why race still matters

8/16/2019 Why Race Still Matters

1/15102 Dædalus Winter 2005

Why has race mattered in so manytimes and places? Why does it still mat-ter? Put more precisely, why has there

been such a pervasive tendency to applythe category of race and to regard peopleof different races as essentially differentkinds of people? Call this the ‘½rst ques-tion.’ Of course there are many morequestions that one must also ask: Whyhas racial oppression been so ubiqui-tous? Why racial exploitation? Why ra-cial slavery? Perhaps we tend to think of races as essentially different just becausewe want to excuse or to justify the domi-nation of one race by another.

I shall proceed with the ½rst questionby canvassing ½ve possible answers to itthat variously invoke nature, genealogy (inthe sense of Michel Foucault), cognitive science, empire, and pollution rules.

One ½nal preliminary remark is inorder. Most parts of this essay couldhave been written last year or next year,

but the discussion of naturalism, medi-cine, and race could only have been writ-ten in November of 2004, and may wellbe out of date by the time this piece isprinted.

Why has the category of race been sopervasive? One answer says that the dis-tinction is just there, in the world for all

to see. Super½cial differences betweenraces do exist in nature, and these arereadily recognized.

The naturalist agrees at once that thedistinctions are less in the nature of things than they once were, thanks to in-terbreeding among people whose ances-tors have come from geographically dis-tinct blocks. Racial distinctions are par-ticularly blurred where one populationhas been translated by force to live in themidst of another population and yet hasnot been assimilated–slaves taken fromWest Africa and planted in the SouthernUnited States, for example. The natural-ist notes that traditional racial distinc-tions are less and less viable the morechildren are born to parents whose geo-graphical origins are very different.

Sensible naturalists stop there. Thebelief that racial differences are anything

Ian Hacking

Why race still matters

Ian Hacking, a Fellow of the American Academy

since 1991, holds the chair of Philosophy and

History of Scienti½c Concepts at the Collège de

France. His work spans the philosophy of science,

the philosophy of language, the theory of proba-

bility and statistical inference, and the socio-his-

torical examination of the rise and fall of disci-

plines and theories. His most recent books are

“Mad Travellers” (1998), “The Social Construc-

tion of What” (1999), and “Historical Ontol-

ogy” (2004).

© 2005 by the American Academy of Arts& Sciences


2/15

more than super½cial is a repugnanterror. John Stuart Mill was the wisestspokesman for this position.

Here, in modern terminology, is hisdoctrine: (1) Nature makes differencesbetween individuals. These differencesare real, not constructed. (2) We classifythings according to differences we ob-serve. Classi½cations are made by peopleand encoded in social practices, institu-tions, and language. (3) Some classes aresuch that their members have little incommon except the marks by which wesort them into those classes–call thosesuper½cial kinds. (4) Other classes havemembers with a great many things incommon that do not follow from the

marks by which we sort them into class-es. These are “real Kinds.”1

Examples? “White things,” he wrote,referring not to race but to the color it-self, “are not distinguished by any com-mon properties except whiteness; or if they are, it is only by such as are in someway dependent on, or connected with,whiteness.” But horses, to use one of his

other examples, have endless properties

in common, over and above whatevermarks we use to distinguish them fromother animals or other kinds of things.Horses form a real Kind, but the class of white things is a super½cial kind.

The contemporary philosophical con-cept of a ‘natural kind’ is a descendentof Mill’s notion. Nonphilosophers whohave come across this phrase may sup-pose it refers to a well worked out, tech-nical, and stable concept. I argue else-where that it does not.2

Mill himself was as notable a profemi-nist and antiracist as can be claimed fora white nineteenth-century man. Al-though he argued that real Kinds exist,he at once went on to ask whether the

races and sexes are real Kinds, or if theyare merely super½cial, like the classi½ca-tions “Christian, Jew, Musselman, andPagan.” The religious confessions arenot real Kinds, he argued, because thereis no property that Christians have andMuslims lack, or vice versa, except what-ever follows from their faiths.

What about race? Most anthropolo-

gists of Mill’s day held that there were½ve races, named geographically but rec-ognized by color: Caucasian, Ethiopian,Mongolian, American, and Malayan.According to Mill, color and certainother physiological traits are the marksby which we distinguish members of the different races. Races would be realKinds if there were endlessly many other

differences between the races that didnot follow from the marks by which wedistinguish them. Are there endlesslymany such differences?

Well, you cannot rule that out a priori,Mill thought. “The various races andtemperaments, the two sexes, and eventhe various ages, may be differences of Kind, within our meaning of the term. I

Why racestill matters

1 His own words are old-fashioned but lovely.The differences between members of classes“are made by nature . . . while the recognitionof those differences as grounds for classi½cationand of naming is . . . the act of man.” However,“we ½nd a very remarkable diversity . . . be-tween some classes and others.” Only super-½cial resemblances link members of one typeof class, while members of classes of the other

type have a vast number (he said an endlessnumber) of properties they share. Those thatshare an almost endless number of propertiesare his real Kinds. From John Stuart Mill, ASystem of Logic: Ratiocinative and Inductive, ½rstpublished in 1843. The discussion of racial clas-si½cation is found in bk. 1, chap. 7, sec. 4. Thechanges Mill made in later editions of the bookinvolved sex, not race–doubtless because Millhoped to get the questions about sex exactlyright for Harriett Taylor. See chap. 7, on Millon

classi½cation, in my forthcoming book, TheTradition of Natural Kinds (Cambridge Univer-sity Press).

2 This is one of the conclusions urged in mybook The Tradition of Natural Kinds.

Dædalus Winter 2005 103



IanHacking onrace

say they may be; I do not say they are.”Mill believed that only empirical sciencecould determine whether the variousraces, as distinguished by color and afew other features, pick out classes thatare distinct in a great many unrelatedways. “If their differences can all betraced to climate and habits [or, headded in later editions, to some one ora few special differences in structure],they are not, in the logician’s view, spe-ci½cally distinct.” He would have beenpleased by Anthony Appiah’s careful dis-cussion of very much the same questionusing more recent terminology. Sciencemight have revealed an endless numberof differences between the races that are

not consequences of the marks by whichwe distinguish them, namely color andphysiognomy. But science has not doneso, and almost certainly will not. Mill,like Appiah, thus concludes that theraces are not real Kinds.

This conclusion, however, does notanswer, or aim at answering, the speci½cquestion I raised at the outset, of why

there is such a pervasive tendency toapply the category of race. Maybe Millthought the answer was obvious. Thedesire of one racial group to dominate,exploit, or enslave another demands le-gitimacy in societies that, like modernEurope and America, are committed toversions of egalitarianism. Race scienceswere devised to discover a lot of differ-

ences between races that do not followfrom the marks of color and structure bywhich we distinguish them. You do nothave to treat people equally, if they aresuf½ciently different.

Although it takes us some distancefrom the ‘½rst question,’ some recentevents force us to clarify the naturalist

position on race. In an important edito-rial on the U.S. census published in theyear 2000, Nature Genetics stated: “That

race in this context is not a scienti½cterm is generally acknowledged by sci-entists–and a message that cannot berepeated enough.” An editorial in 2001observed that “scientists have long beensaying that at the genetic level there ismore variation between two individualsin the same population than betweenpopulations, and that there is no biologi-cal basis for ‘race.’”3 Now–in Novem-ber of 2004–this selfsame journal hasproduced a special supplement on themedical and genetic uses of racial andethnic classi½cation. And the November11 issue of The New England Journal of Medicine highlights the news of the ‘race-based’ drug targeted at African Ameri-

cans suffering from certain types of heart failure. All this is breaking news.Hence what follows cannot be de½ni-tive, but one may hope that a perspectivesomewhat distanced from media discus-sion can be useful even in the midst of it.

We must ½rst update Mill with a littlelogic. When he wrote about differencesbetween classes, he had in mind proper-

ties that serve to distinguish members of one class from another in a uniform way.A uniform difference between cows andhorses is something that is true in themain of any cow but not true in themain of any horse–digestion by rumi-nation, for example. There are ever somany such differences between horsesand cows; hence they are real Kinds.

Call them uniform differences. There are agreat many uniform differences that dis-tinguish horses from other kinds of ani-mals, but almost no uniform differencesthat distinguish white things from greenthings, except their color, or Muslimsfrom Christians, except their faith.

Writing in 1843, Mill had little occa-sion to think about statistical differ-

3 “Census, Race and Science,” Nature Genetics24 (2000): 97; “Genes, Drugs and Race,” NatureGenetics 29 (2001): 239.

http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2000)24L.97[aid=189368]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2000)24L.97[aid=189368]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2001)29L.239[aid=6458928]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2001)29L.239[aid=6458928]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2001)29L.239[aid=6458928]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2001)29L.239[aid=6458928]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2001)29L.239[aid=6458928]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2000)24L.97[aid=189368]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1061-4036(2000)24L.97[aid=189368]


4/15



IanHacking onrace

at the time of this writing, in Novemberof 2004–are scared of the idea that thetraditional list of races employed by tra-ditional racists might be statistically sig-ni½cant classes. With good reason!

Ten years ago The Bell Curve by RichardHerrnstein and Charles Murray attracteda great deal of attention. The authorsclaimed that the Gaussian distributionsof iq scores establish a natural distinc-tion of some importance between differ-ent races. They forcefully argued that theclass of African Americans is a statisti-cally signi½cant class–signi½cant withrespect to a property they called intelli-gence, and which they measured with iq tests.

They did not imply that the races arereal Kinds. That is, they did not statethat there is a host of uniform differ-ences between Caucasian Americans andAfrican Americans. Readers not unrea-sonably assumed, however, that the au-thors meant exactly that. At any rate, theauthors clearly were not talking about

mere correlations, namely, disparitiesbetween iq scores within different racialgroups. But they did not establish thatthese disparities are statistically mean-ingful to any biological understanding.

About the same time that The BellCurve was published, ogre naturalists,such as Philippe Rushton in Race, Evolu-tion, and Behavior , made more sweeping

claims to biologically grounded racialdifferences. They claimed that the racesare distinguished by many propertiesrightly prized or feared for differentstrengths and weaknesses. If that weretrue, then races would exactly ½t Mill’sde½nition of a real Kind.

One deplores both Rushton and The Bell Curve, but there is an absolutely fun-

damental logical difference betweenwhat the two assert. Rushton claimedthat the races are real Kinds. One imag-

ines that Herrnstein and Murray thoughtso too, but what they claimed was thatthe races are statistically signi½cantclasses. And they implied that this is sta-tistically meaningful.

Despite the fact that his doctrines havea centuries-old pedigree, we can dismissthe egregious Rushton. We can also re-fute Murray and Herrnstein.4 Mill’stype of naturalism has contempt forboth doctrines. Loathing of these quiterecent doctrines and their predecessorshas, not surprisingly, produced revul-sion against any sort of naturalism aboutrace. Today there is some consternationover the appearance of what is calledrace-based medicine.

The science of medicine was for quitea long time the science of the Europeanmale body, with footnotes for non-European or female bodies. All that haschanged: those footnotes are now chap-ters. But the current situations for thegroups that had been relegated to thefootnotes are quite different. Many

medical differences between males andfemales are uniform, but medical differ-ences between races are almost alwaysonly statistical.

We have long known that some ail-ments are restricted to some gene pools.Tay-Sachs is a hereditary disease (inwhich an enzyme de½ciency leads to theaccumulation of certain harmful resi-

dues in the brain and nerve tissue, oftenresulting in mental retardation, convul-sions, blindness, and, ultimately, death)that almost exclusively affects youngchildren of eastern European Jewish de-

4 There is a tendency among proper-thinkingpeople to dismiss The Bell Curve cavalierly, asboth wrong-headed and refuted, without actu-ally saying why. Many things wrong, and one

has an obligation to say what. My own ‘genea-logical’ objections are stated in a piece in The

London Review of Books, January 26, 1995.


6/15

scent. ‘Ashkenazi’ is a valuable geo-graphical, historical, and social classi½-cation. It is geographical because it indi-cates where members of this class, ortheir near ancestors, came from, namely,eastern Europe. It makes a contrastwith Sephardic Jews, whose roots arein Spain. In modern Europe and NorthAmerica, social differences between theAshkenazi and Sephardic hardly matterto most people, but they remain signi½-cant in North Africa and West Asia. Un-til further interbreeding makes it totallyobsolete, Ashkenazi is a statistically sig-ni½cant and a statistically meaningfulclass with respect to Tay-Sachs disease.

There are similar geographical-histor-

ical indicators for lactose intoleranceand for an inability to digest fava beans.West African ancestry is an indicator forbeing a carrier of the sickle-cell anemiatrait, which confers some immunityagainst malaria. This trait was oftenstigmatized as simply ‘black.’ In fact,it is primarily West African, although itshows up in Mediterranean populations

where malaria was a major selector forsurvival. The indicator was abused forracial reasons in widespread screening.

“Drug approved for Heart Failure inAfrican Americans”–headline on the½rst business page of The New York Times, July 20, 2004. Here we go again? Quitepossibly. “The peculiar history [of thisdrug] on the road to the market presents

a wide array of troubling and importantissues concerning the future status of race as a category for constructing andunderstanding health disparities inAmerican society.”5 For a stark remind-er of the commerce, the Times reportedthat the previous day the stock of thedrug’s maker, NitroMed, rose from $4.31

to $10.21, and had reached $16 at midday.This story has been ongoing for a decadein medical, commercial, and regulatorycircles.

There are real problems about the ra-cially targeted heart drug. BiDil is a mix-ture of two well-known heart medica-tions. Scienti½c papers assert, ½rst, thatother medicines are not as good for Afri-can Americans with heart failure as theyare for other Americans with this prob-lem, and, second, that BiDil works betterfor African Americans with certain spe-ci½cs than any other drug on the mar-ket.6 In fact, randomized trials were dis-continued because the drug was mani-festly effective on black patients. No-

body well understands why. The reasonscould be at least in part social and eco-nomic (including dietary) rather thanhereditary. The correlation is stronglysigni½cant, but it is not statisticallymeaningful at present from a geneticor other biological point of view.

Even if one is a complete skepticabout, for example, a genetic basis for

the differential ef½cacy of the drug, thedrug does appear to be statistically usefulin treating the designated class of pa-tients. That means that race may be auseful indicator to a physician of the po-tential effectiveness of this rather thananother drug–under present social andhistorical conditions.

Now turn to leukemia. Bone marrow

transplants help an important class of patients. Donors and recipients musthave matching human leukocyte anti-gens (hlas); at present, doctors try tomatch six different types of them. If apatient has no relative to serve as a do-nor, matches are hard to come by. The

5 Frederick Kahn, “How a Drug Becomes ‘Eth-

nic’: Law, Commerce, and the Production of Racial Categories in Medicine,” Yale Journal of

Health Policy, Law, and Ethics 4 (2004): 46.

6 Anne L. Taylor, “Combination of Isosorbide

Dinitrate and Hydralazine in Blacks with HeartFailure,” New England Journal of Medicine 351(2004): 2049–2057.



http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1535-3532(2004)4L.46[aid=6458927]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1535-3532(2004)4L.46[aid=6458927]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1535-3532(2004)4L.46[aid=6458927]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-4793(2004)351L.2049[aid=6458926]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-4793(2004)351L.2049[aid=6458926]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-4793(2004)351L.2049[aid=6458926]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-4793(2004)351L.2049[aid=6458926]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-4793(2004)351L.2049[aid=6458926]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1535-3532(2004)4L.46[aid=6458927]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1535-3532(2004)4L.46[aid=6458927]



IanHacking onrace

relevant antigens are unevenly distrib-uted among ethnic and racial groups.7

There exist registries of possible donors–truly generous persons, for at presentdonation of bone marrow is quite har-rowing. Happily, free-floating stem cellsin the blood also help, but the donormust take a lot of drugs to boost thosestem cells. Another source of cells is um-bilical cord blood. But this, like all theother options, requires antigen match-ing.

In the United States, the NationalBone Marrow Program maintains themaster registry. Most people in existingregistries have tended to be middle-agedand white, which means that whites

have a good chance of ½nding a match.Hence there have been racially targetedprograms for Asian and African Ameri-cans. In the United States and Canadathere is also the Aboriginal Bone Mar-row Registries Association, and in theUnited Kingdom there is the AfricanCaribbean Leukemia Trust. Asians forMiracle Marrow Matches has been very

successful, especially in the Los Angelesregion. The African Americans Unitingfor Life campaign has been less success-ful, for all sorts of historical reasons. AnAfrican American with leukemia has afar worse chance of ½nding a match intime than members of other populationshave. That is a social fact, but there isalso a biological fact: there is far greater

heterogeneity in the human leukemiaantigen in persons of African originsthan in other populations.8 (This fact½ts well with the hypothesis that all

races are descendants of only one of many African populations that existed atthe time that human emigration beganout of Africa–populations whose char-acteristics have continued to be distrib-uted among Africans today.)

If you go to the websites for the organ-izations that maintain the registries, youwill see they do not shilly-shally in somedance of euphemistic political correct-ness about race. For them it is a matterof life and death. Without the Asian reg-istries there would have been manymore dead Asian Americans in the pastdecade. For lack of more African Ameri-cans on the registries there will be moredead African Americans in the next few

years than there need be.We certainly lack a complete under-

standing of the distribution of humanleukemia antigens in different geograph-ically identi½ed populations. But we dohave some biological understanding of the underlying causal differences. Andrace is a very useful quick indicator of where to look for matches, just as the

bmi is a useful quick indicator of poten-tial health problems.

So when, if ever, is it useful to speakin terms of the category of race, on thegrounds that the races in some contextsare not only statistically signi½cant butalso statistically useful classes? To an-swer this question, we can use our dis-

tinctions:• The Bell Curve may show that iq is a

statistically signi½cant characteristic

7 This also matters to renal transplants. SeePauline C. Creemers and Delawir Kahn, “AUnique African hla Haplotype May Identify aPopulation at Increased Risk for Kidney GraftRejection,” Transplantation 65 (1998): 285–288.

8 For hla differentiation, see T. D. Lee, A. Lee,and W. X. Shi, “hla-a, -b, -d and -dq Antigens

in Black North Americans,” Tissue Antigens(1991): 79–83. For maps, see, for example,one of the essays in the November NatureGenetics issue referenced in the text: Sarah A.Tishkoff and Kenneth K. Kidd, “Implications

of Biogeography of Human Populations for‘Race’ and Medicine,” Nature Genetics Supple-ment 36 (2004): 521–527.

http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0041-1337(1998)65L.285[aid=2512052]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0041-1337(1998)65L.285[aid=2512052]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-2815(1991)0L.79[aid=6458925]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-2815(1991)0L.79[aid=6458925]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0041-1337(1998)65L.285[aid=2512052]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-2815(1991)0L.79[aid=6458925]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-2815(1991)0L.79[aid=6458925]


8/15

of some American subpopulations, butit is neither meaningful from a biologi-cal point of view nor useful for anywell-de½ned purpose.

• Some medications may be less effec-tive, and BiDil may be more effective,for African Americans with certain

types of heart failure. If so, this is sta-tistically signi½cant and statisticallyuseful for helping patients, but (in myopinion) it is at present not statisticallymeaningful.

• The relationships between humanleukemia antigens and race are statis-tically signi½cant, statistically mean-ingful for a biological understanding,

and statistically useful in making mar-row matches possible for minoritygroups.

It is not a good idea, in my opinion, tospeak of BiDil as a race-based medicine,as do The New York Times and othermedia. The drug is not in the least based on race. It is quite possible that the rea-son it is more useful for African Ameri-

cans than for other large and looselycharacterized groups has less to do withthe inherent constitution of their cardio-vascular systems than with a mixture of social factors. If we had reliable data onthe relevance of diets shared by a sub-class of white and black Americans, wemight be able to help whites with similardiets. The drug would not then be ‘diet-

based’ but ‘diet-targeted.’ If you ½nd ituseful to use the word ‘race,’ say ‘race-targeted’ medicine.

I should have thought that the differ-ential distribution of human leukocyteantigens would be esoteric enough toescape notice. Not so. The StormfrontWhite Nationalist Community, whosebest-known ½gure is the neo-Nazi Da-vid Duke, is having a good time on one

branch of its website discussing hladiversity. In my opinion, the correct

strategy is not to play down the differen-tial distribution of hla, but to make itcommon knowledge that speci½c differ-ences among peoples may be used inhelping them–in much the same waythat white Australians, given their so-cially induced tendency to overexposethemselves to the sun, should be target-ed to cut down on the rate of death dueto skin cancer.

I have introduced these remarks tomake plain that naturalism about race,far from being an atavistic throwback toan era well left behind, is a topic for to-day, one about which we have to becomeclearer. Not because the races are realKinds, denoting essentially different

kinds of people. But because already weknow that the races are not only statisti-cally signi½cant classes for some dis-eases, but also statistically useful. Somecorrelations are statistically meaningful.There is every reason to believe thatmore statistically meaningful correla-tions will be discovered.

Every time such a phenomenon is

found useful, the racists will try to ex-ploit the racial difference: witness theneo-Nazi use of differential antigens.Hence we need to be fully aware of whatis involved.

A historian may well despise the com-placency of naturalism. Differences be-tween the races have seemed inevitable

in the West, it will be argued, because of a framework of thought whose originscan be unmasked only by a genealogy.Classi½cation and judgment are seldomseparable. Racial classi½cation is evalua-tion. Strong ascriptions of comparativemerit were built into European racialclassi½cation and into evaluations of human beauty from the beginning. Andso the Caucasian face and form weredeemed closest to perfect beauty.





IanHacking onrace

That is the vein in which Cornel Westhas sketched a genealogy of modern ra-cism.9 Though his is not exactly a deepgenealogy in the spirit of Nietzsche andFoucault, it is an excellent résumé of events. I wish only to comment on hisstarting point, less to correct it than toencourage rethinking the connection be-tween race and geography.

According to West, “the category of race–denoting primarily skin color–was ½rst employed as a means of classi-fying human bodies by François Bernier,a French physician, in 1684. He dividedhumankind into four races: Europeans,Africans, Orientals and Lapps.” Notethat none of these is named by color

and that the ½rst three are identi½ed bywhere they live or come from. It hardlymatters now, but the fourth name,“Lapp” (probably derived from a wordmeaning simpleton), for the people whocall themselves Sami, is about as racist adesignation as there is. Bernier seems tohave met only two Lapps, and he foundthem loathsome, and he simply reports

that other unnamed travelers told himthat the inhabitants of Laponia were“vile animals.”10

There are certain emendations to bemade in Cornel West’s account. Bernierdid not designate a race restricted to Eu-ropeans. What he called the “½rst race[ sic ]” included Europeans (the disgust-ing Lapps aside), North Africans, and

the peoples of West and South Asia.

With some hesitation, he also includedNative Americans of both hemispheresin that category.

He did not classify by color but mostlyby facial features. Although he countedMongols, Chinese, and Japanese aswhite ( véritablement blanc ), he felt theyhad such differently shaped faces andbodies that they constituted a differentrace. Indigenous Americans were alsowhite. South Asians were less white (oli- vâtre), he thought, because of the torridclimate. When his categories (minus theLapps) were expressed in terms of colorduring the next century, they became‘white,’ ‘yellow,’ and ‘black’–categoriesstill going strong in Mill’s day. It may

come as some surprise that for high-brow race science, whites includedArabs, Turks, everyone on the Indiansubcontinent, and maybe Americans,that is, the indigenous ones.

Bernier does discuss color, but mostlywhen noting the existing hierarchy inthe Indian subcontinent, where thelighter skin of the Moghul elite puts

them ahead of the browner Hindus. Ber-nier’s observations of Africans seemedto be based almost entirely on Africanslaves, especially at Turkish or Arabslave markets (where of course he sawwhite, mostly female, slaves too). Yes,(sub-Saharan) Africans were black, butthey contrasted with the ½rst race chieflyin other aspects of the body, especially

the hair and lips. “Here Bernier,” SiepStuurman writes, “surely anticipates lat-er racial discourse.”11

In 1685, the year after Bernier pub-lished both his classi½cation of races andhis abridgement of Gassendi, Louis XIVpromulgated the rules of the Transat-lantic slave trade, the Code noir , makingthe effective identity of blackness and

9 Cornel West, “A Genealogy of Modern Ra-cism,” in West, Prophesy Deliverance!: An Afro-

American Revolutionary Christianity (Philadel-phia: The Westminster Press, 1982), 47–65.

10 François Bernier, “Nouvelle division de laterre,” Journal des Sçavans (April 24, 1684): 148–155. A de½nitive account of this paper is Siep

Stuurman, “François Bernier and the Inventionof Racial Classi½cation,” History Workshop Jour-nal 50 (2000): 1–21.

11 Stuurman, “François Bernier and the Inven-tion of Racial Classi½cation,” 4.

http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1363-3554(2000)50L.1[aid=6441930]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1363-3554(2000)50L.1[aid=6441930]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1363-3554(2000)50L.1[aid=6441930]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1363-3554(2000)50L.1[aid=6441930]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=1363-3554(2000)50L.1[aid=6441930]


10/15

slavery a point of law, in no need of anyrace science to legitimate it.12

In West’s important subthesis aboutaesthetics and human beauty, he showsthat Bernier’s conception was not simplythat black Africans were uglier than the½rst race. There was also the element of sexual exoticism. Bernier raved aboutAfrican women on display for sale inTurkey, naked. He regretted only thatthey cost so much.

West wanted to write a genealogy inpart because he had the insight to ad-dress an intellectual problem that is sel-dom stated: The oceanic empires of Eu-rope, chiefly France and Britain, and theUnited States in their wake, are unique

in world history in that the dominanttendency of their moral and politicalphilosophy from the start emphasizedequality. Backsliding and self-interestare apparent beyond exaggeration, butthe propensity for egalitarianism hasbeen permanent and progressive. At thesame time, West cites numerous cele-brated egalitarians and reminds us of

their persistent racism. In justice, Millhimself does not escape criticism.

How can racism and egalitarianismcoexist? Because equality is amongthose who are essentially the same. If races are essentially different, they neednot be treated alike. The framework forthis alliance was established at the be-ginning, West urges, and became en-

trenched as Western thought passedfrom the ½rst stage described in his ge-nealogy to the second. One can envisagebroadening West’s analysis into some-thing with the same form as Michel Fou-cault’s A History of Insanity in the Age of Reason–a history of racism in the age of equality. Stuurman, whom I have citedas the authority on Bernier, has impor-

tantly contributed on the other side, inhis newly published François Poulain and the Invention of Equality.

Now we turn to the universalist ap-proach favored in the cognitive sciences.It is proposed that human beings are

born with an innate capacity not only tosort other people along racial lines, butalso to act as if the differences distin-guished are essential characteristicsof people. This capacity is ‘prepro-grammed’ by a genetic inheritance andmatures and becomes operational early,say, at three or four years of age. A fur-ther proposal is that children are bornnot only with an ability to sort items

into speci½c types of classes, but alsowith a predisposition to identify certainproperties as essential to speci½c classes.

Lawrence Hirschfeld is an anthropolo-gist who works at the intersection of cognitive science and developmentalpsychology–to use proper names,the improbable intersection of NoamChomsky and Jean Piaget.13 Hirschfeld

draws on the work of psychologists,child-development experts, anthropolo-gists, linguists, philosophers, neurosci-entists, and others to postulate the dis-tinct innate cognitive modules withwhich all of us are born. These modules

13 Lawrence Hirschfeld, “The Conceptual Poli-tics of Race: Lessons from our Children,” Ethos25 (1997): 63–92. See also his book Race in the

Making: Cognition, Culture, and the Child’s Con-ception of Human Kinds (Cambridge, Mass.: mitPress, 1996). The expression ‘human kind’ isobviously derived from ‘natural kind.’ I regretthat it was I who put the phrase into circulationwith this use, in Paris in 1992, at a conferenceon culture and cognition attended by Hirsch-feld: Ian Hacking, “The Looping Effects of Hu-man Kinds,” in Dan Sperber, David Premack,and Ann James Premack, eds., Causal Cognition:

A Multidisciplinary Approach (Oxford: Oxford

University Press, 1995), 351–383. I have aban-doned this terminology, partly, and only partly,because it was modeled on the unsatisfactoryidea of a natural kind.

12 Louis Sala-Molins, Le Code noir, ou, le calvairede Canaan (Paris: Presses Universitaires deFrance, 1987).



http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0091-2131(1997)25L.63[aid=6458924]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0091-2131(1997)25L.63[aid=6458924]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0091-2131(1997)25L.63[aid=6458924]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0091-2131(1997)25L.63[aid=6458924]



IanHacking onrace

enable infants to acquire speci½c abili-ties. There is not just an all-purposemodule for sorting things accordingto their resemblances, but speci½c mod-ules for classifying living things, formaking judgments of number, for sort-ing according to motion, and so forth.

Where does race enter? Hirschfeldproposes a module that enables childrento distinguish different kinds of people.Some of the earliest distinctions chil-dren make using this module involve ra-cial traits, primarily stereotypical skincolor and a few facial characteristics.There is the further proposition thatdue to an innate disposition, the races,like any classes recognized using this

module, are treated as if they were es-sential characteristics of people. Experi-ments show that children believe thatchanging a person’s race, as marked bystereotypical features such as color,would change the kind of person thatthat individual is. In these ½rst experi-ments, children were asked only aboutblack and white individuals, illustrated

by simple cartoon representations.Hirschfeld’s initial data were drawnfrom experiments on school childrenin Ann Arbor, Michigan, but they nowappear to be con½rmed in results frommore diverse groups.

This cognitive theory proposes thatthe tendency to regard racial classi½ca-tions as essential is a corollary of a devel-

opmental fact about the human mind.We have a phenomenon on the order of the cognitive fallacies known from Tver-sky and Kahneman’s studies of decisionunder uncertainty. Whatever evolution-ary value our human kind module mighthave had, it made disastrous racist prac-tices all too easy. But this proposalstands wholly apart from ogre natural-

ists’ claim that the alleged differencesbetween the races are grounds for mak-ing social arrangements that discrimi-

nate between the races. The cognitivescientists will say their results show howhard we must ½ght to control our innatetendencies to ½nd essential differencesbetween races.

Hirschfeld’s analysis may be queriedon grounds speci½c to race. Experiment-ers are vigilant not to confuse culturalfrom cognitive input. They highlight theissue in titles such as Culture and Cogni-tion, which is the present approved wayto express the nature-nurture debate. Yetone cannot but suspect that they under-estimate how quickly very young chil-dren catch on to what is wanted of them.One might say, with a whiff of irony,that children have an innate ability to

½gure out what adults are up to, andhence to psych out the experimenters.

In any event, nurture has prepro-grammed very young Americans to at-tend to race. Well-intentioned televisionprogramming for children constantlyemphasizes that the characters, even if they are not human, are of differentraces. From infancy, children watch tele-

vision cartoons that show, for instance, ahappy black family playing with a happywhite family. The intended message isthat we can all get on well together. Thesubtext is that we are racially different,but should ignore it. Experimenters dis-cover that small children expect parentsof any color to have children of the samecolor. Is that proof of innate essentialism

or of the ef½cacy of television?

It is time to turn away from cognition,and back to institutions and history.

Categories become institutionalized,especially by censuses and other typesof of½cial tagging. It is important to re-member that the ½rst working Europeancensuses were carried out in colonies–

Quebec, New Spain, Virginia, and Ice-land. Categorization, census, and em-pire: that is an important nexus.


12/15

I turn to empire in part for personalreasons. Race, as a category, has its ownmanifest meanings in the United States.For me, race has of course the Americanconnotations, but other ones as well.The primal racial curse for me as a Cana-dian is my country’s history of relationswith the native peoples. Now I work inFrance, where the chief racial issue con-cerns people of North African descent.Despite all their differences, the Canadi-an, French, and American racial obses-sions have a single historical source:Empire. Conquest and control–whetherof North Africans, West Africans, or the½rst nations of North America.

On Webster ’s de½nition, empire–“a

state that has a great extent of territoryand a great variety of peoples under onerule”–is about the conquest of peoples.With it comes an imperial imperative toclassify and enumerate the conqueredpeoples. Thus the words cast in stonethree times–in Old Persian, Elamite,and Babylonian hieroglyphics–on theGreat Staircase of Persepolis at the hey-

day of the Persian Empire:A great God is Ahuramazda, who created

this earth, who created yonder heaven,

who created man, who created welfare for

man, who made Xerxes king, one king of

many, one lord of many. I am Xerxes the

great King, King of Kings, King of the

countries having many kinds of people,

King of this great earth far and wide, the

son of Darius the King, the Achaeme-nian.14

Xerxes (?519–465 b.c.e.) inherited thePersian Empire in 485. The lapidary in-vocation to his power, thought to datefrom the beginning of his reign, includescarved processions of the many peoples

he ruled. First come the Medes bearingvessels, daggers, bracelets, coats, andtrousers. Then twenty more stereotypesof peoples, each similarly accompaniedby their characteristic tribute. Theyprocess in the following pecking order:Medes, Elates, Parathions, Sogdians,Egyptians, Bactrians, Armenians, Baby-lonians, Cilicians, Scythians, Thracians,Assyrians, Phoenicians, Cappadocians,Lydians, Afghans, Indians, Macedo-nians, Arabs, Somalis, and Ethiopians.Surprise, surprise, the blackest comelast.

Empires have a penchant for classify-ing their subjects. Doubtless there areadministrative reasons: some conquered

societies furnish goods, some furnishsoldiers. But over and above practicalexigencies, there seems to be an impera-tive to classify subject peoples almost asan end in itself. Or rather, the end is tomagnify the exploits, glory, and power of the ruler. Classi½cation, as an imperialimperative, invites stereotyping.

Persepolis has seen other empires,

other conquests, a fact to which graf½tion the remaining walls of the city (ren-dered mostly by bored British soldiersfrom the eighteen and early nineteenthcenturies who identify themselves bytheir names, dates, and regiments)attests. There is only one inscription torival Xerxes’ own: an enormous dia-mond carved into the side of the only

standing entrance door of the royal gate.It is inscribed,

stanley

new york herald

1870

In the unvarnished words that describeHenry Morton Stanley in the 1911 editionof The Encyclopaedia Britannica, “In geo-graphical discoveries Stanley accom-plished more than any other explorer of Africa, with which continent his name is

14 Ali Sami, Persepolis (Takht-Jamshid ), 9th ed.,trans. R. Sharp (Shiraz: Musavi Printing Of½ce,1977), 35.




13/15


14/15

ple, sexual selection trumps pollutionrules, or vice versa.

How much more powerful pollutionand the imperial imperative becomewhen history puts them together! Pollu-tion rules are important for maintainingthe imperial group intact. As soon aspollution rules break down, men of themaster group sire children with womenfrom subjugated groups, and a new kindof person–the half-breed–emerges.The etymology of words such as ‘Eur-asian’ embodies this phenomenon. Welearn from the trusty 1911 Encyclopaediathat ‘Eurasian’ was “originally used todenote children born to Hindu mothersand European (especially Portuguese)

fathers.” There are pecking orders be-tween conquerors, as well as among theconquered–and this British word was aput-down meant to keep the Portuguesein Goa in their place. Note also the dom-inance order between the sexes: a Hindufather and a European woman wouldyield, at least in the of½cial reckoning,a Hindu, not a Eurasian.

The French noun métis, derived froma Portuguese word originally used forEurasians, dates back to 1615. In FrenchCanada it signi½ed the children of whitefathers and native mothers. Early in thenineteenth century it was adopted inEnglish to denote the offspring of French Canadian men, originally trap-per/traders, and native women. In other

words, ‘Eurasian’ and métis alike meantthe children of males from conqueringgroups of lower status and females fromthe totally subjugated groups–and thenthe offspring of any of those children.

For a few generations, one can be pre-cise in measuring degrees of pollution.At that the Spanish and Portuguese Em-pires excelled. First came ‘mulattoes,’the children of Spanish or Portuguesemen and South American Indian wom-en. With the importation of black slavesfrom West Africa, the label was trans-

ferred to the children of white mastersand black slaves, and then to mixed racein general. The oed says it all: the Eng-lish word is derived from Portugueseand Spanish, “mulato, young mule, henceone of mixed race.”

The Spanish cuarteron became the Eng-lish ‘quadroon,’ the child of a white per-son and a mulatto. The few quotationsgiven in the oed are a record of colo-nial history. Here is the ½rst, dated 1707:“The inhabitants of Jamaica are for themost part Europeans . . . who are theMasters, and Indians, Negroes, Mulatos,Alcatrazes, Mestises, Quarterons, &c.who are the slaves.” The next quotationin the list is from Thomas Jefferson.

And so on: from Spanish the Englishlanguage acquired ‘quintroon,’ meaningone who is one-sixteenth of Negro de-scent. The 1797 Encyclopaedia Britannicahas it that “The children of a white and aquintroon consider themselves free of all taint of the negro race.” More impor-tantly, from an 1835 oed citation, “‘Thechild of a Quintroon by a white father is

free by law.’ Such was recently the West-Indian slave code.” Better to have awhite father than a white mother.

In real life, interbreeding was endem-ic, so such classi½cations were bound tobecome haphazard. Only one option wasleft. The American solution was de½ni-tive. One drop of Negro blood suf½cedto make one Negro. Which in turn im-

plied that many Americans could makea cultural choice to be black or not, achoice turned into literature in ToniMorrison’s Jazz and, more recently, inPhilip Roth’s The Human Stain. The onedrop of blood rule perfectly harmonizesthe imperial imperative and the preser-vation of group identity by pollutionprohibitions.

Why is there such a widespread ten-dency to regard people of different races Dædalus Winter 2005 115



15/15

as essentially different kinds of people?That was our ½rst question.

I have argued that naturalism of thesort taken for granted by John StuartMill has more going for it than is com-monly supposed, and I have also ex-plained why it may make sense in the

context of medicine to regard races asstatistically signi½cant and also statisti-cally useful classes. But neither of theseforms of naturalism explains the wide-spread tendency to regard people of dif-ferent races as essentially different.

There is the cognitive answer, that es-sential distinction by race is the resultof a universal human kind module. Ihave discounted that, and have alsodismissed what I call ogre naturalism,which claims that races are real Kinds.Note, however, that if there is any ves-tige of truth in any type of naturalism,that could only reinforce the effect of other considerations.

We are left with Cornel West’s geneal-ogy of modern racism, pollution rules,and the imperial imperative. Together

they describe the foundation of the ra-cial predicament of the Western world.The imperial imperative employs a par-ticular type of pollution rule to reinforcecaste distinctions and degrees of subjec-tion within an empire. The racial essen-tialism of the European empires andtheir American continuation are to beregarded as a special case of the imperial

imperative.One speci½c feature of modern racism

–race science–results from a central as-pect of modern European history. Froma world-historical point of view, onlyone feature of early modern Europestands out. It is the coming into beingof modern science. The ½rst stage of West’s genealogy of modern racism is

wholly embedded in that period whenearly modern science developed. As biol-ogy emerged in the second stage, around

1800, so did race science, that strangeblend of evolutionary biology and statis-tical anthropology. In the heyday of pos-itivism, race science repainted old pollu-tion rules, the ones selected as suitingthe imperial imperative, with a veneerof objective fact.

There are two strands of thought inthe human sciences, the one universal-ist, the other emphasizing contingen-cies. They seldom harmonize. Here theydo. West’s genealogy is a wholly contin-gent account of the reasons for the per-vasive tendency to regard racial distinc-tions as essential. In contrast, the use of pollution rules is a universal techniquefor self-stabilizing a human group. Clas-si½cation of peoples by a category of race is an integral part of the controlnecessary to organize and maintain anempire, and it employs pollution rules.These observations suggest a fruitfulway to combine contingent and univer-sal theories that help to explain why thecategory of race remains so pervasive.

IanHacking onrace

why race still matters

Documents