{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Goodman_article

Goodman_article - Sebastifio Salgado Brasil 1984 BRED IN...

Info iconThis preview shows pages 1–6. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Background image of page 6
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Sebastifio Salgado, Brasil, 1984 BRED IN THE BONE? For all their claims about the usefulness ofrace, physicians and forensic experts leave a trail of misdiagnoses and misidenty‘ications in their wake BY ALAN H. GOODMAN N THE MORNING OF MAY 30, 1995, RES— cue workers in Oklahoma City made a final, melancholy sweep through the ruins of the Alfred P. Murrah Federal Building. In the weeks after the building was bombed, 165 victims had been discovered and removed, but three more bodies had been lodged in places too unstable to reach. Rather than risk more lives in a futile rescue—any survivors of the blast would have long since died of starvation or suf— focation—workers simply had marked the three locations with Day—G10 orange paint, before bringing down the rest of the building with dynamite. Now they picked method~ ically through the rubble, searching for glimpses of orange. Clyde Snow, a forensic anthropologist with a long his— tory of identifying victims of war crimes, was stationed in the state morgue at the time, listening to reports from the bomb site. “Everything was going swimmingly,” he later recalled. “When they got down to level zero, people could hear them talking on their mobile phones: ‘Okay, we have one, two, three bodies. . . . Fine, wrap it up, we can all go 20 THE SCIENCES - Mania/April 1.997 home.” The rescue team, events soon showed, was jump— ing the gun just a bit. Two or three minutes after the third body had been found, a voice suddenly broke back over the airwaves: “Hey wait a minute! We’ve got a leg down here. A left leg.” During the explosion and its aftermath, about twenty—five of the victims had been dismembered. Snow assumed, at first, that the leg must belong to one of those. “In all the confusion, with bodies going back and forth for X rays, I thought somebody just overlooked that one body had a left leg missing,” he said. “So we’ll just match it up.” But one recount after another yielded the same number: 168 right legs, 168 left legs; none of the survivors was missing a leg. “We went through autopsy records, pathology reports, body diagrams, and photographs. I did it twice, the pathologist did it twice,” Snow said. “It was just a mathematical paradox.” Baflled, Snow took a closer look at the leg itself. Sheared off just above the knee by the blast, it still wore the remains ofa black, military—style boot, two socks and an olive—drab blousing strap. Its skin, Snow said, suggested “a darkly com— plected Caucasoid.” By measuring the lower leg and plugging the numbers into computer programs that categorize bones by race and sex, Snow confirmed his hunch: the leg probably came from a white male. An attorney for the prime suspect in the bombing, Timothy J. McVeigh, pounced on the news, sug— gesting that the leg belonged to the “real bomber.” Snow wondered if it might belong to one of the transients who hung out on the first floor of the building. Fred B. Jordan, the Chief Medical Examiner for the state of Oklahoma, guessed that the leg belonged to a person walking alongside the truck carrying explosives. As it turned out, the leg belonged to none of the above. Its owner was one Lakesha R. Levy of New Orleans, an Airman First Class, stationed at Tinker Air Force Base in Midwest City, Oklahoma. On April 19 Levy had gone to the Mur— rah building to get a Social Security card and gotten caught near the epicenter of the blast. Levy was five feet, five inches tall, twenty-one years old and female. She was also, in the words of one forensics expert, “obviously black.” With that dis- closure, McVeigh’s attorney declared, “no one can have confidence in any of the forensic work in this case.” Just a few weeks before the leg was found, in the pages of this magazine, Snow had said that he could accurately discern a victim’s race from its skull 90 percent of the time [see “Murder Most Foul,” by Clyde Snow, May/June 1995]. True, a skull provides more clues to its owner’s identity than a leg does, and Levy’s leg was discovered and examined under extremely trying conditions. But the leg was still covered in skin, only part— ly decomposed, and skin is the most com— mon indicator of“race.” In fact, numerous examples suggest that mistakes like the one in Oklahoma City are common. They are common not because forensics experts do shoddy work—they don’t, the errors in Okla— homa City notwithstanding—but be— cause their conclusions are based on a deeply flawed premise. As long as race is used as a shorthand to describe human biological variations—variations that blur from one race into the next, and are great— est within so—called races rather than among them—misidentifications are inevitable. Whether it is used in police work, medical studies or countless every- day situations where people are grouped Vernon Fisher, White Hunter, 1991 biologically, the answer is the same: race science is bad science. HIRTY YEARS AGO, THE I American paleontologist George Gaylord Simpson declared all pre—Darwinian definitions of humanity worthless. “We will be better 0E,” he wrote, “if we ignore them com— pletely.” The scientific concept of race— an outgrowth of the Greek idea of a great chain of being and the Platonic notion of ideal types—i5 anti—evolutionary to its core. It should therefore have been the first relic consigned to the scrap heap. Race should have been discarded at the turn of the century, when the Amer— ican anthropologist Franz Boas showed that race, language and culture do not go hand—in—hand, as raciologists had contended. But race persisted. It should have vanished in the 19305, when the “new evolutionary synthesis” helped explain subtle human variations. Yet between 1899, when William Z. Rip- ley published Rates of Europe, and 1939, when the American anthropologist Carleton S. Coon published a book by the same name, the concept of race as type persisted almost unchanged. (Coon, on the eve of the Second World War, went to some lengths to ponder the essence of Jewishness. “There is a qual— ity of lookingJewish,” he wrote “and its existence cannot be denied”) Race should have disappeared in the 19505 and 19605, when physical anthropologists switched from studying types to study— ing variations as responses to evolution- ary forces. But race lived on. To Coon, for instance, races just became popula— tions with distinct adaptive problems. Most anthropologists today acknowl— edge that biological races are a myth. Yet the idea survives, in a variety of forms. A crude typology of world Views goes something like this. At one end of the spectrum are the true believers: At the University of Western Ontario in Lon— don, for example, the psychologist J. Philippe Rushton asserts that there are three main races—Mongoloid, Negroid and Caucasoid—and he ranks them according to intelligence and procreative ability. Here, sure enough, the old racial stereotypes leak out: the two traits allegedly appear in inverse proportion. You can have either a large brain or a large . . . (insert sexual organ of choice). Rushton’s Mongoloids rank as the most March/April 1997 ' THE SCIENCES 21 intelligent; Negroids allegedly have the strongest sexual drive; Caucasoids fit into the comfortable middle. At the other end of the spectrum are two groups who agree that races are a myth, but draw radically different con— clusions from that premise. The politically conservative group, known for proclaiming a “color—free society,” argues that if races do not exist, sociopolitical policies such as affir— mative action ought not to be based on race. Social con— structionists, on the other hand, realize that race—as—bad—biol— ogy has nothing to do with race—as-lived—experience. Social policy does not need a biological basis, especially when a dark—skinned American is still roughly twice as likely to be denied a mortgage as is a light—skinned person with an equiv— alent income. True races may not exist, but racism does. A fourth group, the confused, occupies the middle ground. Some do not understand why race biology is such bad science, yet they avoid any appeal to race because they do not want to be politically incorrect. Others apply race as a quasi—biological, quasi—genetic category and cannot fig— ure out what is wrong with it. Still others think the stance against racial biology is political rather than scientific. That middle category of the confused is huge. It includes nearly all public health and medical professionals, as well as most physical anthropologists. Moreover, the continued “soft” use of race by that well—meaning group acts to legit— imize the “hard” use by true believers and scientific racists. And if most professionals are confused about race, most of the public is both dazed and confused. There is no single, stable or monolithic public perception about race, but races are generally thought to be about genes (or blood) and (only slightly less permanent) cultural ties. Regardless, race is con— sidered to be deep, primordial and constant: in short, indis— tinguishable from its nineteenth—century definition. N 1992 THE FORENSIC ANTHROPOLOGIST NOR— man]. Sauer of Michigan State University in East Lansing published an article in the journal Social Science and Medicine provocatively titled, “Forensic Anthro— pology and the Concept of Race: If Races Don’t Exist, Why Are Forensic Anthropologists So Good at Identifying Them?” Race may be unscientific, Sauer argued, but people of one socially constructed racial category still tend to look alike— and different from the people of another “race.” The bio— logical anthropologist C. Loring Brace of the University of Michigan in Ann Arbor explains Sauer’s paradox in a slight— ly different way. Forensic scientists are good at estimating race, Brace says, because so—called racial variations are statis— tically confounded with real regional differences. People do vary in a systematic way depending on their environment. Both arguments make sense, and forensic anthropolo— gists do important work. But how good are they, really, at identifying race? Like Snow, the authors of forensic texts and review articles typically maintain that the race of a skull can be correctly identified between 85 and 90 percent of the time. The scientific reference for those estimates—if cited as anything other than common knowledge—is a sin~ gle, groundbreaking study by the physical anthropologists Eugene Giles, at the University of Illinois in Urbana—Cham— paign, and Orville S. Elliot, at the University of Victoria in British Columbia. In the early 19605 Giles and Elliot 22 THE SCIENCES - A'Iuri'II/xlpril I997 measured the skulls of modern, adult blacks and whites who had died in Missouri and Ohio, many of them at the turn of the century, as well as Native American skulls from a prehistoric site in Indian Knoll, Kentucky. Using a statis— tical equation known as a discriminant function, they then identified a combination of eight measurements that could determine a skull’s “race” once its sex was known. When Giles and Elliot applied the formula to additional skulls fitom the same collections, it agreed with the race assigned to the deceased at death between 80 and 90 per— cent of the time. To be useful, however, the formula has to work in places other than Missouri, Ohio and prehistoric Kentucky. I have found four retests of the Giles and Elliot method, and their results do not inspire confidence. Two of the retests restricted themselves to Native American skulls: in one of them almost two—thirds of the skulls were correctly classified as Native Americans; in the second, only 31 per— cent were correctly classified. For the two other studies, in which the skulls were of mixed race, skulls were correctly identified as Native American just 18.2 percent and 14.3 per— cent of the time. Thus in three of the four tests, the formu— la proved less accurate than a random assignment of races to skulls—not even good enough for government work. Contemporary Native American skulls may be particu— larly hard to classify because the formula is based on a very old sample. But the four retests were carried out on com~ plete crania that had already been sexed, a necessary pre— requisite to determining race. Forensic anthropologists often have much less to go on. Moreover, Native Americans are easier to classify than Hispanics or Southeast Asians, not to mention infants, children or adolescents of any race. At best, in other words, racial identifications are depressingly inac— curate. At worst, they are completely haphazard. How many e: s, : .% % % §~ . a v wwmwaamnmm bodies and body parts, like Lakesha Levy’s leg, are sending investigators down wrong paths because the wrong box was checked off? Forensic anthropologists usually blame such mistakes on the melting pot. Yet distinct racial types have never exist— ed. What changes are social definitions of race—the color hn%and human biology. \X/hites in Cleveland in 1897 were different from whites in Amarillo, Texas, in 1997. Science 101: generalizations ought not be based on an ill—defined, constantly changing and contextually loaded variable. KULLS AND CORPSES, ONE COULD ARGUE, HAVE ceased to care to which race they belong— though their families and friends might disagree. But when physicians base their actions on perceived racial categories, their patients ought to care a great deal. Does Lorna Simpson, Wigs (detail), 1995 race, however imperfect a category, help physicians diag~ nose, treat, prevent or understand the etiology of a disease? Before the Second World War, physicians were often blinded by the conviction that certain races suffered from certain diseases. People who had sickle—cell anemia, for instance, were assumed to have “African blood.” In 1927 the American physician J. S. Lawrence discovered a case of the disease in a “white” person. “Special attention was paid to the question of racial admixture of negro blood in the family but no evidence could be obtained,” Lawrence wrote in the journal of Clinical Investigation. “There must be some caution in calling this sickle—cell anemia because no evidence of negro blood could be found.” Evelynn M. Hammonds, a historian of science at the Massachusetts Institute of Technology, has brought to my attention some early diagnoses of ovarian cysts that express the same logic. In 1899 the American physician Thomas R. Brown reported that he often heard surgeons say that tumors found in black women had all the features of ovar— ian cysts, “but inasmuch as the patient is a negress it is cer— tainly not so, as multilocular cysts are unknown in the negress.” The following year Daniel H. Williams, the emi— nent African—American physician and the first American to perform successful heart surgery, quoted a physician from Alabama speculating that: “Possibly the Alabama negro has not evoluted to the cyst—bearing age.” Williams went on to show unambiguously, in a study, that ovarian cysts are common in black women including women from Alaba— ma. He noted that white physicians have a history of ignor— ing black women, then offered examples of black women whose cysts swelled to 100 pounds or more before they were diagnosed. ODAY THE PARADIGM OF RACIALLY DISTINCT diseases has been replaced by the more flexi— ble idea of race as disease risk factor. Yet the medical effects are the same. Some 25 million Americans are said to suffer from osteoporosis, a progressive loss of bone mass that leads to 1.5 million fractures a year. Since the nineteenth century, blacks have been thought to have thicker bones than whites have and to lose bone mass more slowly with age. (A few years ago, when a dentist Visited my laboratory, he was shocked to find that neither one of us could tell a black jaw from a white one.) In the journal Seminars in Nuclear Medicine, a review titled “Osteoporosis: The State of the Art in 1987” listed race as a major risk fac- tor. The section on race begins: “It is a well—known fact that blacks do not suffer from osteoporosis.” That “fact” is backed by a single reference, a seminal paper by the American physical anthropologist Mildred Trotter and her colleagues titled “Densities of Bones of White and Negro Skeletons.” Trotter and her colleagues evaluated the bone densities of skeletons from forty adult blacks and forty adult whites. They excluded skeletons with obvious bone diseases, but they did not describe how they chose the cadavers or whether the samples were matched for causes of death, diet or other known risk factors for osteoporosis. Of the ten bones they studied in each skele— ton, Trotter and her colleagues found that six tended to be denser in blacks than in whites; the other four showed no differences by race. Furthermore, the authors wrote, the decline in density took place at “approximately the same rate” for each sex—race group. Trotter and her colleagues may have realized that their data could be overinterpreted. In later publications they present scatterplots with age on one axis and bone density on the other. The scatterplots confirm that bone densities tend to decline with age: the clusters of data points slope downward. It is a challenge, however, to discern any dif— ference between the densities of bones from blacks and those from whites. The six lowest radius densities, for exam— ple, were found in bones of blacks. Let me be clear: I am only following citations to see if the data say what the references say they say. But my con— clusion is dismaying. If the “well—known fact that blacks do not suffer from osteoporosis” is based on poorly inter- preted data, then black women may not be getting enough March/April 1997 ' THE SCIENCES 23 preventive care, are not targeted in the media and are under— diagnosed as osteoporotic. In every instance I have cited, a double leap of scientif— ic faith seems to have taken place. First, a serious medical condition (sickle—cell anemia, ovarian cysts, osteoporosis) is regarded as genetic, even though environmental factors have not been adequately examined. Second, anything genetic is assumed to imply a panracial phenomenon. Thus, what might be true in a statistical sense is assumed true for all members of a so—called race. All blacks are protected from osteoporosis. All blacks are less prone to heart disease. By the same logic, Native Americans have some special predisposition to obesity and diabetes, though, in truth, rates vary wildly among groups and regions. HY ARE MY FINDINGS MORE THAN IDIO— syncratic examples? Why does race not work as a shorthand for biological varia— tion? The answer lies in the structure of human variation and in the chameleon-like concept of race. 0 Most traits vary in small increments, or clines, across geographic areas. Imagine a merchant walking from Stockholm, Swe— den, to Cape Town, South Africa, in the year 1400. He would notice that the skin colors of local people darkened until he reached the equator, then slowly turned lighter again. If he took a different route, perhaps starting in Siberia and wandering all the way to Sin— gapore, he would observe the same phenomenon, though none of the people he passed on this second route would be classified as white or black today: all of them would be “Asian.” Race, in other words, does not determine skin col— or, nor does skin color determine race. As Frank B. Living— stone, an anthropological geneticist at the University of Michigan in Ann Arbor, put it more than th'nty years ago: “There are no races, there are only Clines.” 0 Most traits are nonconcordant. That is, traits tend to vary in different and entirely independent ways. If you know a person’s height, you can guess weight and shoe size because tall people tend to be heavier and have bigger feet than short people. Those traits are concordant. By the same token, however, you could guess nothing about the per— son’s skin color, facial features or most genes. Height is nonconcordant with nearly every other trait. If you know skin color, you might be able to guess eye color and per— haps (but surp...
View Full Document

{[ snackBarMessage ]}