The Blackwell Companion to Sociology (53 page)

racial and ethnic minorities at greater risk for the negative consequences of

frequent transfusions. The term phenotypically matched blood basically means

that it is possible to use the social appearances of race as a rough approximation (of likely antigens) to screen to minimize antibodies (along with ABO and Rh).

Transfusion therapy for sickle cell anemia is limited by the development of

antibodies to foreign red cells (Vichinsky et al., 1990). In one important study, the researchers evaluated the frequency and risk factors associated with alloimmunization, and obtained the transfusion history, red cell phenotype, and devel-

opment of alloantibodies in 107 black patients with sickle cell anemia who

received transfusions. They then compared the results with those from similar

studies in 51 black patients with sickle cell disease who had not received

transfusions and in 19 non-black patients who received transfusions for other

forms of chronic anemia.

220

Troy Duster

We assessed the effect that racial differences might have in the frequency of

alloimmunization by comparing the red-cell phenotypes of patients and blood-

bank donors (n = 200, 90 percent white). Although they received transfusions less frequently, 30 percent of the patients with sickle cell anemia became alloimmunized, in contrast to 5 percent of the comparison-group patients with other forms of anemia (P less than 0.001). Of the 32 alloimmunized patients with sickle cell

anemia, 17 had multiple antibodies and 14 had delayed transfusion reactions.

Antibodies against the K, E, C, and Jkb antigens accounted for 82 percent of the alloantibodies.

They then go on to conclude: ``These differences are most likely racial. We

conclude that alloimmunization is a common, clinically serious problem in sickle cell anemia and that it is partly due to racial differences between the blood-donor and recipient populations'' (Vichinsky et al., 1990).

Note that Vichinsky and his colleagues concludè`that it is partly due to racial

differences between the blood-donor and recipient populations.'' True enough,

this may not bè`race'' in any essentialist conception, but that is precisely the point. Over eighty years ago, Hirschfeld and Hirschfeld (1919, p. 675) posited

that when introducing the blood of one species into that of the same species

``those antigen properties which are common to the giver and receiver of blood

can not give rise to antibodies, since they are not felt as foreign by the immunized animal.'' While the Hirschfelds were talking about dogs, they were drawing a

straight line toward humans, human classification, and racial taxonomy: `Ìf we

inject into dogs the blood of other dogs it is in many cases possible to produce antibodies. By means of these antibodies we have been able to show that there

are in dogs two antigen types. These antigen types, which we recognize by means

of the iso-antibodies, may designate two biochemical races'' (Hirschfeld and

Hirschfeld, 1919, p. 675±6).

Using this hypothesis, they went on to perform the first systematic and

comprehensive serological study of a variety of ethnic and racial groups. As I

indicated above, their classification system did not survive the test of time, but `à way of thinking'' persists (Marks, 1995). Moreover, with the data reported in the Vichinsky study (given that the increased blood donations from ``blacks'' is a key policy goal for those suffering from a relatively common genetic disease, sickle cell anemia), the resuscitation of ``race'' through blood antigen theorizing and empirical research cannot be far behind. That persons who are phenotypically

``white'' can and do have sickle cell anemia complicates any essentialized racial theorizing, to be sure ± but for the purposes of further action (blood donation

requests and transfusion direction), racial phenotyping as à`short-hand'' is back with us at the beginning of the twenty-first century.

This provides a remarkably interesting intersection. While the full range of

analysts, commentators, and scientists ± from postmodern essayists to molecular

geneticists to social anthropologists ± have been busily pronouncing ``the death of race,'' for practical clinical purposes the concept is resurrected in the conflation of blood donation frequencies, by ``race.'' I now want to make it clear that I am not merely trying to resurrect ``race'' as a social construct (with no biological The Sociology of Science

221

meaning) ± no more than I am trying to resurrect ``race'' as a biological construct with no social meaning. Instead, I am arguing that when ``race'' is used as a

stratifying practice (which can be apprehended empirically and systematically)

there is often a reciprocal interplay of biological outcomes that makes it impossible to completely disentangle the biological from the social. While that may be obvious to some, it is completely alien to others, and some of thosèòthers'' are key players in current debates about the biology of race.

In late September 1996, Tuskegee University hosted a conference on the

Human Genome Project, with specific reference to the Project's relevance to

the subject of race (Smith and Sapp, 1997). In attendance was Luca Cavalli-

Sforza, a pre-eminent population geneticist from Stanford University and per-

haps the leading figure behind the Human Genetic Diversity Project. (This

project has been concerned with tracing human populations through evolution-

ary history of many centuries, and is not be to be confused with the Human

Genome Project, the goal of which is to map and sequence the entire human

genome.) Cavalli-Sforza had appeared on the cover of Time magazine a few

years earlier, as something of a hero to the forces that were attacking the genetic determinism in The Bell Curve, a popular book by Richard Herrnstein and

Charles Murray (1994). At this conference, he repeated what he had said in

the Time article: `Òne important conclusion of population genetics is that races do not exist'' (Smith and Sapp, 1997, p. 53).

If you take differences between two random individuals of the same population,

they are about 85% of the differences you would find if you take two individuals at random from the whole world. This means two things: (1) The differences between

individuals are the bulk of the variation; (2) the differences among populations, races, continents are very small ± the latter are only the rest, 15%, about six times less than that between two random individuals of one perhaps very small population (85%). Between you and your town grocer there is on average a variation

which is almost as large as that between you and a random individual of the whole world. This person could be from Africa, China, or an Australian aborigine. (Smith and Sapp, 1997, p. 55)

Cavalli-Sforza is speaking here as a population geneticist, and in that limited

frame of what is important and different about us as humans he may be

empirically correct. But humans give meaning to differences. At a particular

historical moment, to tell this to an Albanian in Kosovo, a Hutu among the

Tutsi, a Zulu among the Boers, or a German Jew among the Nazis may be as

convincing, for the purposes of further action, as telling it to an audience of

mainly African Americans at Tuskegee University, as Tuskegee was the site of the infamous syphilis experiments on black males, where the Public Health Service

of the US Government studied the racial effects of how disease ravages the body

of blacks in contrast to whites (Jones, 1981). Indeed, David Botstein, speaking

later in a keynote address, had this to say about The Bell Curve:

So from a scientific point of view, this whole business of The Bell Curve, atrocious though the claims may be, is nonsense and is not to be taken seriously. People keep 222

Troy Duster

asking me why I do not rebut The Bell Curve. The answer is because it is so stupid that it is not rebuttable. You have to remember that the Nazis who exterminated

most of my immediate family did that on a genetic basis, but it was false. Geneticists in Germany knew that it was false. The danger is not from the truth, the

danger is from the falsehood. (Smith and Sapp, 1997, p. 212)

David Botstein is also a pre-eminent molecular geneticist at the vanguard of

his field. In this statement, he takes the position that if people just understood the genetic truth, that would be sufficient and even corrective of racist thinking and action. But he acknowledges that the German geneticists knew that Nazi

claims about Aryan racial purity were false. Even though people may someday

come to understand that they are basically similar at the level of the DNA, RNA, immunology, or kinds of blood systems, it is the language group, kinship,

religion, region, or race that is still far more likely to generate their pledge of allegiance.

The American Anthropological

Anthropological Association

Association Statement

on ``Race''

In May 1998 the American Anthropological Association issued its own state-

ment on ``Race.'' It attempts to address myths and misconceptions, and in so

doing takes à`corrective'' stance toward folk beliefs about race. The statement

strongly states the position that ``physical variations in the human species have no meaning except the social ones that humans put on them.'' But in casting ``the problem'' in this fashion, it gives the impression that the biological meanings

that scientists attribute to race are biological facts, while the social meanings that lay persons give to race are: (a) either errors or mere artificial social constructions; and (b) not themselves capable of feedback loops into the biochemical,

neurophysiological, and cellular aspects of our bodies, which in turn can be

studied scientifically. The statement of the Anthropological Association is con-

sistent with that of the UNESCO statement on race. However, formulating the

matter so that `ìt is only the social meanings that humans provide'' implies that mere lay notions of race provide a rationale for domination, but have no other

utility.

There is profound misunderstanding of the implications of à`social contruct-

ivist'' notion of social phenomena. How humans identify themselves, whether in

religious or ethnic or racial or aesthetic terms, matters for their subsequent

behavior. Places of worship are socially constructed with human variations of

meaning, interpretation, and use very much in mind. Whether a cathedral or

mosque, a synagogue or Shinto temple, thosè`constructions'' are no less ``real''

because one has accounted for and documented the social forces at play that

resulted in such a wide variety of ``socially constructed'' places of worship

(Fujimura, 1998). ``Race'' as social construction can and does have a substantial effect on how people behave. One important arena for further scientific exploration and investigation is the feedback between that behavior and the biological

The Sociology of Science

223

functioning of the body. It is now appropriate to restate the well known social

analytic aphorism of W. I. Thomas, but to refocus it on human taxonomies of

other humans: if humans define situations as real, they can and often do have

real biological and social consequences.

Molecular Genetics and the New Conflation

Conflation of Race

and Forensics

If ``race'' is a concept with no scientific utility, what are we to make of a series of articles that have appeared in the scientific literature over the past seven years, looking for genetic markers of population groups that coincide with common

sense, lay renditions of ethnic and racial phenotypes? It is the forensic applications that have generated much of this interest. Devlin and Risch (1992a)

published an article on `Èthnic differentiation at VNTR loci, with specific

reference to forensic applications'' ± a research report that appeared prominently in the American Journal of Human Genetics.

The presence of null alleles leads to a large excess of single-band phenotypes for blacks at D17S79 (Devlin and Risch, 1992b). . . . This phenomenon is less important for the Caucasian and Hispanic populations, which have fewer alleles with a

small number of repeats . . . it appears that the FBI's data base is representative of the Caucasian population. Results for the Hispanic ethnic groups, for the D17S79

locus, again suggest that the data bases are derived from nearly identical populations, when both similarities and expected biases are considered. . . . For the allele frequency distributions derived from the black population, there may be small

differences in the populations from which the data bases are derived, as the

expected bias is .05. (Devlin and Risch, 1992a, pp. 540, 546)

Some explanations are in order. Alleles are different versions of one gene. For

example, while thè`generic gene'' will instruct the proteins and cells to make an eye, a particular allele may produce a blue, brown, or grayish-green eye. The

same can be said for the allelic variation that produces the epicantic fold (or lack of it) over the eye, and a wide range of other human physiognomic characteristics, including hair texture and skin color. While some aspects of skin color are of course environmental, these account for only about 20 percent of the variation. Moreover, this is not to suggest that single genes determine skin color, but that allelic variations even from full siblings can produce considerable variation.

When researchers try to make probabilistic statements about which group a

person belongs to, they look at variation at several different locations in the

DNA ± usually from three to seven loci. For any particular locus, there is an