color). As skin colors become lighter or darker, heights do not change in a predictable, dependent, or concordant way. Similarly, as heights increase or decrease, there is no predictable change in skin color. One trait is independent of the other.

Indeed, skin color appears to be independently distributed compared to all size traits and to virtually all phenotypic and genetic traits. The obvious exceptions are those few such as eye and hair color that are likely to share pigmentation genes with skin color. These exceptions can literally be seen with the eyes, and perhaps because they are so obvious, one might think that these are more common than they are. In fact, they are one of just a few known exceptions to the rule of trait independence.

The significance of this pattern of independent variation is that generally one cannot predict the distribution of one trait from that of another. Skin color does not reveal deeper (genetic) variation. More generally, from a biological perspective, racial traits such as skin color are just skin deep: they do not have deeper biological meaning.

HUMAN BIOLOGICAL VARIATION

The notion of discordant trait variation can be traced back to at least the 1950s. Evolutionary biologists observed that instead of varying together, geographic variation in one trait is usually discordant with geographic variation in other traits. One implication of this finding was a debate over the taxonomic classification “subspecies.” In an influential 1953 article, “The Subspecies Concept and Its Taxonomic Application,” zoologists Edward O. Wilson and William L. Brown Jr. observed that because so few traits are concordant (vary together), a focus on different traits produces different subspecies. They also observed that most traits used to define subspecies are arbitrary.

In 1964, anthropologist Ashley Montagu edited a groundbreaking volume titled The Concept of Race in which zoologists Paul Ehrlich and Richard Holm used the terms concordance and discordance to describe the distribution of human traits relative to each other. They noted that if traits are largely concordant, then the study of the variation in one trait will reveal the pattern in the other. However, if the opposite is true, and traits are discordant or nonconcordant, then population variation must be studied one trait at a time.

Along with anthropologists C. Loring Brace and Frank Livingstone in the same volume, Ehrlich and Holm noted that rather than abrupt ruptures or boundaries in traits, the mean probabilities of traits vary continuously or gradually over geographic space. This type of geographic variation is now commonly referred to as clinal or continuous variation. They further noted that these traits tend to vary in different ways from each other. Thus, for example, human skin color tends to vary along a north to south gradient, with darker skin, on average, near the equator and lighter skin, on average, toward the poles. However, the distribution of other traits tends to follow different and less predictable patterns.

Trait independence is also a rule of variation on an individual (as well as a group) level. Thus, if one knows that someone is tall, one cannot predict if that person is dark or light, handsome or ugly. Tall, dark, and handsome is just as unlikely as tall, light, and handsome or tall, light, and ugly.

A useful diagrammatic illustration of trait independence was also published by Ehrlich and Holm (Figure 1). This diagram of discordant variation in four traits or characteristics has been frequently reproduced and redrawn in many textbooks. Each layer in the cube represents geographic variation in a trait. The top trait, for example, might be skin color. In this case, the individuals to the right have light skin, whereas the individuals to the left have darker skin. Similarly, the next layer down represents another trait, such as eye color, the next still another trait, such as the percent of type A blood, and the bottom layer represents a fourth and last trait.

Each core represents a specific individual or the average of a group of individuals in a region. (Remember: trait independence works at both an individual and a group level.) If the top layer represents skin color, then the two cores (individuals, groups) to the right appear to be the same for skin color. In the other levels, it becomes clear that skin color does not predict for the variation in the other traits. For example, the two individuals who

have the same skin color differ with respect to all other traits. One might imagine millions of traits and any number of individual or group cores. Any one level/trait will predict very little about the other levels/traits.

Although the general point of nonconcordant traits is modeled very well in Figure 1, in reality traits are rarely either completely independent or completely concordant. Another way to think about trait independence is in relationship to statistical concepts of correlation and association. If two traits are perfectly positively correlated, then as one increases, the other increases in a proportionate fashion. Similarly, if two traits are perfectly inversely correlated, then as one increase, the other decreases in a proportionate fashion. This is rare, but some traits are highly correlated, such as length of legs and height or hair color and eye color. On the other hand, most traits are nearly totally independent of each other; that is, as one trait changes, the change in the other trait is for all intents and purposes random and unpredictable. Trait independence is both a general taxonomic rule and a hallmark of human variation.

Why, then, would traits vary independently? The answer depends on the specific traits under consideration. But, one can say that trait independence is probably related either to the forces of natural selection, if the trait is adaptive, or to random genetic factors.

In the case of skin color, variation appears driven by the relative adaptive advantages of dark skin pigmentation under conditions of maximal solar radiation and the advantages of lighter skin pigmentation under conditions of reduced solar exposure. The distribution of sickle cell anemia follows a pattern of variation that seems to be related to areas where malaria is greatest. The ability to digest lactose (milk sugar) in adulthood seems to be related to histories of dairying.

The majority of genetic and phenotypic traits appear to be adaptively neutral. Their distributions are also generally continuous across the globe. While the distribution of other traits seems more random and unpredictable, all of these traits are clinally distributed. Finally, the key point is that all of their clinal distributions are different. They are not concordant. They are independent.

NONCONCORDANCE AND RACE

The idea of biological races presupposes that traits are concordant; that is, the distribution in one trait tells us something or predicts for the distribution in another. In a sense, the idea of race is based on the notion that one can “read” a phenotype. For example, seeing someone who is overweight is often “read” in our society to mean someone who has an impulse-control problem. In this racist culture, seeing someone as dark skinned is “read” to conjure up a number of negative attribute associations. This is how the ideology of racism works.

As anthropologist Alan Goodman (1997) has noted, nonconcordance variation helps to make the case against meaningful biological races. For example, just as Wilson and Brown found that a subspecies could be redefined according to different traits, Jared Diamond (1994) created racial classifications based on the distribution of fingerprint patterns that were strikingly different from traditional models based on skin color.

Further, the overwhelming nature of nonconcordant variation tells us that what one sees does not predict for deeper characteristics. Genetic research suggests that traits such as skin color do not necessarily reflect shared evolutionary lineages. Nor are they useful for predicting behavior. Despite the social importance they may carry, racial traits are not useful markers for understanding human variation.

In describing the relationship between traits, nonconcordance captures part of the structure of human variation. That part is at complete odds with the idea of race.

BIBLIOGRAPHY

Brace, C. Loring. 1964. “A Nonracial Approach Towards the Understanding of Human Diversity.” In The Concept of Race, ed. Ashley Montagu, 103–152. New York: Free Press of Glencoe.

Diamond, Jared. 1994. “Race Without Color.” Discover 15 (11): 82–89.

Ehrlich, Paul, and Richard Holm 1964. “A Biological View of Race.” In The Concept of Race, ed. Ashley Montagu, 153–179. New York: Free Press of Glencoe.

Goodman, Alan H. 1997. “Bred in the Bone.” The Sciences. March–April: 20–25.

Livingstone, Frank B. 1964. “On the Nonexistence of Human Races.” In The Concept of Race, ed. Ashley Montagu, 46–60. New York: Free Press of Glencoe.

Park, Michael Alan. 2002. Biological Anthropology, 3rd ed. Boston: McGraw-Hill.

Wilson, Edward O., and William L. Brown Jr. 1953. “The Subpecies Concept and Its Taxonomic Application.” Systematic Zoology 2: 97–111.

Alan Goodman

Joseph Jones

Encyclopedia of Race and Racism