SciELO - Scientific Electronic Library Online

vol.104 issue11-12Latest information on Sterkfontein's Australopithecus skeleton and a new look at AustralopithecusSurface polarity determination of wood fibres after different pre- treatments and bisulphite pulping author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


South African Journal of Science

On-line version ISSN 1996-7489
Print version ISSN 0038-2353

S. Afr. j. sci. vol.104 n.11-12 Pretoria Nov./Dec. 2008




What can we learn about the meaning of race from the classification of population groups during apartheid?



Yvonne ErasmusI, *; George T.H. EllisonII

ICentre for Critical Research on Race and Identity, George Campbell Building, Howard College Campus, University of KwaZulu-Natal, Durban 4041, South Africa
IISt George's, University of London, Cranmer Terrace, London SW17 0RE, U.K.




This paper examines what might be learnt about the meaning of race from the formalization of racial classification and reclassification under apartheid, generated by the 1950 Population Registration Act. It draws on 69 (re)classification appeals heard by the South African Supreme Court between 1950 and 1991, and in-depth interviews with a civil servant, expert witness and scientist involved in the (re)classification process. The Supreme Court data indicate that the three classificatory criteria set out in the act (appearance, descent, and acceptance) were ambiguous and subject to substantial debate and reinterpretation by the courts, which principally relied on acceptance. This is supported by the interviewees, who lamented the 'unscientific' classifications they were obliged to perform, yet accepted these as the inevitable consequence of the role social practices play in determining and accepting the classification applied. These findings suggest that there was not a single concrete definition of race during apartheid. Instead, race was whatever people understood or wanted it to be, and racial classification could be attained through 'performing' an identity with sufficient proficiency to 'get away with it'. This provides a crucial insight into the meaning of race elsewhere—as simply a flexible, yet pragmatic and 'acceptable' social classification of group identity.




Notwithstanding the ongoing utility of race as an analytical concept in genetics and biomedical research, the resilience and contemporary salience of racial categories largely reflect the way in which our bodies are 'categorized and valued according to socially constructed rules about privilege, power and domination'.1 Nowhere was this more self-evident than in apartheid South Africa—a context which provided perhaps the most powerful contemporary example of how racial categories tend to be embedded in power relations and discriminatory practices. However, while there is substantial consensus about the consequences of racial classification, not least in South Africa but also elsewhere, the meaning of race seems to remain elusive wherever it occurs.2 The aim of this analysis was therefore to examine what more might be learnt about the meaning of race by studying the process of racial classification within a context in which this was formalized through legal statutes—apartheid South Africa.

The legislative cornerstone of racial classification during apartheid was the Population Registration Act (No. 30 of 1950). This act divided the population into three main 'race' or population groups, namely 'white', 'coloured' and 'native',3 and in 1959 subdivided the coloured group into seven subcategories.4 The act also laid down the framework through which appeals against racial (re)classification (whether by the Director of Census, or subsequently, the Secretary for the Interior) could be made, in the first instance to a Race Classification Appeal Board, and thereafter to the South African Supreme Court.3

To assess what might be learnt about the meaning of race from the procedures required to sustain racial classification during apartheid, the analyses that follow explore how the Supreme Court interpreted the legal definitions and classificatory criteria set out in the act, and how classifiers and experts working both inside and outside government made their classificatory judgments.



Data for this analysis were drawn from two sources. The first comprises the judgments of 69 Supreme Court cases which dealt with race (re)classification between 1950 and 1991. These cases were published in the South African Law Reports—a legal journal that has appeared monthly since 1947 and contains, amongst other things, the judgments of all precedent-setting cases heard at the Local, Provincial and Appellate divisions of the South African Supreme Court.5 Cases dealing with racial (re)classification were located by hand- searching and cross-referencing four of the Law Reports' different indexes using a range of search terms, in both English and Afrikaans, such as: 'Population Registration Act, No. 30 of 1950'; 'Secretary for the Interior'; and 'Race Classification Appeal Board'.

The second source of data comprises three in-depth interviews with individuals who were directly involved in classifying others: a civil servant who had worked as classifier in a local office of the Department of the Interior for a short period and in an office two doors down from the classification sub-department for a number of years; another who had acted as an 'expert witness' in appeals against race (re)classification both at Race Classification Appeal Boards and Supreme Court hearings; and a third who had offered professional scientific advice to individuals who were concerned about their racial classification and an orphanage that needed to classify babies of mixed or unknown parentage. The first of these interviewees was traced from a newspaper article on reclassification. The second and third interviewees were identified by networking with academics working on race/ethnicity in South Africa. Consent was obtained from each of the interviewees to tape record their interviews for subsequent transcription and analysis, on the basis that their experiences would be analysed confidentially and presented anonymously.



Race through the eyes of the law and the courts

Although the act laid down three criteria for classifying race—appearance, descent, and acceptance—the judgments of the 69 Supreme Court cases examined in this study demonstrate that all of these criteria were ambiguous, complex and difficult to apply. In particular, appearance was viewed as 'deceptive'.6 Its ambiguity was evident in the range of different characteristics involved (including anatomy, clothing and language) and in a number of disagreements between the Race Classification Appeal Board and the Supreme Court regarding an appellant's 'appearance'. Nonetheless, appearance was primarily used as a precursor to descent or acceptance—identifying when these other criteria were relevant and how they should be applied and interpreted. And although using appearance in this way (to tacitly screen appellants prior to the application of descent or acceptance) is likely to have reduced inherent contradictions between the different criteria, it did not entirely solve the inherent ambiguity of descent or acceptance. Indeed, when applying descent to determine race there was substantial confusion, not least about the relative merits of the terms 'full-blooded descent' and 'a preponderance of blood' (and what constituted a 'preponderance'), for example:

What preponderance would be required, two-thirds, three-quarters, or seven-eights of full-blood ?7

Likewise, the test of acceptance could be interpreted in a range of different ways. On the one hand the Race Classification Appeal Board sought to compartmentalize people's lives into different and very specific domains, in each of which it was necessary to provide evidence of acceptance by other members of the racial group concerned. In contrast, the Supreme Court judges seemed to take a more balanced view of each appellant's life, ignoring domains that were irrelevant or relatively unimportant in their particular case:

The true approach seems to me to be to view all the various factors pertaining to each individual including the various facets of his life, some of which may be important and others unimportant, and then having regard as a whole to the picture of acceptance by others, decide whether or not he is generally accepted as white…8

This more individualistic approach to acceptance meant that, in the absence of evidence of non-acceptance, Supreme Court judges were prepared to accept evidence of a more general sense in a smaller subset of domains as sufficient grounds for acceptance.

Administering race: the scientist, the expert witness and the civil servant

It was therefore within this largely ambiguous and somewhat contested legal framework that the classifiers and experts interviewed had needed to make their classificatory judgments. Interviewees alluded to two separate, yet related, influences on their classificatory decisions: science and society. Perhaps the most startling finding was the lack of explicit references to science. Indeed, in all 69 court cases, science was referred to only three times, and the scientist consulted by clients who were concerned about their racial classification confirmed that the courts refused to admit his genetic reports as evidence. He considered that:

…the judges probably didn't want to get involved in resorting to simple scientific evidence; they were going on appearance and acceptance and descent…. I think [they] quite rightly, knew that it was going to get very messy in cases of multiple race contributions.

This apparent reluctance of the Supreme Court judges to accept or invoke scientific evidence to support their judgments suggests that they thought a more 'scientific' approach to racial classification was not possible. Nonetheless, for the interviewees, the key principles of scientific practice (reliability and validity) were implicit in the way they described the need for consistency and accuracy of measurement. For example, the scientist described how he

would go through the steps of looking at the criteria that are used to define race. So I would look at the colour in various places … the face … then I would look at certain areas of the body.

Elsewhere, in one of his case files, a colleague had commented that

…there are a number of [genetic] markers which are incontrovertible indicators of one kind of descent or another which are socially meaningless….

Despite a desire for quasi-scientific rigour, the interviewees gave the impression that what they were doing was not felt to be scientific. In this regard the scientist described how 'We sure did make a mockery of it', and cited an American colleague who was also involved with classifying race, who used to say: 'This is not science; this is perverted sociology.'

However, just because the interviewees recognized that what they were doing was not strictly 'science', this did not mean that they felt there could or should be no role for science in racial classification. Indeed, the expert witness felt that the sorts of genetic markers cited by one of the scientist's colleagues would have helped improve the reliability of (re)classification:

…if a specific genetic marker (or even a few) was used as definitive in classification, then there could [have been] consistency….

Nonetheless, what is clear is that even though some of the interviewees thought that science might have been able to shed some light on classification, this was not the approach that they had explicitly adopted. Instead, the data suggest that there was substantial reluctance on the part of the interviewees to apply strict scientific criteria where this was at odds with a desire to do the best they could for the people concerned. As such, the scientist described his approach to classification as

…very pragmatic… to help them live as successfully as possible in whatever category they wanted to be situated....

and he considered that:

…our conscience should dictate what we reported….

Clearly, the classifiers and experts were not free from personal or social influences, not least because they sought to avoid providing any unintentional support for apartheid practices. In this regard the scientist felt that:

If you have scientists who have these hang-ups that we see very clearly in politicians perhaps, then they may well be doing things, almost subconsciously… certainly unintentionally, which reinforces political thinking.

Moreover, the interviewees were fully aware of the social nature of race: for example, one of the scientist's colleagues had added a comment to one of his case files to the effect that 'classification [is] a social rather than a biological matter', and the scientist himself described what he had tried to do as '…trying to help in a situation which was man made – that was Nationalist Party made, and society made it long before the Nationalist Party…'.

In this sense the interviewees acknowledged that the tests used to classify people were socially constructed and essentially subjective. Thus, when discussing how the law provided substantial leeway for the development of classificatory tests, the scientist described how '…a lot of it was obviously open to subjective interpretation'. Likewise, the civil servant described how, when she had asked her boss how she should classify people, she was told: 'Oh, you've got eyes'.

Common-sense, pragmatism and the social acceptability of racial classification

The overriding importance of acceptance over appearance and descent in the judgments made by Supreme Court judges, together with the inherent subjectivity of the classificatory tests applied by classifiers and experts, do provide compelling evidence that apartheid's racial categories were explicitly social constructs and were widely recognized as such. However, the interviewees also provided some unexpected insights into the broader role that society played in the way these categories and classificatory processes operated. In particular, society played a key role in accepting or rejecting classificatory decisions, and thereby actively participated in decisions about which members of society fitted into which racial categories. For example, the civil servant considered that the third-party appeals enshrined in the legislation (where a 'third party' could approach the Department of the Interior to question the classification of another person) meant that racial classification was subject to the approval of neighbours and colleagues and was continually subject to social surveillance. In a similar fashion, the civil servant described how officials from the Department of the Interior would visit schools and "would sit and watch the children during play breaks... The questions that they asked were, 'Were they accepted by their peers? Were they shunned?' " Likewise, the issue of acceptability was fundamental to a case recounted by the scientist concerning a baby that had initially been classified as white but had subsequently been brought back to the adoption agency because it was apparently not acceptable to the white family that had adopted the child.


Discussion and conclusion

These analyses demonstrate the multiplicity of meanings that can be assigned to the concept of race even within the context of apartheid South Africa, where racial categories and racial classification were enshrined within legal statute. Indeed, the varied interpretations offered by the courts concerning the meaning and relevance of appearance, descent and acceptance, coupled with the different approaches to classification taken by the interviewees, discredits any notion that a single working definition of race existed. As such, these findings reflect both the elusiveness of race as a concept, and the difficulty of establishing criteria capable of measuring race consistently and accurately.

Taylor9 suggests that race, like language, is 'intuitive, or practical: using [it] involves knowing how more than knowing that.' Perhaps race functioned in a similar manner during apartheid? Whether through 'commonsense'10 or intuition, it seems likely that most South Africans found it relatively easy to classify one another on a day-to-day basis, but as soon as attempts were made to explain or formalize the underlying rules involved, their arbitrariness, elusiveness, complexity, and ambiguity became all too apparent.

In the various contexts in which racial (re)classification took place (in the courts, by classifiers and by experts), it seems clear that the absence of scientific method was matched by a reliance on, and general acceptance of, social construction. For the most part, the only role played by science was implicit—being evident in the desire for consistency and accuracy (that is, reliability and validity). This seems extraordinary given the role of scientific positivism in the development of racial hypotheses from the mid-19th century to the present day.2 However, as Dubow11 found in his analysis of racial science in South Africa, the architects of apartheid seemed wary of drawing too closely on the 'science' of classification and measurement because they were well aware that this was fallible and would not sustain the more nuanced social categories apartheid sought to reify and apply. Thus, while the absence of explicit references to science might largely reflect the specific practices of the courts and interviewees, it also fits with Posel's12 view that during apartheid race was seen as a 'judgement of social standing made on the strength of prevailing social conventions about difference' rather than an assessment based on any established tenets of 'racial science'. In this sense the most important mechanism through which racial identities were socially negotiated and determined was through what Stone13 calls 'review', in which for any identity to 'achieve significance', it 'must match the review of others'—very much as racial identity was subject to the test of social acceptance during apartheid, as applied by the Supreme Court and in everyday life.

These findings may not seem terribly surprising to those who already dismiss race as a social construct. Indeed, it has recently become a sign of 'intellectual sophistication', as Zack14 described it, to claim that there is no scientific and biological support for the concept of race, and thereafter to 'move on to more pressing matters, as though that matter were not the most pressing'. However, it does seem extraordinary that the true social nature of race becomes so obvious within a context in which the classification of race was legally sanctioned and ostensibly formalized. Despite its enthusiasm for defining race and the criteria it established for operationalizing racial classification, the National party failed to pin down a workable definition of race. Yet, by exploring this failure, the 'true' meaning of race emerges as whatever people expected, wanted or needed it to be. In essence then, racial identity in apartheid South Africa was defined by, and subject to, whatever everyone understood race to be. As such, racial identity was partly determined by whatever it was possible to 'get away with' when 'performing' an identity with sufficient proficiency to be accepted as such. These analyses therefore offer insight into what race might also mean elsewhere—simply a flexible, yet pragmatic and acceptable social classification of identity.

This paper was first presented in the Social Science plenary session of the 4th annual conference of the Africa Genome Education Institute held at Somerset West, South Africa, on 22–24 March 2006. The conference was sponsored by: the Charles Stewart Mott Foundation, Innogen, Media 24, the Nelson Mandela Foundation, Sanlam, the Department of Trade and Industry, the Department of Science and Technology, the African Medical Research Council, and the Wellcome Trust.


1. Dalmage H.M. (2004). In Skin/Deep: How race and complexion matter in the 'color-blind' era, chap. 5, p. 83. University of Illinois Press, Urbana and Chicago.        [ Links ]

2. Malik K. (1996). The Meaning of Race: Race, history and culture in Western society. Macmillan Press, London.        [ Links ]

3. Union of South Africa (1950). Population Registration Act, No. 30 of 1950, in SA Government Gazette.        [ Links ]

4. Union of South Africa (1959). Proclamation No. 46 of 1959.        [ Links ]

5. Du Plessis J.R. and Kok L. (1989). In An Elementary Introduction to the Study of South African Law, 2nd edn, p. 44. Juta, Cape Town.        [ Links ]

6. Goliath v. Director of Census and Another (1956[3]), p. 101, S.A. 95–102 (T).        [ Links ]

7. Mabitle v. Secretary for the Interior (1968(1]), pp. 31–32, S.A. 29–32 (C).        [ Links ]

8. Francisco v. Secretary for the Interior and Another (1967[2]), pp. 287-288, S.A. 283–291 (C).        [ Links ]

9. Taylor P.C. (2004). In Race: A philosophical introduction, chap. 1, pp. 6-7. Polity, Oxford.        [ Links ]

10. Posel D. (2001). Race and common sense: racial classification in twentieth century South Africa. Afr. Stud. Rev. 44(2), 87–113.        [ Links ]

11. Dubow S. (1995). Illicit Union. Scientific racism in modern South Africa. Cambridge University Press, Cambridge.        [ Links ]

12. Posel D. (2001). What's in a name? Racial categorizations under apartheid and their afterlife. Transformation, pp. 53. 47, 50–73.        [ Links ]

13. Stone G. (1962). Appearance and the self. In Rose A.M., Human behaviour and the social process cited in Rockquemore K.A. and Brunsma D.L. (2004). In Skin/Deep: How race and complexion matter in the 'color-blind' era, chap. 6, pp. 104–105. University of Illinois Press, Urbana and Chicago.        [ Links ]

14. Zack N. 2002. In Philosophy of Science and Race, preface, p. xi. Routledge, New York and London.        [ Links ]



* Author for correspondence. E-mail:

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License