Geoffrey C. Bowker
Susan Leigh Star
Cambridge, MA: MIT Press, 1999.

Table of Contents

Introduction: To Classify is Human
Chapter One: Some Tricks of the Trade in Analyzing Classification

Part One: Classification and Large Scale Infrastructures
Chapter Two: The Kindness of Strangers - Kinds and Politics in Classification Systems
Chapter Three: The ICD as Information Infrastructure
Chapter Four: Classification, Coding and Coordination

Part Two: Classification and Biography, or System and Suffering
Chapter Five: Of Tuberculosis and Trajectories
Chapter Six: The Case of Race Classification and Reclassification under Apartheid

Part Three: Classification and Work Practice
Chapter Seven: What a Difference a Name Makes - The Classification of Nursing Work
Chapter Eight: Organizational Forgetting, Nursing Knowledge, and Classification

Part Four: The Theory and Practice of Classifications
Chapter Nine: Categorical Work and Boundary Infrastructures; Enriching Theories of Classification
Chapter Ten: Why Classifications Matter

Chapter Six: The Case of Race Classification and Reclassification under Apartheid

Sir de Villiers Graaff asked where the sudden danger to the White group was that had caused the Minister to decide to close off the human stud-book he had tried to create. He was endeavoring to classify the unclassifiable. (Horrell, 1968: p. 27)

The stubborn survival of racial categories attests to the enduring power of the old race paradigm, as well as the fact that new insights and methodologies take time to be fully incorporated and internalized. (Dubow, 1995: 106)

As information scientists, the theoretical and practical issues of racial and ethnic group categorization, naming, and meaning can be viewed as empirical data about problems associated with the organization of knowledge, representation, classification, and standards setting. (Robbin, 1998, p. 3 of MS)

In the last chapter, we examined the detailed interactions between people, institutions and categories with respect to tuberculosis. Each has a trajectory, and the trajectories may pull, or torque each other over time if they move in different directions or at different rates. The threads which tie category to disease, to science, to bureaucracy and thus to person, often become twisted and tangled in the long process of the disease. The texture of classification here is composed of thick filiations, encompassing much of a person's life, imposed from outside, and filled with uncertainty and contradiction.

This chapter examines another similarly torqued group of filiations between people and classifications, that which tied racial categories to persons under apartheid in South Africa. Here, race classification and reclassification provided the bureaucratic underpinnings for a vicious racism. Here too, the attempt to create a normalized, systemic bookkeeping system was embedded in a larger program of human destruction. There are enduring lessons to be drawn about moral accountability in the face of modern bureaucracy. The ethical concerns are clearly basic questions of social justice and equity; at the same time, their very extremity can teach us about the quieter, less visible aspects of the politics of classification. We walk here a line similar to that of Hannah Arendt in her Eichmann in Jerusalem: A Report on the Banality of Evil (1963). The quiet bureaucrat "just following orders" is in a way more chilling than the expected monster dripping grue. Eichmann explained what he was doing in routine, almost clerical terms; this was fully embedded in the systematic genocide of the Holocaust.

One of the central arguments of this book is that classification systems are often sites of political and social struggles, but that these sites are difficult to approach. Politically and socially charged agendas are often first presented as purely technical. Secondly, they are difficult even to see. As layers of classification system become enfolded into a working infrastructure, the original political intervention becomes more and more firmly entrenched. In many cases, this leads to a naturalization of the political category, through a process of convergence. It becomes taken for granted. (We are using the word naturalization advisedly here - since it is only through our infrastructures that we can describe and manipulate nature.) In this chapter, we shall emphasize the stubborn refusal of "race" to fit the desired classification system suborned by its pro-apartheid designers. So doing, we will further develop the concept of torque to describe the interaction of classification systems and biography.


From the early days of Dutch settlement of South Africa, the de jure separation and inequality of people co-existed with interracial relationships. In the mid-19th century charter of the Union, it was simply stated that "equality between White and coloured persons would not be tolerated" (Suzman, 1960). Various laws were enacted which reinforced this stance. However, when the Nationalists came to power in 1948, a much more detailed and restrictive policy, apartheid, was put into place. In 1950 two key pieces of legislation, the Population Registration Act and the Group Areas Act were passed. These required that people be strictly classified by racial group, and that those classifications determine where they could live and work. Other areas controlled de jure by apartheid laws included: political rights, voting, freedom of movement and settlement, property rights, right to choose the nature of one's work, education, criminal law, social rights including the right to drink alcohol, use of public services including transport, social security, taxation, immigration (Cornell, 1960; United Nations, 1968). The brutal cruelty, of which these laws were the scaffolding, continued for more than four decades. Millions of people were dislocated, jailed, murdered and exiled.

The racial classification which was so structured in the 1950s sought to divide people into four basic groups: Europeans, Asiatics, persons of mixed race or Coloureds, and "Natives" or "pure-blooded individuals of the Bantu race" (Cornell, 1960). The Bantu classification was subdivided into eight main groups, with Xhosa and Zulu the most numerous. The Coloured classification was also complexly subdivided, partially by ethnic criteria. The terribly fraught (and anthropologically inaccurate) word Bantu was chosen over preference to African (or Black African), partly to underscore Nationalist desires to be recognized as "really African." (i)

State authorities, touching every aspect of work, leisure, and education obsessively enforced apartheid. In a bitter volume detailing his visit to South Africa, Kahn notes:

Apartheid can be inconvenient, and even dangerous. Ambulances are segregated. A so-called European injured in an automobile accident may not be picked up by a non- European ambulance (nor may a non- European by a European one), and if a white man has the misfortune to lead to death before an appropriate mercy vehicle materializes, he can comfort himself in extremis by reflecting that he will most assuredly be buried in an all-white century. (Nonwhite South African doctors may not perform autopsies on white South African corpses.) (Kahn: 1966: 32)

"Separate development" was the euphemism used by the Nationalist Party to justify the apartheid system. They argued from a loose eugenic basis that each race must develop separately along its natural pathway, and that race mingling was unnatural. This ideology was presented in state-sanctioned media as a common-sense policy (Cell, 1982).

Despite that fact that it was required by law, it often took months or years for Blacks to acquire passbooks, during which time they were in danger of jail or being deported to one of the Black homelands (Mathabane, 1986). Horrell recounts a story about the early years of apartheid, and a group of Black people waiting for hours outside the registration office:
The Native Affairs Department official tried to pacify them. He told them to come next Monday, next Wednesday. But no, they said. They had waited yesterday, and now losing their pay for today, and the rent would soon be due, and if they were not classified they would be arrested - "Hulle sal ons optel. My vriend was gister opgetal," they said. The word came from every quarter of the yard, -- 'optel, optel'.(ii) (Horrell, 1958: 59)

Any form of interracial sexuality was strictly forbidden by a series of immorality laws, some of which pre-dated apartheid. Ormond states that: "Between 1950 and the end of 1980 more than 11,500 people were convicted of interracial sex; anything from a kiss on up" (1986:33). These sexual borders were vigorously patrolled by police. Ormond continues:
Special Force Order 025A/69 detailed use of binoculars, tape recorders, cameras and two-way radios to trap offenders. It also spelled out that bedsheets should be felt for warmth and examined for stains. Police were also reported to have examined the private parts of couples and take people to district surgeons for examination. (Ormond, 1986: 33)

The South African Institute of Race Relations (Horrell, 1969) shows a typical year for charges and convictions under the Immorality Act in 1966-67. The racialized gender biases speak for themselves.

Figure 17. Charges And Convictions Under The Immorality Act During The Year Ending June 1967.

 Charged  Convicted
 Charged  Convicted


 671  349
 18  11
 20  5
 264  126
 11  4
 20  13
 8  5
 338  180

Source: Horrell, Muriel. A Survey of Race Relations. South African Institute of Race Relations, Johannesburg, January, 1969: 36.

For black South Africans, the system of segregation included a legal requirement to carry a pass book, a compilation of documents attesting to birth, education, employment history, marriage, and other life events. The books were over fifty pages long. No black was allowed to be in a white area for more than 72 hours without special permission, including government authorization for a work contract (such as that for a live-in servant). The consequences of transgression were severe, as Frankel (1979: 205) notes:

The inestimable number of 'illegals' in the urban areas live a life of harassment that is Kafkaesque in its proportions, yet even those fortunate enough to qualify for urban status are faced with a harsh and insecure daily existence where the loss of a document, some technical violation of the mass of administrative decrees, or some arbitrary (and often vindictive) stroke of the bureaucratic pen, can mean condemnation to perpetual displacement.

A Foucaldian system of control of all people except Whites ensued (although by law the restrictions applied to Whites entering proscribed areas, this was rarely enforced and Whites did not carry pass books) (Black Sash, 1971; Mathabane, 1986). Blacks were the major targets of scrutiny, and the pass book system allowed for comprehensive surveillance of their actions:

The whole system has been extended and rationalised over the years by widening the categories of officials who can formally demand the production of passes, and by linking this up with sophisticated computer technology centred on the reference book bureau of the Department of Plural Relations in Pretoria. (Frankel, 1979: 207)

These data were entered into a centralized database, cross-referenced across the different domains. Kahn (1966) notes that:
Every African over sixteen must have on his person what is called a reference book, a bulky document measuring five by three and a half inches and containing ninety-five pages. As a rule, it is only Africans who are stopped by the police and asked to produce their passes. "The African must be a collector of documents from the day of his birth to the day of his death," says a publication issued by the Black Sash (iii). His passbook must contain particulars about every job he has had, every tax he has paid, and every X-ray he has taken. He would be well advised, the Black Sash has suggested, not to let himself get too far away from his birth certificate, baptismal certificate, school certificates, employment references, housing permits, hospital and clinic cards, prison discharge papers, rent receipts, and, the organization has added sarcastically, death and burial certificates. (p. 91)

Insert Figure 18. South African Passbook.

Horrell (1960) relates a story of an illiterate man "D.L." of Natal, who was arrested for having removed pages from his passbook. He was fined £10 or two months in jail (a huge sum for a Black man at that time); unable to pay the fine, he went to jail. After being released, he could no longer find work, as he now had a prison record. A sympathetic literate friend investigated the case and found that the printers of the passbook had by accident eliminated pages 33 to 48, and instead had produced two sets of pages 49 to 64. D.L. had to appeal the conviction up to the Supreme Court level (again a costly and time-consuming business), where it was finally set aside.

In addition to the pass book system regulating the lives of Black South Africans, the state attempted to enforce many other forms of segregation. Christopher Hope, in his novel A Separate Development, writes of petty apartheid such as the segregation of buses and benches:

One lived, of course, surrounded by such signs and notices. Most of them, however, served some clear purpose, the point of which everyone recognized as being essential for their survival: WHITES ONLY on park benches; BANTU MEN HERE on nonwhite lavatories; or INDIAN BENCH; or DEFENSE FORCE PROPERTY: PHOTOGRAPHS FORBIDDEN; or SECOND-CLASS TAXI; or THIS PLAYGROUND IS RESERVED FOR CHILDREN OF THE WHITE GROUP. And, of course, people were forever being prosecuted for disobeying one or other of these instructions. (Hope, 1980: 20)

For apartheid to function at this level of detail, people had to be unambiguously categorizable by race. However, despite the legal requirement for certainty in race identification, this task was not to prove so easy. Many people did not conform to the typologies constructed under the law: especially people whose appearance differed from their assigned category, or who lived with those of another race, spoke a different language from the assigned group, or had some other historical deviation from the pure type. New laws and amendments were constantly being debated and passed (see, e.g. Rand Daily Mail, 1966). By 1985, the corpus of racial law in South Africa exceeded 3,000 pages (Lelyveld, 1985: 82).

Both the scientific theories about race and the street sense of terms were confused. Both prototypical and Aristotelian senses of categorization were used simultaneously, as with the example of the ICD in Chapter 3. The original official sorting by race after the 1950 Population Registration Act derived from the categories checked on the 1951 census returns. An identity number was given to each individual at that time (Horrell, 1958: 19). The Director of Census was in charge of deciding everyone's racial classification, on the basis of the census data, and, where necessary, other records of vital statistics. Horrell notes: "But this classification is by no means formal. Section Five(3) of the Population Registration Act provides that if at any time it appears to the Director that the classification of a person is incorrect, after giving notice to the person concerned, specifying in which respect the classification is incorrect, and affording him or her an opportunity of being heard, he may alter the classification in the register" (1958: 4). So in the case of apartheid, we have the scientistic belief in race difference on the everyday level, and an elaborate formal legal apparatus enforcing separation. At the same time, a much less formal, more prototypical approach uses an amalgam of appearance and acceptance -- and the on-the-spot visual judgments of everyone from police and tram drivers to judges -- to perform the sorting process on the street.

The conflation of Aristotelian and prototypical categories for race classification has deep historical roots in South Africa and elsewhere. The concept of racial types took firm hold in the nineteenth century across a range of natural and social sciences, and was embraced by the architects of apartheid. At the same time, of course, the pure type existed nowhere, and racism everywhere. As Dubow writes about the scientific history of South African racial theories:

The typological method is at the heart of physical anthropology. It was based on empiricist principles of classification taxonomy originally developed in the natural sciences. The conception of race as "type "encouraged a belief in the existence of ideal categories and stressed diversity and difference over similarity and convergence. This was overlaid by binary-based notions of superiority and inferiority, progress and degeneration. One of the many problems associated with the typological method was its fissiparous character. The search for pure racial types could not easily be reconciled with the evident fact that, in practice, only hybrids existed. New fossil discoveries led to a proliferation of variant racial types and ever more theories were developed to explain their affinities. (Dubow, 1995: 114-115)

Such difficulties are always present when trying to place people in racial categories (see e.g., López, 1996; Robbin, 1998; Harding, 1993). As Donna Haraway says of racial taxonomy in the United States:

In these taxonomies, which are, after all, little machines for classifying and separating categories, the entity that always eluded the classifier was simple: race itself. The pure Type, which animated dreams, sciences, and terrors, kept slipping through, and endlessly multiplying, all the typological taxonomies. The rational classifying activity masked a wrenching and denied history. As racial anxieties ran riot through the sober prose of categorical bioscience, the taxonomies could neither pinpoint nor contain their terrible discursive product. (1997: 234)

Although a vague conception of eugenics and other forms of scientific racism are woven throughout the debates about apartheid, this lack of a scientific definition of race appears repeatedly. Dr. M. Shapiro, at a meeting of the Medico-legal Society in Johannesburg in 1952, wryly notes that:

Where for purposes of legal classification, the question arises whether a person is White, Coloured, Negroid or Asiatic, the policeman and the tram conductor, unencumbered by biological lore, can make an assessment with greater conviction, and certainly with fewer reservations, than can the geneticist or anthropologist. Indeed, the law being traditionally intolerant of uncertainty in matters of definition, the evidence of the scientist on the subject of race can only prove an embarrassment to the Courts if not to himself. (quoted in Suzman, 1960: 353)

In a legal article reviewing race classification in 1960, Suzman concludes: "As the present study has revealed, the absence of uniformity of definition flows primarily from the absence of any uniform or scientific basis of race classification. Any attempt at race classification and therefore of race definition can at best be only an approximation, for no scientific system of race classification has as yet been devised by man. In the final analysis the legislature is attempting to define the indefinable" (Suzman, 1960: 367). Landis (1961) similarly notes that the definition of the law was inherently ambiguous; she argues this was intentional, and that the ambiguity shifted the burden of proof to the individual. In a case where a person could not be proved to be either European or non-European, the burden of proof would fall on disproving the non-European side.

Of course the lack of scientific definition had no bearing on the brutal consequences of the classification, despite the fissiparous -- branching and dividing -- nature of the scientific problem.

Conflicting Categories in South Africa

Different aspects of apartheid law could classify a person differently. Where a woman lived, for example, often depended on her husband's classification, although movement from Bantu to White was not possible this way. So she might be of Indian national origin classified as Asian, married to a man classified as Coloured, and live in a Coloured zone but only be able to work or go to school in an Asian zone. This could be impossible or very arduous due to distance and the segregated transportation infrastructure:

Under the Population Registration Act, the children of mixed unions are, it appears, generally being classified according to the "lower "of the two layers involved -- that is, the group carrying fewer privileges..... But under the Group Areas Act, the children of Coloured and African parents, or Asian and African parents, would while they were minors presumably be classified according to the racial group of the father in order that they might live with him and his wife in his group area. The child of an Indian father and an African mother might, thus, be brought up in an Indian environment, but, on reaching the age of 16 and receiving his identity card might be forced to leave his parents and change his mode of living and his associates to those of the African group. (Horrell, 1958:12-13)

Again, the racialized gender structure is prominent here, where patrilinearity and patriarchal definitions of the couple's race are followed. Oddly, at times the multiple, contradictory methods of classifying could be used subversively to work in favor of the individual who lived between the categories. As Horrell recounts:

A third case is described to show the absurdities that may arise. Mr. T. is in appearance obviously Coloured, and his sons and daughter are near-White: his sons, in fact, served as Europeans in the army. Both of them now live as Coloured men and were so classified. But Mr.T. trades in an African location and wants to continue doing so. It is said that he asked the official to classify him as an African: certainly this was done (1958: 53).

In one infamous example, a jazz musician, Vic Wilkinson of Cape Town was born to a Coloured man and a White woman, and originally classified White. After apartheid he was reclassified as Coloured, and then twice more reclassified as he married women of different races and moved to different neighborhoods. (Note that the remarriages took place outside of South Africa for legal reasons.) Finally, both he and his Asian wife Farina were reclassified Coloured, allowing them and their children to live together. At the age of fifty, Vic actually received a new birth certificate -- and crossed the race lines for the fifth time (Sunday Times, 1984).

Figure 19. Vic Wilkinson and his Family. Source: Johannesberg Sunday Times, 11/4/84: 21.

The barriers to movement to a less privileged class were of course more permeable that those to passing "up." The language originally used to encode the classifications was itself inconsistent as well. Officials entering vital statistics in the pre-apartheid era frequently used the term "mixed." In many cases, this caused later confusion:

It was mentioned earlier that some White people, on sending for the first time for their birth certificate find that their racial group has been entered as "mixed," but that, on further investigation, this may be found to imply nothing more than that one parent was, for example, an immigrant from Sweden, while the other was an Afrikaans-speaking White woman. (Horrell, 1958: 73)

Again, we see here the conflation of prototypical ("mixed") categories with the attempted Aristotelian definition (the precise, exclusive categories aspired to by Nationalists). In this example, as with many in the book, the formal-informal mixture itself produces organizational conditions that favored both structural and face-to-face ad hoc discrimination, the one reinforcing the other. Star (1989a) describes a similar case with scientific anomalies, where anomalies arising in one sector of research may be answered by non-anomalous research drawn from another, thus obscuring the original difficulty. In this case, as with the conflation of prototypical and Aristotelian categories, biases become deeply embedded in both practice and infrastructure. The conflation gives a terrible power of ownership of both the formal and the informal to those in power. The use of both simultaneously is precisely at the heart of the banality of evil of which Arendt wrote. It differs from the mindless adhesion to formal rules found in Heller's Catch -22 (1961). The demoralization which it may produce is more like the Japanese-American Men of Company K of which Shibutani writes so eloquently -- a thorough cultural demoralization produced by unbearable role strain (1978).

The South African Coloured Population and Reclassification

Approximately 1,500,000 South Africans fell into the Coloured category at the time of the Population Registration and Group Areas Acts. It was among this group that the majority of borderline cases appeared -- most often in the form of a person labeled Coloured and desiring to be labeled White (or European). Within the internal logic of apartheid, an apparatus had to be constructed to adjudicate these cases. The Registration Act had a proviso that if the person objected to their classification, he or she had 30 days in which to appeal it. Several local administration boards were set up to hear borderline cases and reconsider classifications. Their decisions could be appealed up to the level of the Supreme Court, a costly and time-consuming business. The average waiting time for an appeal was 14 months (many were longer), during which time the person existed in limbo. For example, if someone wanted to be classified White, but was classified Coloured, she or he could not go to a White school. If they enrolled at a Coloured school, this could later become legal evidence that they were Coloured. Several took correspondence courses as a solution, as apartheid apparently did not hold long distance (interestingly, without a face-to-face component it was not enforced).

The havoc wreaked in the lives of those in between was considerable. A broadside issued by the South African Institute of Race Relations stated:

While the Population Registration Act of 1950 did not affect the circumstances of the vast majority of the South African population, it created the utmost confusion as to the destiny of the small minority of people whose appearance, associations, and descent do not happen to coincide. The South African Institute of Race Relations pleads with all the power at its command that this small number of persons should be allowed to remain in the racial category in which they feel most at ease (1969).

But this was not to be allowed by the Nationalist government. "As from 1 August 1966 it became compulsory for all citizens of the Republic over the age of 16 years to be able to produce identity cards, on which the racial group of the holder is. Until then, large numbers of people on the racial borderline had apparently not submitted themselves for classification." (Horrell, 1968: 23)

There were several factors involved in weaving the texture of categories in the lives of those in the borderlands of apartheid. Often there were long bureaucratic delays in assigning a racial classification to those who seemed ambiguous. The Associated Press cites a case in which two pre-school children were held in detention for three years while they awaited a government decision about their race (6 April 1984).

Approximately 100,000 people applied for reclassification (Brookes, 1968; Horrell, 1958). Few were approved. Bamford notes that: "The board would seem to have been over-strict against the subject -- the court has upheld its decision in only one of the ten reported cases involving the merits of reclassification. " (1967: 39) A typical year is shown below:

Figure 20. Numbers Of Objections To Racial Classifications Under The Population Registration Act, 1968.

 Total number of reclassifications  Of the total, number made as the result of representations by the person concerned
White to Coloured  9  1
Coloured to White  91  91
Coloured to Bantu  29  3
Bantu to Coloured  136  136

By May 1956 officials had dealt with 18469 cases "in which objection had been raised to the classification claimed by the person concerned. Of these 1182 had been classified as White, 9642 as Coloured, and 7645 as Bantu" (Brookes, 1968: 23).

Some years later, the figures had risen slightly but the basic direction of the changes remained the same. In 1981-82, 997 people changed races; in 1983, 690. In 1984, 795 people were re-classified. Of these, 518 went from Coloured to White; two Whites became Chinese and one became Indian; 89 Black Africans became Coloured, and 5 Coloured people became African (Ormond, 1986). A man from Durban won his reclassification appeal against his designation as Coloured when a judge declared that he was a "white of the Mediterranean type" (Horrell, 1968: 22)

The reclassification process was fraught in myriad ways, and was completely internally inconsistent. At first the Board ruled out descent (or "blood" as it was commonly called) as the determining factor. Instead, it used a mixed criteria of "appearance and general acceptance and repute." This was in explicit contrast with the American one-drop rule, (Davis, 1991) presumably for the reason that nearly all White South Africans had some traceable Black African ancestry (iv). Bamford, in an article in the South African Law Journal, attempts to clarify the juridical meanings of "appearance" and "general acceptance" (1967). He notes that:

Appearance is a matter of visual observation and assessment, to be undertaken by the tribunal. This observation and assessment should be made at the start of reclassification proceedings. If the subject is obviously white in appearance the presumption in section 19(1) will operate; if he is obviously not white, no further enquiry is necessary since he cannot be reclassified as white; and if he is neither obviously white nor non-white, the tribunal must proceed to decide on general acceptance.(p. 41)

There was no clear onus of proof about the meaning of general acceptance as White -- in ambiguous cases the Race Reclassification Board would decide after conducting hearings and administering a range of tests of race. Like child custody hearings in American courts, such painful (and often shameful) tests were not stable or guaranteed of permanence:

The concept of general acceptance does not preclude a person's movement from one classification to another by virtue of changing association. The acceptance need not be absolute or without exception, so that the fact that a subject maintains contract with relatives or remains friendly with a Coloured family is not in itself fatal. In such cases: '[The tribunal must] decide whether the non-white history and associations were so overshadowed by the acceptance as white as to constitute general acceptance of the [subject] as white.' (Bamford, 1967: 41).

Acceptance is an ambiguous, highly subjective prototypical concept sitting uneasily in the middle of the attempt for Aristotelian certainty. The New York Times reports the story of one Johannes Botha, a mail carrier living in Durban in 1960. Botha returned home to find his wife in tears following a visit from two investigators from the Group Areas Board whose mandate was to seek out those living illegally outside of their category. The sleuths had:

pounded on the door and then pushed past her and seated themselves in the living room without invitation. They had demanded identity cards, birth certificates and a marriage license and had asked whether her husband was white. Mrs. Botha told them to wait for her husband's return, but they persisted in the interrogation. (Bigart, 1960: 14)

The men questioned neighbors and the couple's 4-year old son, focusing on a visitor to the house on the two previous Sundays. A Coloured man had been seen on their doorstep. Botha, a scrap metal dealer, was interested in purchasing a used car from him. Neighbors had concluded that he was a relative, and thus that the Bothas were passing for White and hiding their "coloured blood." The network of suspicion, spying and the search for purity implied here affected every aspect of South African life, for those in all racial categories.

The actual process of reclassifying was usually done in camera. The procedure was kept highly secret by those at the Population Registration Office. "No observer is in any circumstances allowed to attend. Legal representation is permitted; but as an inquiry by the Board is not analogous to a law suit, the ordinary rules of court do not apply. The officials may ask any question they wish" (Horrell, 1958: 31).

Like the tuberculosis patients discussed in the previous chapter, people's biographical trajectories were severely disrupted by the reclassification process. Many lived for years in limbo; it could take months or years for the appeal to be heard and the person to be reclassified. It was as one of Roth's respondents in the previous chapter declared the case to be with waiting and negotiating for a tuberculosis classification: "an ungraded classroom." A time out of time:

Even if appeals succeed the people concerned have often suffered much anxiety and hardship before their cases are settled. In 1961 a family in Cape Town was classified Coloured. The eldest daughter had to postpone her marriage to a white man, while the second daughter had to leave a white school. For eight years she studied by correspondence: her parents did not want to prejudice their case by sending her to a Coloured school. The son's job was threatened. They all agreed to commit suicide if they could not get the classification altered. On reading their story in a newspaper, someone in Johannesburg made them an anonymous loan of R500 to cover the expense of an appeal, and this proved successful (Horrell, 1970: 27).

Once heard, the process was an open degradation ceremony (Goffman, 1959). That is, the process stripped people of identity, of uniqueness, and publicly humiliated them. "These officials question the people at the head of the queues and fill in forms -- there is no privacy as those behind can hear the questions and answers. On average, the investigation of one case appears to take about twelve minutes." (Horrell, 1958: 66)

Technologies of Classifying

How impracticable it is to try to classify human beings, for all time, into definite categories, and how much suffering has resulted from the efforts made to do this. (Horrell, 1958: 77)

Apart from the categories themselves, the technology associated with the reclassification process was crude. Combs were sometimes used to test how curly a person's hair was. Horrell (1968) notes that barbers were sometimes called as witnesses to testify about the texture of the person's hair.

Insert Figure 21. A Scale for Comparing the Color of Skin. 1958.
(sorry, unavailable online)

One source mentioned expert testimony from the South African Trichological Institute (presumably an organization for the scientific study of hair). Affidavits were taken from employers, clergy, neighbors and others in order to establish general acceptance or repute. "The official may summon any living relative, including grandparents, and question them in a similar way." (Horrell, 1958: 32) Complexion, eyes, hair, features and bone structure were examined by board officials, and they could summon any relative and examine them in this way as well. Horrell (1958) notes: "It is reported that some were even asked 'Do you eat porridge? Do you sleep on the floor or in a bed?' Some Coloured people said that they had been told to turn sideways so that the officials could study their profiles."(p. 62) Folk theories about race abounded - differences in cheekbones, even the notion that blacks have softer earlobes than whites was takes seriously. A newspaper account notes that some Coloured people had reported that "the officials fingered the lobes of their ears -- the theory is that Natives have soft lobes" (Sunday Times, 1955). The same article reported that a Coloured man was stopped by the police in the street and asked to which soccer club he belonged. He named a Coloured team, and then was told, "Only Natives play soccer, not Coloureds."

The "pencil test" was recounted by many who had undergone the reclassification ordeal. Sowden gives us the following passage, quoting at first from an old Black woman describing apartheid to him:

'If you're black and pretend you're Coloured, the police has the pencil test.'
'The pencil test?'
'Oh, yes, sir. They sticks a pencil in your hair and you has to bend down, and if your hair holds the pencil, that shows it's too woolly, too thick. You can't be Coloured with woolly hair like that. You got to stay black, you see.' (Sowden, 1968: 184).

Because of the ambiguous nature of both the notion of general appearance and of general acceptance, the burden of evidence fell on the person desiring to be reclassified. At the same time, the Population Board fostered the system of informers where someone trying to pass (typically as White or Coloured) could be turned in to the classification board for re-examination. Horrell notes the case of Mr. A, who was turned in to the Population Registration Office by an informer, and called before them to prove his Whiteness. Questioners asked if he knew of any Coloured blood in his family, and noted their hair, eyes and skin color.

Mr. A said: "It was a terrible shock to me, and more so to my sons. The whole future of my family know rests on a decision from Pretoria." The worst part, he added was that the very act of trying to prove himself European suggested that there might be some suspicion in the matter." (Horrell, 1958: 34)


They are both White and not White at the same time. They are in a White school and there they "must" be White: the law is witness to that. Yet "everybody" knows that they are not White, not really. They are something in between. But the law, which is an ass, knows no in-betweeness. It dichotomizes inflexibly, imposing a clumsy disjunction upon the subtly variegated flux of reality. (Watson, 1970: 114)

In the early pre-apartheid days, it was easier to change race category than it became later. Kahn notes that "Between 1911 and 1921 ... some fifty thousand individuals disappeared from the colored population rolls" (1966: 51). Many families living in the categorical borderlands went to great lengths to establish themselves as White, keeping photos of White ancestors (sometimes fabricated) (Boronstein, 1988: 55)

Under apartheid, merely associating with someone of the wrong group could become evidence of membership and thus of race. Horrell writes of the collusion between members of Coloured families being torn apart at the same time they collaborated to help some members pass for White:

There are in South Africa many thousands of people who cannot be classified according to a rigid system of racial identification. . . The lightest-coloured members of these ["border-line"] families often "passed "as whites and went to live in separate homes. Their darker relatives have been referred to as "Venster-Kykers" ["window lookers"] because, in order not to embarrass those who had "passed," they made a practice of looking studiously into shop windows in order to avoid greetings should they happen to meet on the streets. (1958: 4)

As noted above, someone's racial classification could be challenged at any time. This was particularly important to the apartheid government in the case of people trying to pass for White, and a crucial location for the operation of the system of informers. Doman notes:

Soon after the introduction of the legislation many people asked for reclassification, with the result that there are today many families split down the middle. The offspring of the 'across-the-line' marriages are not always as White as their parents, and many families have emigrated rather than risk exposure. Today, there is no concerted effort to unearth the skeleton in the family cupboard. Coloured mothers avoid embarrassing their 'White' daughters and do not see them even though they live in the same town. Yet the legislation also lends itself to spite - aggrieved people can get their own back on enemies or people they dislike by exposing a 'mixed' marriage, or informing the police about a couple having an immoral (in terms of law) relationship. (1975: 151)

In an extraordinary study of a school in the suburbs of Cape Town, Graham Watson (1970) wrote of the complex negotiations, subterfuge, and balancing acts performed by parents, students, school principals, and the local Race Classification boards in managing "pass whites". Cape Town is the area in South Africa with the largest population of Coloured people. Over the years, thousands chose to pass for White (or tried to). In order to do so, they changed their primary language from Afrikaans (used at the time by most Coloured people in Cape Town) to English. They changed their social affiliations, as noted by Horrell above. Some passed for White during working hours, and returned to live with their Coloured families in the evening.

In Passing for White: A Study of Racial Assimilation in a South African School, Watson drew a vivid picture of Colander High School based on his ethnographic participation. The high school was one of many buffer schools, which meant that "they admitted as pupils children described in the local School Board minutes as 'slightly Coloured' but colloquially known as 'borderline cases' " (Watson, 1970: 57). These schools predated apartheid, and the practice of accepting light-skinned Coloured children into them was well-established before 1948.

Many borderline students were admitted to Colander High through a process full of ad hoc decisions and negotiations spanning years. This was facilitated in the first instance by the contradictory and arms-length relationships between those parts of the government charged with classifying people. In order to attend a school, a pupil received a classification from the School Board. Before apartheid, the Education Ordinance of 1921 stated that parents must prove European heritage for the child to be defined as White. Prior to 1963, the Superintendent-General of Schools was not bound by the decision of the Director of Census and Statistics (who managed the Population Registration Act). Thus, it was possible for a child whose parents both carried White identification cards, but who her or himself was dark-skinned, to be rejected by a White school. Similarly, the Superintendent-General, via the school principals, could act to let a child into the nominally White school. Principals who were in doubt about the race of the pupil could request an interview with both of the parents and with the School Committee. Sometimes the parents and even grandparents were asked to produce birth and marriage certificates. The hearings, like those of the Race Classification Board, were held in camera. Watson notes that there was often disagreement among members of the School Committee about the classification decisions. Of the many cases heard by the Committee of Colander High School, twenty were rejected. Of these, 7 were rejected because one or both parents failed to appear; the rest were refused primarily on grounds of appearance (Watson, 1970: 43).

The number of appeals from these decisions were small, as the principals and committee members were loath to confront parents directly. Watson notes that parents were not actually told of the real reason for the child's rejection from the school, and often the blow was softened by simply saying, "we're full." He continues:

Moreover, the parents of rejectees, whether they have been summoned before the Committee or not, are not informed of the real reason for their rejection--- they are normally told simply that the school is full. Not even the School Board -- to whom the Principal has been instructed to disclose the reason for each refusal -- is told unequivocally that a child has been refused on the grounds of colour: in correspondence addressed to the Board the Principal covers himself by claiming that 'In the first instance, inability to accommodate is the reason for refusal.' In answer to verbal queries from the Board the Principal is reticent. 'He led me to the brink,' he said, recounting his response to such a query, 'but I wouldn't say it. I told him we were full up with thirty or forty in each classroom and we weren't prepared to take anyone until we got an increase of staff. Mind you, he must have taken one look at the boy and seen there was less milk than coffee and known perfectly well that wasn't my reason, but he couldn't say so.' (1970:44-45)

However, appeals to the Board about the Committee's decisions were very often successful. Thus, there was a delicate invisible negotiation between parents and school principals/school committees. If no real reason was given for rejection, there would have been no grounds for appeal to the Board. The principals were charged with keeping up appearances as a White school, or they would risk far more serious sanctions from the Population Board, as well as complaints from white parents. As one says, "I can accept that child...but what do I do when I have a school function and the rest of the family comes along?" (1970: 47). Presumably, he speaks here to the racism of the local White families. In addition: "For the Principal, however, the child represents a sinister threat to the White status of his school, to his ability to attract teachers and pupils of sufficient number and satisfactory quality, and, ultimately, to his own personal prestige." (1970: 48-49) Multiple convergent systems are operating here. One school psychologist had the belief that the school ratings were lowered by Coloured children as they performed more poorly on IQ tests. Such was the worry about this sort of status difficulty that principals often even rejected those darker-skinned children who carried formal White identification cards. On the other hand, the schools also needed to keep their numbers up, and some more liberal principals sometimes wanted to help the applicants. Overall, this juggling represents another example, as we saw with the ICD, of distributing the residual categories, the others. For those who don't quite fit the given categories, distribute them around the buffer schools, rather than having them all attend school at one place and thus threaten the White status of the school.

As mentioned, some officials willingly collaborated in the passing process. "It takes two (or more) to complete the process of passing for White" (Watson, 1970: 55). Watson writes of a zone of ambiguity in face-to-face decisions, "Is it incumbent upon me, in the circumstances, to decide whether or not this person is White? If I decide that he is White, will others go along with my estimation? And what's in it for me?" (Watson, 1970: 55-56) If the zone of ambiguity remains intact, often the overhead of trouble incurred by refusing someone the claimed White status is too much. At other times, officials find workarounds with which they can sidestep any confrontation. At one hospital, for example, a patient came in who was unconscious, and who looked neither European nor Coloured, but somewhere in between. The hospital authorities wavered on whether to put her in a European or a Coloured ward, and finally put her alone in a side ward -- in a living architecture of a residual category.

Watson (1970: 59) hypothesizes that those who successfully pass as White interact "segmentally with members of the superordinate group, thus allowing the superordinate-group members leeway in which innumerable ad hoc decisions cumulatively favourable to the aspirant can be made." The person wishing to pass for White manages a kind of shell-game sequence: first obtain employment in a Whites-only occupation that is not too fussy about identity cards (such as being a tram director). The next step is to move to a mixed neighborhood, and quietly join local White associations. Working with the fact that even racist Whites may find it difficult to confront a person face-to-face as passing, pass-whites are able to manage many face-to-face interactions such as attending White churches. Over time, this establishes a track record that can be used as leverage for reclassification based on general acceptance and repute. With possession of a White identity card, the person has the nominal protection of the law.

The more rigid the system of racial segregation and inequity, the more important passing became to those living in the categorical borderlands. At the same time, with the rise of the Black Consciousness movement in South Africa in the 1970s, a new ideology of Black unity across Black African, Asian and Coloured lines became powerful there. In an 1975 article, Unterhalter (1975: 61) notes that most Coloured people that she surveyed had disapproved of passing. This disapproval was based primarily on the need to remain loyal to the Coloured group. This contrasted with studies done twenty years earlier, where the harmful effect of passing on families was the primary reason (Watson, 1970: 61). Negative attitudes toward Black Africans who try to pass for Coloured remained unchanged.

Given the disparities in power and privilege, it is not surprising that so many Coloured people wanted to pass as White. Because passing is a partially secret, interactive process, and because it does require ad hoc mixtures of prototypical classifying and confrontations with Aristotelian categories of law, it is a crucible for the issues discussed throughout this book. Another kind of implosion (Haraway, 1997) occurs where people try to be reclassified, or who fall in other ways between the categorical imperatives of apartheid.

Reclassification and Borderlands

In one family, one twin was classified as Coloured and the other as African. (Horrell, 1958: 70)

The ground zero of race categories appeared in the case of an infant of indeterminate appearance who was abandoned on the steps of a hospital in Johannesburg. Because of the requirement of general acceptance and repute as determining one's race classification, it was ludicrous to try to use those criteria to decide the race of Lize Venter, named after the nurse who found her. An article in the Rand Daily Mail (7/10/83: 10) in 1983 announced; "Lize? A flat nose and wavy hair could decide her fate." It noted that "The law makes no provision for abandoned, newborn babies." Lize, more than any other person, represented the moment when the gaps created by the enforced mingling of prototype and Aristotelian category are laid bare, and the absurdities of apartheid law made clear.

The case of David Wong displays a more strategic exposure of the laws' internal contradictions. Wong was born in China of Chinese parents, and lived in a White neighborhood of Durban (a city with a large ethnically Asian population). His neighbors, taking advantage of the general acceptance clause of the Population Registration act, swore a series of affidavits stating that he was White. This was no doubt as well an anti-apartheid gesture, read over all. Wong received a White registration card on the basis of the affidavits. Brought to the attention of the press and government, it prompted an outraged reaction by MP deKlerk, who thundered:

It now appears, however, that there are certain White persons in this country who, again for reasons of their own, are prepared by means of affidavit to assist a person who admits that he is a full-blooded Chinese by descent, that he looks like a Chinese and who in appearance is obviously a Chinese, to be accepted as a White person by declaring an oath that he is accepted as a White person. This happened in spite of the indisputable fact that that was not the opinion of the community, and challenge any honorable member on the other side to take this Chinese, David Wong, out of the environment in Durban where he is living and to placing in any other environment in Cape Town or Pretoria or Johannesburg and to get the verdict of public opinion as to whether he is a white person. (DeKlerk, 1962: 10)

An amendment to the Population Registration Act was thereby passed, where anyone who claimed to be of a certain racial descent would be so categorized by the Board. The nested absurdities of the search for purity here are apparent -- yet the search for purity remained strong in the popular White racist opinion through the 1980s.

Language and Race as Conflicting Categories

There are thousands of ironic and tragic cases where classification and reclassification separated families, disrupted biographies, and damaged individuals beyond repair. The rigid boxes of race disregarded, among other things, important linguistic differences, especially between African tribal languages. Let us review a few of the more extreme borderline cases. Collectively, they provide a powerful ethical argument against simple-minded pure-type categories, and for the positive value of ambiguity and complexity when applying racial categories to human beings.

The filiations of appearance and linguistic group become tangled in the case of "Dottie," a girl born to Black African parents in the Randfontein area. She "happened to be lighter-skinned than are most Africans and to have long, wavy, copper-coloured hair. Because of this she was rejected by principals of African schools, and cannot attend a Coloured school because she can speak only Sotho" (Horrell, 1968: 21). A similarly cruel situation appeared in the case of the Griqua group. This group has a distinctive physical appearance, with "yellowish skin, high cheekbones, hair growing in little curly clusters" (Horrell, 1958: 53-54). Many of this group married other native African tribal groups. They were classed by the Population Registration Act as African. This meant that they would be ruled in terms of education by the Bantu Education Act, and thus educated in one of the indigenous African languages. This was "completely foreign to most of the Griquas, who speak Afrikaans" (Horrell, 1958: 53-54).

There were layers of invisibility being enacted here by pro-apartheid forces. The idea of a separate development required that Black people fit into mythic categories of pure tribal groups. The basis on which these groups were established and reported only partly respected actual tribal affiliations and not at all the conditions of people's lives. The hypothetical types were adorned with many natural features such as language and customs. In turn, each hyper-prototypical tribal group must have its own language, its own land, and its own unique customs. There was no room for people or circumstances who did not fit this image.

Again there is resonance with the ways in which Americans have enacted race in different regions. There are thousands of Native American tribes, and most of them have been displaced from their indigenous lands. Unknown numbers have intermarried outside of and across tribal lines. Yet the registration system of the US Government/Bureau of Indian Affairs for counting who is really a member of a recognized tribe (and thus deserving of government benefits) is as contorted as tribal counting under apartheid. Munson (1997), for example, writes of the laws in the state of New Mexico which seek to protect Native American artisans from non-Indian imitators claiming to be "genuine Indian art." The laws have the ironic effect that a legitimately registered member of a tribe from, say, Wisconsin, could come to the area, have no prior knowledge of local tradition, but legally sell "genuine" Indian artifacts. At the same time, a local from an unrecognized tribe, having lived in New Mexico all his or her life, would not be able to do so. In an imposed, purified system of categories, both under apartheid and elsewhere, there are many ironies and much individual suffering.

Sudden Changes

Another sort of ironic twisting of the categorical landscape leading to acute torques occurred when race classification was suddenly, unexpectedly shifted. For example, Ronnie van der Walt was a famous boxer in South Africa. At the age of 29, he was suddenly reclassified from White to Coloured, on the eve of a big match. One presumes from the Newsweek article reporting the case that someone had informed the race classification board, and they timed their inspection to be maximally embarrassing, an object lesson for others. The local race classification board's decision "was based on an inspection of Ronnie, Rachel and their two children. . . One man there, Ronnie recalls, walked around us peering at us from every angle like you do when you buy an animal. He said nothing, just looked. . . Interior Minister P.M.K. Leroux insisted that the ruling on Ronnie would stand. 'He has never been a White person,' sniffed Le Roux. Then, with logic reminiscent of the Mad Hatter the minister added 'And I do not believe he will ever become one'" (Newsweek, 2/27/67: 42).

Van der Walt's biography and career were suddenly bisected by the revision of his race classification. Other cases were reported which illustrated the precarious nature of race purity. Two White children, Jane-Anne Pepler and Johanna de Bruin, had severe malfunctions of the adrenal glands, which caused their skin to turn brown. Jane-Anne had an operation to remove the glands at the age of fifteen; in a short period of time her skin and hair went from fair to dark brown. Her mother reported that: "only close friends and family who knew her before the operation know she is white" (Newsweek, 7/3/70: 31). In a statement rich with an unconscious, ironic pragmatism, her mother said, "Some of her school friends have ostracized her completely -- just as though she were a real non-white" (Newsweek, 7/3/70: 30). She noted that: "All this is particularly embarrassing for us because we are a purely Afrikaner family and strong Nationalists. We believe in white supremacy " (Newsweek, 7/3/70: 31).

Johanna was an infant when she contracted Addison's disease. "No white school would accept her when she reached school-going age. Her father told a reporter that he intended applying to the Education Department for a tutor to teach her at home until she had passed Standard V, after which she would be able to take correspondence courses. " (Horrell, 1970: 26) Her mother "lives in constant fear that, because of the past difficulties, 'someone' will come and take her daughter away from her" (Wannenburgh, 1969). In both cases, the children are stuck with the rigidity of the Aristotelian definition of race -- both were born with parents with White identity cards and were thus White -- tempered with the prototypical face-to-face judgments of skin color, which would render them Coloured.

 Ten-year-old Sandra Laing slipped unnoticed into the school cloakroom. She made sure she was alone, then picked up a can of white scouring powder and hastily sprinkled her face, arms and hands. Remembering the teasing she had just endured in the schoolyard during recess, she began scrubbing vigorously, trying to wash off the natural brown color of her skin. (Ebony, 1968: 85)

Perhaps the most famous case of sudden identity change was that of Sandra Laing, who was brought up by White parents and was evicted at the age of ten from her White school for being Coloured. The United Nations reported that when she was expelled by school officials under the Population Register Act, it became illegal for Sandra to attend her Piet Retief boarding school, which was all-White (United Nations Office of Public Information, 1969: 4). This reclassification of course denied her access to all other White institutions. The only way Sandra could continue living with her family was by being registered as a servant.

Scientists explained Sandra's appearance as resulting from a "dormant 'throwback ' gene". It was posited that among the six color-determining genes, this throwback was responsible for Sandra's coloring (Ebony, 1968: 85). Sandra's parents rejected the idea that this made her Coloured, however. Sandra's father stated that she had been brought up "naturally" as a White child (Ebony, 1968: 90). They attempted to tutor her at home, while appealing her case up to the level of the Supreme Court. After two years, she was reclassified again as White, and was legally permitted entry back into White schools. Ebony continues:

Out of the fire and back into the frying pan. That's what it all amounted to, because it had been Sandra's appearance, not her legal classification, that had aroused the bigotry of white parents two years before. And they still didn't want their children in school with a dark child. Now matter what the statute books ruled, Sandra was still a kaffir(v) to them (1968: 88) (vi)

The informal categories of racism and the formal classification system meet once again, this time tearing Sandra Laing's biography apart. Sandra's parents, clinging to the formal definition of her race, refused to tell her exactly what was going on or why she was so treated. For two years she "acted out fantasy rather than face the bitter truth. Until recently, she dutifully got up in time for school every morning since her dismissal in March of 1966, dressed herself in her school uniform, then sat around the house and waited, trying not to believe what was happening to her" (Ebony, June 1968: 86).

In 1983 the newspaper, Rand Daily Mail reported that Sandra had become completely alienated from her family and community. She "eventually lived with a Black man and, ironically, applied to be reclassified so she could live legally with her lover" (Rand Daily Mail, 7/23/83: 10).

Hope's novel, A Separate Development, centers on a dark-skinned boy who grew up White and who suddenly, upon reaching adolescence, becomes defined as Coloured. A bus conductor throws him off a White bus, calling him a "white kaffir." He says bitterly:

The thing is that this entire country has always based itself on two propositions, to wit: that the people in South Africa are divided into separate groups according to their racial characteristics and that all groups are at war with each other. Before you're clear about your groups, you must be sure you're clear about your individuals. As they teach the kids to chant: 'An impure group is a powerless group! ' . . . Preserve the bloodlines. That was the rallying cry for generations. May your skin-tones match the great colour chart in the sky. Anyone who broke the bloodlines, who wasn't on the chart, was a danger to the regular order of things. You fought such renegades, mutants, throw-backs and freaks with the power of definition. When in doubt, define. Once defined, the enemy could be classified, registered and consigned to one of the official, separate racial groups which give this country its uniquely rich texture. 'White kaffir': the words have a ring to them. I came to be grateful for them. Up until then I hadn't any proper idea what I was. What the conductor gave me was an identity. Ever since, I've been an identity in search of a group. (1980: 28)

Those who live in the borderlands, as Sandy Stone argues about gender order in her "The Empire Strikes Back: A Transsexual Manifesto," illuminate a larger architecture of social order (1991). Transsexuals, those who cross over from male to female or vice-versa, become in her metaphor a blank piece of paper upon which may be written anyone's fantasy of what a perfect women (or man) should be. Stone herself, a male-to-female transsexual, was initially refused surgery for refusing to wear makeup and high heels and behave "like a woman." Emily Ignacio, writing about Filipino/a identity in diaspora, notes a similar struggle -- this both with and against stereotypes of a "real Filipino" (1998). Wherever ethnic identity exists, such struggles for and against purity exist. South Africa's Nationalists tried to classify in a completely Aristotelian fashion, and to make racial borderlands and ambiguity impossible. In so doing, the texture of the filiations they created were knotted, twisted and often torn: another nightmarish texture (Star, 1991).

Earlier, we used the term torque to describe the twisting which occurs when a formal classification system is mismatched with an individual's biographical trajectory, memberships, or location. We have now probed more deeply into how this unfolds -- the prototypical and Aristotelian are conflated, leaving room for either to be invoked in any given scenario (especially by those in power). The South African case represents an extreme example. For those caught in the racial reclassification system, it constituted an object lesson in the problematics of classifying individuals into life determining boxes, outside of their control, tightly coupled with their every movement and in an ecology of increasingly densely classified activities. Each borderland case became a projection screen for the stereotypical fantasies of those enforcing the borders themselves. The stories of Sandra Laing, Dottie, and Ronnie van der Walt are ones that help illuminate what can happen when such a classification system is enforced and policed.

In some ways the South African stance is a mirror image of the current American dilemma. In the mid 1990s, a group of Americans held a march on Washington, with the goal of having the the option of choosing multiple racial categories added to the US Census.(vii) This would replace the vague, and to some insulting, "other" category. They argued on both scientific and moral grounds that multiracial was the appropriate designation, one which would not force individuals to choose between parts of themselves. Yet, many civil rights groups vigorously opposed them. Robbin (1998) recounts the struggle over the decision taken by the Office of Management and Budget (OMB) in 1997 to allow people to choose more than one racial category on the US Census and in other official US Federal Government forms. This decision was known by the innocuous name of Statistical Directive 15. Arguments over the nature of racial classification in the US Census go back over many decades but with the advent of Affirmative Action and other similar measures in the late 1960s and early 1970s, race classification became even more consequential -- and contested. Robbin identifies three major issues involved in arguing for or against changing the categories -- the controversy about whether or not to name (and how to name) racial and ethnic groups in government data; the exclusion of minority populations from the decision-making process within the Census bureau; and the difficult questions of data quality and measurement. For example, where geographical location and race have been confounded, as with the Hispanic-American groups, people have in the past often simply opted to identify themselves as "other". The size of this residual category, and the potential sampling errors it represents, remains an unsolved difficulty in collecting Hispanic-American census data (pp. 43-46 of MS).

In the controversy surrounding the OMB decision, many African-American leaders (among others) argued that if everyone went by the scientific classification and coded themselves as belonging to multiple races, valuable demographic information and resources could be lost for many African Americans (Frisby, 1995-96). Regardless of the scientific or genetic basis for the category, they said, racism against people with any Black African ancestry was real, and the category was necessary in order to obtain resources and justice. This stance, sometimes called strategic essentialism by critical race theorists, lives precisely at the Pragmatic junction between that which is perceived as real, and the consequences of that perception. Other leaders approved of the category choice change, seeing it as potentially liberating. Keen debate about the nature of these categories continued for months, including issues such as whether the indigenous people of Hawaii should be grouped with Native Americans, and how to categorize people of more than one national and ethnic heritage. In October, 1997, OMB decided to allow people to identify themselves as of more than one race, that is, to check more than one box. They could not, however, identify themselves as multiracial. The enormous expense and inertia of the decision is striking. It is estimated that it will cost millions of dollars. It is not often that individual categories are championed as social movements; even more rarely does an entire schema come under scrutiny.


The South African case relates directly to all questions of information systems design where categories are attached to people. It is an extreme case -- but at the same time, a valuable one for thinking about the ethics and politics of information systems. Not all systems attempt to classify people as globally, or as consequentially, as did apartheid; yet many classify users by age, location, or expertise. Many are used to build up subtle (and not-so-subtle) profiles of individuals based on their filiations to a myriad of categories. In the process of making people and categories converge, there can be tremendous torque of individual biographies. The advantaged are those whose place in a set of classification systems (of race etc) is a powerful one and for whom powerful sets of classifications of knowledge (so, for example, the ICD or the world as constituted of philosophy, politics and economics) appear natural. For these people the infrastructures that together support and construct their identities operate particularly smoothly (though never of course fully so). For others, the fitting process of being able to use the infrastructures takes a terrible toll. In order to 'act naturally' they have to reclassify and be reclassified socially.

i We shall at this point, following Dubow (1995), stop putting quotation marks around words such as race or coloured. We trust the reader to recognize that the entire argument here opposes any essentialist or simplistic interpretation of these terms, or acceptance of racist constructions! As well, except in direct quotes, we adapt the terms Black, White, and Coloured in capital letters. The South African usage was not standardized.

ii "They will count us. My friend was checked yesterday. Count us, count us!!" (authors' translation from the Afrikaans)

iii An early anti-apartheid organization noted for greeting officials at airports and the like wearing black sashes of protest.

iv Although DeKlerk in 1962 attributes this to purely technical reasons, "In order to use descent as a test it would have meant digging far back into the past for proof, and the moment on has to start digging into the past one becomes lost in a labyrinth. . . in earlier years there was no reliable record of many of these facts. For example, in some of the provinces facts of this kind this have only been noted since 1915 in registrations of births and deaths. In other words one cannot trace the origin and the race of the person." (De Klerk, 1962: 9)

v Kaffir is a rude word equivalent to nigger.

vi Unsurprisingly, the school officials were of little help. "The Mother Superior of a convent explained: "If could have had my way I would have admitted the little girl. But we depend on public goodwill, and as I see it we would only have trouble if we admitted her. We have to consider the feelings of our parents and children." (Ebony, June 1968: 88).

vii See . "This classification provides a minimum standard for maintaining, collecting, and presenting data on race and ethnicity for all Federal reporting purposes. The categories in this classification are social-political constructs and should not be interpreted as being scientific or anthropological in nature. They are not to be used as determinants of eligibility for participation in any Federal program. The standards have been developed to provide a common language for uniformity and comparability in the collection and use of data on race and ethnicity by Federal agencies. The standards have five categories for data on race: American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White. There are two categories for data on ethnicity: "Hispanic or Latino," and "Not Hispanic or Latino."

1. Categories and Definitions

The minimum categories for data on race and ethnicity for Federal statistics, program administrative reporting, and civil rights compliance reporting are defined as follows:

-- American Indian or Alaska Native. A person having origins in any of the original peoples of North and South America (including Central America), and who maintains tribal affiliation or community attachment.

-- Asian. A person having origins in any of the original peoples of the Far East, Southeast Asia, or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam.

-- Black or African American. A person having origins in any of the black racial groups of Africa. Terms such as "Haitian" or "Negro" can be used in addition to "Black or African American."

-- Hispanic or Latino. A person of Cuban, Mexican, Puerto Rican, Cuban, South or Central American, or other Spanish culture or origin, regardless of race. The term, "Spanish origin," can be used in addition to "Hispanic or Latino."

-- Native Hawaiian or Other Pacific Islander. A person having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands.

-- White. A person having origins in any of the original peoples of Europe, the Middle East, or North Africa.
Respondents shall be offered the option of selecting one or more racial designations. Recommended forms for the instruction accompanying the multiple response question are "Mark one or more" and "Select one or more."

2. Data Formats

The standards provide two formats that may be used for data on race and ethnicity. Self-reporting or self-identification using two separate questions is the preferred method for collecting data on race and ethnicity. In situations where self-reporting is not practicable or feasible, the combined format may be used.

In no case shall the provisions of the standards be construed to limit the collection of data to the categories described above. The collection of greater detail is encouraged; however, any collection that uses more detail shall be organized in such a way that the additional categories can be aggregated into these minimum categories for data on race and ethnicity.

With respect to tabulation, the procedures used by Federal agencies shall result in the production of as much detailed information on race and ethnicity as possible. However, Federal agencies shall not present data on detailed categories if doing so would compromise data quality or confidentiality standards.

a. Two-question format
To provide flexibility and ensure data quality, separate questions shall be used wherever feasible for reporting race and ethnicity. When race and ethnicity are collected separately, ethnicity shall be collected first. If race and ethnicity are collected separately, the minimum designations are:

-- American Indian or Alaska Native
-- Asian
-- Black or African American
-- Native Hawaiian or Other Pacific Islander
-- White

-- Hispanic or Latino
-- Not Hispanic or Latino
When data on race and ethnicity are collected separately, provision shall be made to report the number of respondents in each racial category who are Hispanic or Latino."