Two Questions About Race
Published on: Jun 07, 2006

Alan Goodman is professor of biological anthropology at Hampshire College and co-editor of Genetic Nature/Culture: Anthropology and Science Beyond the Cultural Divide and Building a New Biocultural Synthesis: Political-Economic Perspectives on Human Biology. He is president-elect of the American Anthropological Association.

The billion or so of the world's people of largely European descent have a set of genetic variants in common that are collectively rare in everyone else; they are a race. At a smaller scale, three million Basques do as well; so they are a race as well. Race is merely a shorthand that enables us to speak sensibly, though with no great precision, about genetic rather than cultural or political differences.—Armand Leroi, The New York Times 3/14/05

Instead of obsessing about race, we could try to build a race-blind society. Instead of feeding the fires of neuroticism, we could start teaching people to forget about race, to move on. But to do that, first we must sideline the entire race relations industry - whose only function, it seems, is to make us all deeply anxious about 'race' - a concept they simultaneously believe has no objective reality.—Sean Thomas, Sunday Telegraph (London) 3/13/05

While Armand Leroi’s (2005) editorial in The New York Times proposes that race is everywhere as a shorthand for genetic variation, Sean Thomas’ (2005) essay in the Sunday Telegraph, published just a few hours earlier, advocates for forgetting about race, “to move on.”1 Leroi’s concept "with no great precision" harks back to a 19th century understanding of race as all-explaining and biologically based. Conversely, Thomas asserts that race has “no objective reality.” The problem is not so much that Leroi and Thomas have nearly diametrically opposed views on race. The problem is that they are both wrong.

In this essay, I propose that debates about race often ride upon two questions: “Is race a useful categorization to describe human biological variation?” and “Is race a useful categorization for tracking sociopolitical injustices?” Contra Leroi, I argue that race is no longer the right way to describe biological variation. Contra Thomas, I argue that race is not a mere social construct, but as a lived experience has devastatingly real effects.

As in the case of Leroi, many scientists in the US would probably answer the first question in the affirmative. In doing so, they conflate the idea of race with the reality of human geographic variation. Since Lewontin (1972), it has been clear that race fails to explain the vast majority of human genetic diversity. Moreover, processes such as evolution and cultural history better explain what is statistically left over to race. Armand Leroi’s conceptual error, one of racialization of diversity, has the potential to do harm, especially when coupled with a strong belief in the power of genetics. Moreover, with better explanations available, it is unnecessary to hammer away at questions regarding genetic diversity using the same blunt and dull instrument of race.

Thomas (2005) does not directly disagree with Leroi (2005) because he is presumably focused on the social reality of race, the second question. In suggesting that we ought to forget about race, Thomas makes two entirely different errors. First, he treats race as a biosocial unity that cannot be broken apart. Thus, if race is not real biologically, it must not be real at all. Second, he ignores the real consequences of racialization and racism.

Leroi’s Error: Conflating Race and Biodiversity

Although the word is somewhat unfashionable, and may even be considered politically incorrect, race is a good short word…—Teresa Overfield, Biological Variation in Health and Illness, 1995:1.

Nearly every Western natural historian/physician/scientist once took for granted that the idea of race was the same as human biological variation. The idea that humans were divisible into racial types, a European folk idea, was assumed to be scientifically right, and very few scientists or nonscientists questioned the assumption. Through constant use, the idea of race was fixed and reified as human diversity.

In the middle of the 20th century, a few scientists realized that race was not the same as human biological variation. In Man’s Most Dangerous Myth: The Fallacy of Race, Montagu (1942) considered race to be a myth because race was a typological rather than an evolutionary concept. Human variation was unstable and races could not be reliably defined. Two decades later, Brace (1964) proposed a nonracial approach to human variation that emphasized continuous or clinal variation and Livingstone (1962) wrote that there are no races, only clines. Ehrich and Holm (1964) clearly pointed out that traits tended to be nonconcordant. Thus, for example, skin color would not predict other “deeper” human characteristics. Race was only skin deep.

Until the 1970s, one could be excused from thinking race both categorized and explained human biological variation because there was not very much data to think beyond the existing, powerful and thoroughly reified racial worldview (Smedley, 1999). Since Lewontin (1972), however, an accumulation of data on human genetic variation has made it easier to think beyond race. Nonetheless, as a tribute to the power of the idea of race, many such as Leroi and Overfield still believe that race is a useful way to characterize human variation. In a nursing textbook Overfield writes on the very first page “race is a good short word.” Reliability and validity, upon which race scores very poorly, are critical criteria for a scientific concept. Conversely, I am fairly certain that word length is not.

True Believers and De Facto Believers

There is a spectrum of views among contemporary scientists like Leroi who still assert race is a useful classification for human biological variation. On one end stand scientific “true believers” who treat races as natural entities. The primary difference between this position and racial typologists of the early 19th century is that 21st century scientists assert that evolution, rather than god, created “races.” These racial “true believers” include evolutionary anthropologists Vince Sarich (Sarich and Miele, 2004) and psychologist J. P. Rushton (1995), now the president of the Pioneer Fund. Rushton is infamous for promoting the idea that races evolved either brains (meaning big heads), or brawn (meaning large muscles, and sex organs).

On the other end of the spectrum are the “de facto believers,” an otherwise respectable group of scientists that encompasses Leroi. What separates this group from the true believers is that they understand races as statistical approximations rather than natural types, asserting race as a de facto stand-in for the messy patterns of human biological variation. Sally Satel (2002), for example, the author of a prior New York Times piece titled “I am a Racially Profiling Doctor,” concurs that humans do not vary much genetically and race is a crude approximation of this human variation. She goes on to say that race might not be necessary in a near future of individualized genetic analysis. According to the “de facto believers” of race, we are merely passing through an awkward adolescent phase in which we still need to racialize human variation. While they grasp some of the limits of race, they neither grasp the potential harm nor the lack of necessity to racialize human variation.

Is race a reasonably useful shorthand for human biological variation? A first test of this question would examine whether race statistically explains or correlates with a significant portion of human genetic variation. A second test considers whether the idea of race explains the process by which variation comes about.

Test 1: Is Human Variation Racial?

In a classic article published in 1972 Richard Lewontin estimated the proportion of human variation that could be statistically explained by races. If the relative degree of variation among races is great compared to the variation within a single race, then one might suggest that subspecies/races are statistically real. On the other hand, if the relative degree of variation among races is small compared to the variation within a race, then races are less statistically real.

Lewontin took data on blood group polymorphisms (those that have two or more alleles in high frequency such as blood types A, B, AB and 0) and tested how much variation was explained at three levels: within local groups, within a race but among local groups, and among races. He found that on average 85.4% of variation was explained at the local level and only 6.3% among races. Since Lewontin, a series of papers using larger and larger data sets have replicated his results (see review of Brown and Armelagos, 2001) demonstrating once again that “race” does not statistically explain much. Humans fail the test for biological races (Templeton, 1998).

It may seem surprising that a species with such a wide geographic range would display so little variation among races or continental groups. However, the apportionment of variation makes sense when one considers the history of our species and in particular its youth, steady mobility, and constant openness to ideas and other peoples.

Test 2: Explaining Human Variation

Statistically explaining “a little bit” about something may actually end up doing more harm than good if one begins to forget the "lack of precision" of the concept. This is the first problem when one substitutes race for human variation: one tends to forget about the 94% of variation that race fails to statistically explain. The test I now put to race-as-genetics is not statistically, but conceptually. Is race merely a poor correlate of human genetic variation or does it help to explain the underlying processes by which variation comes about? Consider the following.

  • Racial definitions and boundaries change over time and place. Thus, race is an inherently unstable and unreliable concept. That is fine for local realities but not so for a scientific concept. The importance of this point is that a bio-racial generalization that appears true at one time and place is not necessarily as true in another time and place. We just don’t know. One of the first lessons of science is to not base a generalization on a shifting concept, which is exactly what race is.

  • The idea of race can only divide human diversity into a small number of divisions. That is the limit. This might have been all that one could do before the advent of parametric statistics, multivariable analyses, and computers. But, now we can do so much more.

  • Because race is used in medicine and other fields as a way to categorize both genetics and lived experience, what passes as the result of genetic difference may actually be due to interactions or some aspect of lived experience. Using race tends to conflate genetics and lived experience (Goodman, 2001).

  • I am pessimistic about how the subtle reuses of race in genetics will eventually merge with virulent racists. This does not mean that I want to hide anything about human variation. Rather, it means that we need to study human variation precisely.

I advocate for de-racializing biological variation simply because there is always a more precise and meaningful way to characterize and explain those myriad variations.

Location, Location, Location

In the real estate industry there is a general rule that three things primarily determine housing prices: location, location and location. A similar refrain applies in the case of human genetic variation. Geographic location is the best single explanation for human genetic variation. There is no more powerful piece of information for predicting the genetic makeup of either an individual or a group than knowing from where on the map they originate. Furthermore, the degree of genetic variation between any two human groups is almost entirely explained by the geographic distance between them: Genetic and geographic distances are almost perfectly correlated (Templeton, 1998).

Although highly correlated with genetic variation, geographic location, however, is not in itself an explanation for genetic variation. Complex questions about human variation come down to specifics about our early evolution and migration out of Africa, subsequent movements of migrating populations, adaptive struggles, and stochastic events. To begin to put together these puzzle pieces, requires multiple lines of evidence and inquiry. Human diversity is the end result of two complex, interrelated and fascinating processes: evolution and history.

For example, one might ask, “Why do some individuals have sickle cell trait? Is it because of their race?” The answer to this question is clearly “no.” Race is a poor explanation for the distribution of sickle cell trait, which occurs in high frequencies only in particular regions of Africa while also occurring in high frequencies in parts of Asia and Europe. Rather, sickle cell trait can be understood as a fascinating history involving agricultural intensification, clearing of lands, breeding grounds for mosquitoes, and so on (Livingstone, 1958).

Sickle cell is but one example of how evolution and cultural history explain not only the distribution of particular traits, but how particular traits come about. This is one specific example of the profoundly biocultural processes of evolution and history. I want to propose that if we think race is an explanation or even if we use it as a statistical proxy, we are less likely to conceptually understand how variation arises and is distributed.

Thomas’s Error: The Color Blind Bind

Americans and much of the world’s population have been conditioned to think of race as a fuzzy jumble of behavior, culture, and biology: a deep and primordial mix of a bit of culture and a lot of nature. Thus, to say that race is not real in one way (as a shorthand for human biological variation) and is real in another way (as a way to group and track lived experience) is indeed confusing. Isn’t race simply real or not? If Leroi is wrong, isn’t Thomas right?

The idea that race is a social construct derives in part from natural scientists like Lewontin who maintain that race is a myth, or more precisely that the concept does not capture human diversity. It also derives in part from a misunderstanding of the notion of historical or social construction. Even though race was invented and made to seem real by social humans, and even though race makes little sense on the genetic level, this does not mean that it is not real in other ways. Thomas makes the mistake of thinking that because race is a social construct, race cannot have real effects. To the contrary, processes of racing, racializing, and practicing racism have enormous and powerful consequence for human wealth and health.

Race Is and Race Isn’t

There are different potential paths to racial justice. Some of my colleagues on the left have taken the path of seeing race as both genetically and socio-politically real (see Mosley, 1997 for example). This strategy involves trying to erase the negative biological association of the concept while providing equal opportunity. This message, that race is real, has the advantage of simplicity, but it may not work in the long run.

My own position is basically the same as that of the American Anthropological Association ( and is only slightly more complex. This position reaffirms the salience of race as lived experience while calling for a new vocabulary and concepts to study human biological variation. Like Mosley’s position, this position is grounded in a commitment to social justice. In contrast to Mosley’s position, it is scientifically correct.

In summary, there is no good scientific reason beyond word length, convenience, and maintenance of the status quo (laziness in short), to continue to racialize human variation. Moreover, doing so may cause harm. In this way, using “race” as shorthand for biological variation is a form of ideological iatrogenesis. Real human suffering may result from poor conceptualization of human variation. Yet, race is real as lived experience.

It is time, at least, to ask the right question. This question is not whether race is real, but in what ways do we make it a reality?


Brace, C. L. (1964). “A Nonracial Approach Towards the Understanding of Human Diversity.” In A. Montagu (Ed.) The Concept of Race. Collier Books, London, pp. 103-152.

Brown, R. and Armelagos, G. J. (2001). “Apportionment of Racial Diversity: A Review.” Evolutionary Anthropology 10: 34-40.

Ehrlich, P. R. and Holm, R.W. (1964). “A Biological View of Race.” In A. Montagu (Ed.) The Concept of Race. Collier Books, London, pp. 153-179.

Leroi, A. M. (2005) “A family tree in every gene.” The New York Times, March 14.

Lewontin, R. C. (1972). “The Apportionment of Human Diversity.” Evolutionary Biology 6: 381-398.

Livingstone, F. (1958). “Anthropological implications of sickle cell gene distribution in west Africa.” American Anthropologist 30: 533-562.

Livingstone, F. (1962). “On the nonexistence of human races.” Current Anthropology 3: 279-281.

Montagu, M. F. A. (1942). Man’s Most Dangerous Myth: The Fallacy of Race. Columbia University Press, New York.

Mosley, A. (1997). “Are Racial Categories Racist?” Research in African Literatures. 28.4: 101-111.

Overfield, T. (1995). Biological Variation in Health and Illness: Race, Age and Sex Differences, 2nd edition. CRC Press, New York.

Rushton, J. P. (1995). Race, Evolution and Behavior. Transaction, New Brunswick, N.J.

Sarich, V. and Miele, F. (2004). Race: The Reality of Human Differences. Westview Press, Boulder, Co.

Satel, S. (2002). “I Am a Racially Profiling Doctor.” The New York Times. May 5.

Smedley, A. (1999). Race in North America: origin and evolution of a world view, 2nd edition. Westview Press, Boulder, Co.

Templeton, A. (1998). “Human races: A genetic and evolutionary perspective.” American Anthropologist 1998; 100: 632-650.


1 The quotes above were culled from essays that appeared hours apart in major newspapers. The conflict in views and the frequency of articles on race is not unusual. Newspaper articles on race are literally published every day: Since 1996, in about 9.25 years, The New York Times published 15,976 items yielding a hit on the search term “race.” Many of these, to be sure, are about election, sports, and other types of “races” rather than about races as are here intended. Yet, the term racism, which would eliminate many articles of interest, yielded 5033 items, an average of over 1 (1.35) per day.