Terms of Estrangement

[Article 3 from Discover's 1994 special issue on the science of race. See my initial post explaining why I'm posting this and other articles from the issue.]
Race is a small but volatile word. It lacks a clear definition or scientific purpose. Yet it persists, not only in the lingo of the streets but in the language of the laboratory.

By James Shreeve

One November morning in 1984, Norm Sauer, a forensic anthropologist at Michigan State University, received a call from the state police. Somebody had found a body in the woods. The decomposed corpse displayed the typical mute profile of an unknown homicide victim: no clothing, no personal possessions at the scene, not even enough soft tissue left to readily identify its sex. The police only knew the body was human. They asked Sauer if he could recover the person's disintegrated identity -- turn the "it" back into a he or a she.

Sauer got into his car and drove up to the hospital where the body was being kept. He examined the form and structure of the skeleton, concentrating on the skull and pelvis, then took a number of measurements with his calipers -- the distance between the eye orbits, the length and width of the skull, for example -- and plugged them into standard forensic equations. Within a few hours he was able to inform the police that the skeleton was that of a black female who had stood between 5 foot 2 and 5 foot 6 and was 18 to 23 years old at the time of her death. She had been dead somewhere between six weeks and six months. With that information in hand, the police were able to narrow their search through the files of missing persons to a handful of cases. Some unusual dental restorations completed the puzzle: the skeleton belonged to a woman who had lived two counties away and had been missing for three months. She'd been 5 foot 3,19 years old, and black.

Age, sex, stature, and race are the cardinal points of a preliminary forensic report, the cornerstones that support the reconstruction of a specific human identity. Three out of four of these characteristics are firmly anchored in empirical fact. A person's sex, age, and height at any given moment are discrete quantities, not matters to be interpreted, revised, or dissected into their constituent parts. Whether I am 6 foot 1 or 5 foot 3 does not depend on who is holding the ruler. If I am male in Milwaukee, I remain male in Mobile. My age, like it or not, is 43; no amount of investigation into my personal history will reveal that I am mostly 43, with some 64 mixed in, and just a trace of 19 from mv mother's side.

But the fourth cornerstone – race -- is mired in a biological, cultural, and semantical swamp. In the United States, most people who are considered to be black trace their ancestry to West Africa; biologically speaking, though, some 20 to 30 percent of the average African American's genetic material was contributed by ancestors who were either European or American Indian. Different jurisdictions, government bureaucracies, and social institutions tend to classify race in different ways -- as do different individuals. Most Americans get to decide which race box to check on a form, and their decision may depend on whether they are filling out a financial-aid application or a country club membership form. A recent study found that in the early 1970s, 34 percent of the people participating in a census survey in two consecutive years changed racial groups from one year to the next.

The classifications themselves are eminently changeable: the Office of Man- agement and Budget, which is responsible for overseeing the collection of statistics for the federal government, recently held public hearings and is currently reading written commentary on the categories used by the Census Bu- reau. In addition to the racial categories now in place -- white, black, American Indian, Eskimo, Aleut, Asian or Pacific Islander, and "other" -- the OMB is considering adding slots for native Hawaiians, Middle Easterners, and people who consider themselves multiracial. If such categories are added, they should be in place for the census in the year 2000.

"Race is supposed to be a strictly biological category, equivalent to an animal subspecies," says anthropologist Jonathan Marks of Yale. "The problem is that humans also use it as a cultural category, and it is difficult, if not impossible, to separate those two things from each other."

SO WHAT IS RACE?

How important is it? Is it a notion rooted in our culture, or a reality living in our genes? Should the word be abandoned by scientists, or would banishing it simply cripple any attempt to help the public understand the true nature of human diversity, forcing us to seek our definitions on the street, in the jaundiced folklore of prejudice?

Everyone agrees that all human beings are members of a single biological species, Homo sapiens. Since we are all one species, by definition we are all capable of interbreeding with all other humans of the opposite sex to produce fertile offspring. In practice, however, people do not mate randomly; they normally choose their partners from within the social group or population immediately at hand and have been doing so for hundreds of generations. As a result, the physical expressions of the genes inherited from an expanding chain of parents and grandparents -- most of whom lived in the same region as one another -- also tend to cluster, so that there can be a great deal of variation from one geographic region to another in skin color, hair form, facial morphology, body proportions, and a host of less immediately obvious traits. Roughly speaking, then, race is the part of one person's variation on the theme of humanity created by the interplay of geography and inheritance.

The problem with this definition rests in the way patterns of human variation have traditionally been packaged and perceived. In the past, most anthropologists unquestioningly accepted the concept of races as fixed entities or types, each of which was pure and distinct. These types were seen as gigantic genetic bushel baskets into which people could be sorted. Admittedly, the rims of the bushel baskets might not be stiff enough to keep some of their contents from spilling out and mingling with geographically adjacent baskets. In the sixteenth century, European colonialism began flicking genes from one basket into other parts of the world; soon afterward the forced importation of large numbers of Africans into the Americas had a similar effect. But until recent decades, anthropologists believed that no amount of interracial mixing could ever dilute the purity of the racial ideals themselves.

In the bushel-basket scheme, races are defined by sets of physical characteristics that cluster together with some degree of predictability in particular geographic regions. Asians, for instance, are typically supposed to have "yellow" skin, wide, flat cheekbones, epicanthic folds (those little webs of skin over the comers of the eyes), straight black hair, sparse body hair, and "shovel-shaped" incisor teeth, to name just a few such distinctive traits. And sure enough, if you were to walk down a street in Beijing, stopping every once in a while to peer into people's mouths, you would find a high frequency of these features.

But try the same test in Manila, Tehran, or Irkutsk -- all cities in Asia -- and your Asian bushel basket begins to fall apart. When we think of an "Asian race," we in fact have in mind people from only one limited part of that vast continent. You could, of course, replace that worn-out, overloaded bushel basket with a selection of smaller baskets, each representing a more localized region and its population. A quick scan through some supposedly Asian traits, though, shows why any number of subcontinental baskets would be hopelessly inadequate for the job. Most inhabitants of the Far East have epicanthic folds on their eyes, for instance -- but so do the Khoisan (the "Bushmen") of southern Africa. Shovel-shaped incisors -- the term refers to the slightly scooped-out shape of the back side of the front teeth -- do indeed show up in Asian and American Indian mouths more often than in other people's, but they also pop up a lot in Sweden, where very few people have coarse, straight black hair, epicanthic folds, or short body stature.

The straightforward biological fact of human variation is that there are no traits that are inherently, inevitably associated with one another. Morphological features do vary from region to region, but they do so independently, not in packaged sets. "I tell my students that I could divide the whole world into two groups: the fat-nose people and the skinny-nose people," says Norm Sauer. "But then I start adding in other traits to consider, like skin color, eye color, stature, blood type, fingerprints, whatever. It doesn't take long before somebody in the class gets the point and says, 'Wait a minute! Pretty soon you're going to have a race with only one person in it.'"

Indeed, despite the obvious physical differences between people from different areas, the vast majority of human genetic variation occurs within populations, not between them, with only some 6 percent accounted for by race, according to a classic study done in 1972 by geneticist Richard Lewontin of Harvard. Put another way, most of what separates me genetically from a typical African or Eskimo also separates me from another average American of European ancestry.

But if the bushel-basket view of race is insupportable, does that mean that the concept of race possesses no biological reality? "If I took a hundred people from sub-Saharan Africa, a hundred from Europe, and a hundred from Southeast Asia, took away their clothing and other cultural markers, and asked somebody at random to go sort them out, I don't mink they'd have any trouble at all," says Vincent Sarich of the University of California at Berkeley, a controversial figure in biological anthropology since the late 1960s, most recently because of his views on the race issue. "It's fashionable to say there are no races. But it's silly."

It's certainly true that indigenous Nigerians, for instance, look different from native Norwegians, who look different from Armenians and Australian aborigines. But would these differences be just as obvious if you could see the whole spectrum of humanity? Since people tend to mate with others in their immediate geographic area, there should be only a gradual change from one region to the next in the frequency of various genes and the morphological features they code for. In this scenario, human variation is the result of a seamless continuum of genetic change across space. The race concept, on the other hand, lumps people into clearly delineated groups. This, says anthropologist Loring Brace of the University of Michigan, is a purely historical phenomenon.

"The concept of race didn't exist until the invention of oceangoing transport in the Renaissance," Brace explains. Even the most peripatetic world travelers -- people like Marco Polo or the fourteenth-century Arabian explorer Ibn Battutah -- never thought in racial terms, because traveling by foot and camelback rarely allowed them to traverse more than 25 miles in a day. "It never occurred to them to categorize people, because they had seen everything in between," says Brace. "That changed when you could get into a boat, sail for months, and wind up on a different continent entirely. When you got off, boy, did everybody look different! Our traditional racial groupings aren't definitive types of people. They are simply the end points of the old mercantile trade networks."

Sarich, however, is not so willing to dismiss race as an accident of history. "I don't know whether Marco Polo referred to race or not," he says. "But I'll bet that if you were able to ask him where this person or that one came from just by looking at their physical features, he would be able to tell you."

If populations were uniform in density all over the world, adds Sarich, then the whole panoply of human variation would indeed be distributed smoothly, and race would not exist. But populations are not so evenly dispersed. Between large areas of relatively high density there are geographic barriers -- mountain ranges, deserts, oceans -- where population densities are necessarily low. These low-population zones have acted as filters, impeding the flow of genes and allowing distinct, discernible patterns of inheritance – races -- to develop on either side. The Sahara, for instance, represents a formidable obstacle to gene flow between areas to the north and south. Such geographic filters have not completely blocked gene flow, notes Sarich -- if they had, separate human species would have developed -- but their influence on the pattern of human variation is obvious.

GIVEN THE layered confusion surrounding the term race -- and its political volatility -- it's no wonder that scientists struggle over its definition and question its usefulness. Surveys of physical anthropologists have found that almost half no longer believe that biological races exist. "Historically, the word has been used in so many different ways that it's no longer useful in our science," says Douglas Ubelaker of the National Museum of Natural History at the Smithsonian Institution. "I choose not to define it at all. I leave the term alone."

The other half, however, contends that simply saying you choose not to define race won't make it go awav. "A popular political statement now is, 'There is no such thing as race,' " notes Alice Brues, a physical anthropologist at the University of Colorado. "I wonder what people think when they hear this. They would have to suppose that the speaker, if he were dropped by parachute into downtown Nairobi, would be unable to tell, by looking around him, whether he was in Nairobi or Stockholm. This could only damage his credibility. The visible differences between different populations of the world tell everyone that there is something there."

And, says Brues, we have to find a way to discuss just what that something is, and why it's there. "There are situations when you have to talk about things, and you have to have words to do it," she says. "Forensic anthropology is one such situation. The police want to know, is this a black person, a white person, maybe an Indian? You have to use words."

Like Sauer, Ubelaker is often asked by law enforcement agencies to identify unknown human remains. If racial divisions are merely cultural artifacts, then how are the two men so readily able to glean a person's racial identity from the purely physical evidence of a fleshless skull? The answer, they say, lies in geography and demographics. "I don't have any problems with the idea that there's human variation that is systematic," says Sauer;

"I can look at somebody and say, 'Your ancestors are probably from Europe.' I know that they're not going to be from South Africa or East Asia. But that still doesn't mean it's reasonable to take the world's population and divide it into three groups."

If the body Sauer identified in 1984 as that of a 19-year-old black female had been found in a different country, he says, he might have come up with a different identification. But American forensic standards are specifically designed to discriminate between people of West African, European, Asian, and American Indian descent, since those are the groups that make up the bulk of the American population. Given the bones' location, Sauer says, the odds were very good that the deceased would have identified herself as African American.

"A lot of us could narrow down the geographic origin of a specimen a lot more," says Sauer, "but I don't do that because the police have a form, and I want my form to match their form."

Anthropologist George Armelagos of Emory University, an outspoken critic of the biological concept of race, says it's a cop-out for anthropologists to continue using racial categories just because that's what law enforcement agencies ask for. "That doesn't seem legitimate to me," says Armelagos. "If we want to educate people on the concept of race, we should be doing it at all levels."

"Engaging a detective in a theoretical discussion on the true nature of human geographic variation isn't going to help him solve a case," counters Sauer. "I've come to the conclusion that if the police want race, I give them race. Maybe afterward, when we're having a beer, we can have a discussion about what race really means."

MEDICAL RESEARCHERS, unlike the anthropologists, seem to have little question about the reality of racial categories. Race, it seems, is quite useful for organizing data; each year dozens of reports in health journals use it to show purported clear differences between the races in susceptibility to disease, infant mortality rates, life expectancy, and other markers of public health. Black men are supposedly 40 percent more likely to suffer from lung cancer than are white men, and a number of recent studies on breast cancer seem to show that black women tend to develop tumors that are more malignant than those found in white women. In the United States, black infants are almost two and a half times more likely to die within the first 11 months of life than white infants. And it's been shown that American Indians are far more likely than blacks or whites to carry an enzyme that makes it harder for them to metabolize alcohol; this would leave them genetically more vulnerable to alcoholism. Other studies claim to demonstrate racial differences in rates of cardiovascular disease, diabetes, kidney ailments, venereal disease, and a host of other pathologies.

Are these studies pointing at genetic differences between the races, or are they using race as a convenient scapegoat for health deficiencies whose causes should be sought in a person's socioeconomic status and environment? The lung cancer statistic, for instance, should really be considered along with the numbers that show that black men are much more likely to smoke than are white men.

A recent study of hypertension in black Americans, conducted by Randall Tackett and his colleagues at the University of Georgia, exemplifies the difficulties found in trying to tease out a single answer to such a question. It's been known for some 30 years that blacks in the United States are almost twice as likely as whites to suffer from hypertension, or high blood pressure -- a condition that carries with it an increased risk of heart failure, stroke, hardening of the arteries, and other cardiovascular diseases. Black men are reported to have a 27 percent higher death rate from cardiovascular disease than white men, and black women a 55 percent higher rate than white women. What causes this discrepancy is still unknown: some investigators have attributed the higher incidence of hypertension in blacks to socioeconomic factors such as psycho-social stress, poor diet, and limited access to health care, while others have suggested a genetic predisposition to the disorder, which is often taken to mean a racial predisposition. Trying to track down a genetic cause, however, has proved even more confounding than it might otherwise have been, since high blood pressure can be the result of a number of factors, ranging from higher dietary sodium levels to increased exposure to psychological insult.

Yet last June, Tackett and his associates reported on a possible physiological mechanism underlying the higher incidence of hypertension in blacks. They exposed veins obtained during heart-bypass operations to chemicals that stressed the tissues and caused them to constrict, and found that veins from blacks were slower to return to normal size than those taken from whites. Veins that stay constricted longer in response to stress allow less blood to flow through and require the heart to work harder -- the essence of hypertension. "This is the first direct demonstration that there are racial differences at the level of the vasculature," says Tackett.

The hope is that these findings will lead the medical community to treat hypertension in blacks even more aggressively, and that they will thus save lives. But whether the findings really say anything about the role of race in disease is another matter altogether. Tackett's sample of African Americans was limited to 22 individuals from southern Georgia; would blacks from Los Angeles or New York, living in different circumstances and with different genetic histories, show the same blood vessel impairment? What about native Africans, who unlike their American counterparts generally have remarkably low rates of hypertension? And what about the Finns and the Russians, who have high rates? What do the findings say about their race? And even if American blacks have a greater susceptibility to hypertension primarily because of their blood vessels and not the inequities in their socioeconomic status, who's to say that those inequities -- environmental stresses that American whites never have to face -- aren't the trigger for the prolonged, potentially lethal constriction? Isn't it possible that the chain of causation leading from blood vessels to blood pressure to heart disease is anchored not in race, but in racism?

After all, Lewonon's study, done more than two decades ago, showed that the concept of race doesn't really have much of a genetic punch. "I'm not denying that the difference Tackett sees is there," says Armelagos. "But race only explains 6 percent of human biological variation. How can he be so sure that the 6 percent accounts for the pathology?"

The techniques used in genetic analysis have improved vastly since Lewontin's 1972 study; though race is responsible for just a small amount of genetic difference, it is now somewhat easier to distinguish one population from another and place an individual by looking at a sample of DNA. Of course, there are still limits. "If you ask me to look at a sample and say whether it came from Wales or Scotland, that would be tough," says Peter Smouse, a population geneticist at Rutgers. "But ask me if somebody is from Norway or Taiwan, sure, I could do that. Humans are tremendously variable genetically across the planet, almost surely representing how long we've been out there and spreading around. Now, whether the piles are nice and neat is not so clear; they're probably not as neat as would be convenient for someone who wanted to make piles."

In the end, says Smouse, no one would deny that there are genetic differences between groups of people. But in comparison with the differences between, say, chimps and humans, those dissimilarities shrink to "totally nothing." It's all a matter of perspective.

"What you make of race depends on what the question is," says Smouse. "And who wants to know."

At the time of this article's publication, James Shreve was a frequent writer for Discover on human evolution and biology. He is the coauthor, with anthropologist Donald Johnson, of Lucy's Child: The Discovery of a Human Ancestor.