Ethnicity in Post-Apartheid Discourse: Deconstructing South Africa’s Racial Narrative by Joshua Richman

This article was written by Joshua Richman and forms part of the South African History Online and the University of York Collaborative Project.

Ethnicity in Post-Apartheid Discourse: Deconstructing South Africa’s Racial Narrative


This paper addresses the falsity of the cultural discourse that stresses ethnicity’s salience in post-apartheid South African conflict. It challenges constructivist research and its tendency to be satisfied with simply identifying ethnicity as a social construct. While this interpretation is correct, such readings often continue to approach ethnicity as a viable, albeit socially manufactured cause of violence. To this end, it builds on the structuralist origins of constructivist work by updating the methodology to a post-structural acknowledgement of discourse’s inability to express reality. Through drawing on Roland Barthes and Jean Baudrillard, it articulates a model that explains the divergence between South Africa’s cultural acceptance of continuing ethnic conflict and empirical evidence suggesting otherwise. It considers the process through which these myths are generated, engaging with examples of cultural contributions to a hyperreal framework.

This essay refutes the thesis that enduring conflict in South Africa can be explained in ethnic terms. Responding to Rogers Brubaker’s observation that recent constructivist research “has grown complacent with success”, this paper answers his call to specify “how ethnicity is constructed” (2002: 175). As such, I expand on Jennifer Todd’s Bourdieusian argument that, although not innate, ethnic identity is a powerful social construct accepted by the public[1]. Recognising her appeal to French structuralist thought as a rare acknowledgement of the movement’s influence over recent constructivist approaches to ethnic identity, the first part of my essay outlines a theoretical model based on the post-structuralist work of Roland Barthes and Jean Baudrillard. Rather than explaining conflict as a natural manifestation of ethnic tension, I will reconcile Barthes’ mistrust of mass culture with Baudrillardian hyperreality in order to deconstruct the misnomer that South Africa’s conflict is a result of ethnicity. Instead, I will draw on evidence that demonstrates such conflict is only perceived as ethnic, incorporating Helen Moffett’s analysis that downplays ethnicity’s salience in sexual violence. My final section considers the process through which such myths are generated. Specifically, I will analyse the language of Thabo Mbeki’s “I am an African” speech alongside two South African novels that depict black-on-white violence: Sindiwe Magona’s Mother to Mother (1998) and J M Coetzee’s Disgrace (1999). Ultimately, I will show how the cumulative output of post-apartheid rhetoric has constructed a false narrative that has nevertheless been absorbed into South Africa’s cultural lexis.

Reconsidering Ethnicity: A Post-Structuralist Model

Brubaker’s objection to the complacency within constructivist research is comparable to Joseph Ruane’s call for a critical explanation that embodies the popularity of primordialism, alongside a constructivist rejection of essentialist approaches[2]. While ethnicity is a construct, how that construct is defined affects the theory’s credibility. Co-authoring a paper with Todd they criticise the inability of constructivist theories to “explain those specific instances when ethnic solidarity is intense and persistent” (2004: 209), as is echoed in Todd’s subsequent Bourdieusian rejection of the “soft constructivist analysis [of] psychology” (2005: 432). They address the issue of why a social construct is experienced as a fixed and not fluid component of identity. To this end, both works shift the agency of identity construction from the cognitive autonomy of the individual to a more useful focus on societal institutions that create rigid ethnic boundaries. While I accept the validity of their shared argument that ethnicity is “part of a continuum with other practical categories” like class (2004: 218), I am more drawn to Todd’s articulation of a model that emphasises ethnic identity as “socially embedded” in an overarching habitus, despite lacking “foundational status” (2005: 433). Identity is read as a construct maintained by societal institutions, not as an arbitrary factor subject to the ephemeral whims of individual actors.

Identifying these structuralist origins sets a useful precedent for addressing constructivist complacency. However, in accepting ethnic identity as a product of societal pressures, Ruane and Todd approach ethnicity as an explanation for conflict. When accepting ethnicity as a social construct, they empathise that these “cultural binary oppositions reinforce ethnic solidarity” (2005: 226) which then instigates violence along ethnic lines. Although this reading may apply to some cases, it is not a helpful approach when considering South Africa. Here societal informants construct the perception of ethnic solidarity by creating a superficial narrative, a distinction that is part of a broader divergence between structuralist and post-structuralist thought. While the former accentuates how discourse and language shape the world, the latter expresses anxiety over discourse’s inability to express reality (Barry, 2009). In emphasising the salience of ethnic constructions as legitimate causes of violence, Ruane and Todd therefore base their explanation on an unstable reference point and accept the cultural myth of ethnic violence. For this reason, I update the habitus model of ethnic identity to a post-structural awareness that their perception of intense ethnic feeling is in itself a rigid illusion masquerading as truth.

In developing a model that challenges the folly of using ethnicity to explain post-apartheid conflict, I initially turn to Barthes’ Mythologies. In analysing mass culture, Barthes builds a semiological argument around the image of “a young Negro in a French uniform” (1957: 115). Here the signifying picture gives meaning beyond the sign itself. The signified connotation “that France is a great Empire that all her sons, without any colour discrimination, faithfully serve under her flag” (ibid.) is repeated in the national narrative to construct an alternative version of the real. The media thus constructs a “sensual reality” that “postulates a kind of knowledge, a past [and] a memory” (1957: 116). Academia concerning South Africa acknowledges such symbols, with Christina Steenkamp noting how the image of an AK47 “has acquired cultural significance” across the Southern African states, now signifying an “association with liberalism” (2009: 48). South Africa itself uses “freedom songs [that] extolled its virtues and those of the fighters carrying it” (ibid.). As the cultural discourse is tainted by symbols that alter the perception of reality, I stress the potential for error in drawing on a binary portrayal of ethnicity as an explanation for conflict. Particular care is needed when considering a country whose history is framed by society’s attachment of axiomatic connotations to race. Rather, an awareness of ethnic violence as a hallucination exposes the category’s limitations.

Invoking Baudrillard’s view that what is natural has been replaced by a culturally produced hyperreal, I argue ethnicity is ascribed false meaning through the repetition of distorted narratives. That is, through a process of four “successive phases of image” a sign no longer bears “any reality ... it is its own simulacrum [or representation]” (1981: 1736). This idea of “the real [being] no longer real” (1981: 1741) draws heavily on Barthes’ identification of mass culture altering one’s perception of reality. Further developing this argument, Baudrillard describes ethnology as a “world of simulation [and] the hallucination of truth” (1981: 1738). Rather than a structuralist acceptance of ethnic conflict, my post-structuralist approach emphasises the artificiality of this socially constructed world. While Ruane and Todd are correct to stress the hardiness of ethnic constructs, South Africa’s cultural discourse should be recognised as polluted by signs that cannot signify the real. It is contaminated and thus incapable of reflecting reality.

The Nature of Conflict in Post-Apartheid South Africa

This paper does not challenge the presence of violence in South Africa. The murder rate of 32.2 for every 100,000 people is significantly higher than the world average of 7.6. Tellingly, over 17,000 South Africans were murdered between April 2013 and March 2014 (South African Police Service, 2014). Such is the extent to which violence has made an impression on South African children that Steenkamp refers to a 2000 children’s art competition where, out of 2000 entries, “at least 70% depicted death, guns, abuse [and] drugs” (2009: 50). Problematically, South African culture, as enunciated through political rhetoric and literature etc., portrays this violence as predominantly black-on-white crime.  South African musician and controversial activist for Afrikaner rights Steve Hofmeyr, wrote an infamous Facebook post stating that white farmers are being killed “like flies” and that “the daughters of our little [white] group are raped as we speak” (2013). This is indicative of a broader cultural consensus of whites as persecuted victims[3]. Melissa Steyn addresses this flawed, albeit popular use of inverted racism by recounting a white engineer’s perception of whiteness as a “formerly privileged, presently endangered, position”, alongside an Afrikaans woman’s belief that, unless whites “keep quiet [blacks] will take [their] home or kill [their] family” (2001: 71)[4].

While Todd’s analysis would adhere to an explanation of conflict as predicated on ethnic division, the consensus of ethnically motivated violence is a hyperreality unsubstantiated by research. In an interview with Action on Armed Violence, The Project Manager at the South African Institute for Security Studies, Lizette Lancaster, stressed that the notion of black-on-white violence “is only the perception” and that only 1.8% of whites have been killed since apartheid (Olgiati, 2013). Comparatively, whites “are disproportionately less killed ... you are more likely to be killed if you’re black” (ibid.). The use of the word “perception” is significant in reinforcing ethnic conflict as an artificial reality that resists empirical fact. She instead points to “socio-economic conditions” as a more realistic cause of violence, echoing John Mueller’s argument that conflict is not conducted on racial grounds but is instead a manifestation of “hooligan opportunists” (2000: 43). That black South Africans continue to be more impoverished than their white counterparts (Steenkamp, 2009) has likely contributed to the cultural misreading of poverty based crime as rooted in ethnic hatred.

Regardless, explaining conflict in South Africa as a result of ethnic division is theoretically supported in post-conflict criticism. Contrary to Steyn’s refreshing analysis that recognises the falsehood of the “colonial binaries that ... provide the scaffolding for reality” (2001: 69), Sven Simonsen emphasises how the “institutionalization of ethnicity has become an important hindrance to peacebuilding” (2005: 297). In analysing Bosnia and Afghanistan, he makes the sweeping assertion that ethnicity is inevitably highly politicised in post-conflict settings, stressing that systems utilising a presidential system are “far from what a society might need” as such forms “can be highly unstable” (2005: 308). This reading of post-conflict situations is similar to Todd’s in that both accept ethnic motivations as a viable cause of continuing violence. However, instead of a habitus where actors reproduce the socially constructed model of ethnicity presented to them, evidence suggests that this perception of an ethnic divide is a falsehood that rarely translates into empirical experience.

Returning to Hofmeyr’s remark on the supposed threat black rapists pose white girls, South Africa does experience prevalent sexual violence. Despite the already high ratio of 118.2 sexual assaults for every 100,000 people (South African Police Service, 2014) Martin SchÁ¶nteichargues that there is 60-70% more violence in South Africa than is reported (2001), while Helen Moffett credits a lack of faith in policing as rendering women “unlikely to report [their] attacker” (2006: 141). Moffett’s analysis presents a South Africa with “the worst figures for gender-based violence for a country not at war”, asserting that “at least one in three South African women will be raped in her lifetime” (2006: 129). However, she then laments how “the narrative about rape continues to be rewritten as stories about race, rather than gender” (ibid.), describing the area as “fraught with ... rape narratives [that] feature a barbaric Other, invariably inscribed as ‘darker’ (literally, morally and figuratively) than the victim”[5] (2006: 137). For Moffett, all ethnicities are equally culpable. Her analysis of racialised assumptions lends itself to my argument that one cannot simply be black in South Africa. Such is the strength of the cultural hyperreality that, despite empirical evidence to the contrary, the act of being black in itself signifies a separation from, and corresponding threat to, a persecuted white minority. Post-apartheid culture has thus continued to stress antithetic views of ethnicity that no longer represent reality.

The Origins of South Africa’s Ethnic Myths

Turning to the process through which such myths are generated, Steenkamp argues that South Africa’s “cultural (re)production takes place through various means involving ... cultural icons, rituals [and] the mass media” (2009:34). Building on the outdated racial prejudices of apartheid, South Africa’s cultural lexis has continued to attach additional meaning to race. Although a comprehensive account of the bases for such misnomers would go beyond the scope of this paper, I will illustrate three examples of South African culture that have contributed to this ethnic hyperreality. Beginning with Mbeki’s 1996 “I am an African” speech, Brubaker emphasises the role of “everyday talk, policy analysis [and] media reports” (2005: 165) in shaping a cultural consensus. Despite his claim that “I know what it signifies when race and colour are used to determine who is human”, Mbeki engages in primordial rhetoric: “I owe my being to the hills and the valleys ... to the Khoi and the San who ... haunt the great expanses” (1996: 252). Such discourse signifies an essential difference between South Africansof indigenous and European descent that is then reproduced by both ethnicities. Once this ethnic split is established, there is potential for a moral significance to be attached to ethnicity, with competing races identifying with Andreas Wimmer’s notion of the “category of the excluded ... chosen people who are ... culturally superior” (2008, 988).

Jan Penrose’s exploration of land’s significance in primordial thought is interesting in relation to Mbeki. For primordialists one’s perception of self becomes “indistinguishable from the soil itself”, thus giving “physical substance and symbolic meaning to notions of ‘us’ and ‘them’” (2002: 281). In declaring his “mind” and “knowledge” to be a product of the environment (1996: 252), Mbeki others the descendents of white colonisers and contributes to the primordial hyperreality of an ethnically divided South Africa. This undermining of [Archbishop] Desmond Tutu’s so-called “Rainbow Nation” narrative of a South Africa indifferent to race would rightly be condemned by Simonsen as an example of a presidential political system encouraging ethnic rhetoric (2005). Mbeki’s speech served to crystallise the notion of ethnic conflict and help write the perceived phenomena into existence. However, I stress the difference between a structuralist acceptance of political rhetoric’s ability to construct genuine ethnic division and a more nuanced awareness that Mbeki’s speech has instead led to a hyperreal cultural perception.

Magona engages with this primordial reading of self in Mother to Mother, a fictionalised account of the 1993 murder of the white American activist Amy Biehl. Published in 1998, the novel is a significant example of apartheid instances of violence being reproduced in post-apartheid culture. In Chapter 10 the novel’s protagonist believes herself to have “knowledge with which I was born ... we sucked it from our mother’s breasts, at the very least; inhaled it from the very air” (1998: 173). While the novel’s form as a fictionalised instant of ethnic conflict between white and black and South Africans suggests Magona possesses a meta awareness of fiction’s role in constructing narratives, superficial readings of the text would accept the presence of primordial portrayals of identity as passed on through material connections to one’s maternal and environmental origins. Through the character of Mandisa, Magona presents an essentialist perception of self that is in turn transmitted into the cultural discourse. In this sense, despite efforts to undermine the notion of innate ethnic characteristics, literature can unwittingly signify the hyperreality of a country split along intrinsic ethnic lines.

Unfortunately, such cursory readings of texts feature far too frequently in the critical discourse surrounding South African literature, a notable danger when one considers the country’s rich literary history. On Nobel Prize Winner J. M. Coetzee, Moffett recalls a visit to the USA where she had to assure an American academic that black South African men were not determined to punitively rape white women, an impression he “gleaned from reading Coetzee’s novel Disgrace” (2006: 135).  Although this incident is the text’s high point, Coetzee challenges the assumed connection between race and rape through the character of David Lurie, a white middle class professor who pursues a predatory relationship with a student. Regardless, these limited readings are recycled into the cultural lexicon and contribute to the hyperreal framework that in this case, is accepted by international observers[6]. Despite the admirable intentions of South African authors, that their novels depict ethnic conflict is enough to reinforce the cultural misnomer of enduring ethnically motivated conflict.


A post-structuralist approach reveals South African ethnic conflict to be a mirage that fades away upon closer inspection. This paper has challenged the misconception that ethnicity continues to be an important factor in post-apartheid conflict.  While Ruane and Todd are correct to point to the strength of socially constructed perceptions of ethnicity, their unwavering faith in the actuality of such instances does not apply here. Through the post-structuralist work of Barthes and Baudrillard, I have developed a model that adheres to Brubaker’s notion of ethnic conflict existing “only in and through our perceptions” (2002: 175). Nevertheless, the prevalence of such conflict remains a prominent myth in South African discourse and the expansive volume of cultural informants is a potential topic for further research. Echoing the continuing ethnic signification of danger from black citizens, a 2012 Victims of Crime Survey indicated that around 35% of households believed crime had increased from 2009 to 2011 (Statistics South Africa, 2012). South Africa must deconstruct this hyperreality before such fears lead to genuine instances of responsive violence. Equally, it must transcend Moffett’s view of a perverse culture where “challenging rape [is] often perceived as racism [due to] the inherent idea that you are blaming the black man” (2006: 135). Therefore, in order to properly address its significant issues, South Africa must raise public awareness to steer citizens away from the misnomer that ethnicity and conflict are one and the same.


[1]French sociologist Pierre Bourdieu argued that individuals internalise the values that structure their social world (a habitus). When responding to social situations people then “reproduce the[se] objective structures” (1992: 62).

[2]Although this primordial reading of ethnicity as an innate quality has lost traction in academic discourse, the citizenry and media tend to accept this model of ethnicity as an entrenched feature of identity in times of conflict. As Todd explains, actions are social products “even if they sometimes appear to individuals as primordial givens” (2005: 434).

[3]Tellingly, this status received over 2000 “likes”.

[4]This idea of the black South African threat underwent international scrutiny last year in the Oscar Pistorius trial. Writing for The Guardian, Margie Orford described the defence as predicated upon the presence of an “imaginary body of paranoid ... suburban South Africa”. Accordingly, the trial “tapped into an old and painful vein in which race, sex, power and violence converge” (2014).

[5]This reading is obviously influenced by Edward Said’s Orientalism (1978), a foundational post-colonial text that built on post-structural thought. The act of Othering is to use one’s ethnicity as a means of differentiating a normative sense of self from an alien Other. In constructing the hyperreal racial narrative of the black aggressor, South African culture has continued to Other its black citizens.

[6]For further examples of South African literature that include interethnic violence see Alan Paton’s 1948 Cry the Beloved Country in which a white civil rights activist is murdered by black intruders.

• Barry, P. 2009. Beginning Theory. Manchester: Manchester University Press.
• Barthes, R. 1957. Mythologies. London: Jonathan Cape.
• Baudrillard, J. 2001. The Precession of Simulacra. In: Leitch, V.B. ed. The Norton Anthology of Theory and Criticism. New York: W. W Norton & Company, pp.1729-1741.
• Bourdieu, P. 1992. The Logic of Practice. Cambridge: Polity Press.
• Brubaker, R. 2002. Ethnicity without groups. Archives Europeennes de Sociologie. 43(2), pp.163-189.
• Coetzee, J.M. 1999. Disgrace. London: Secker & Warburg.
• Hofmeyr, S. 2013. In layman’s terms: my tribe is dying. [Facebook]. 20 June. [Accessed 27 December 2014]. Available from:
• Lancaster, L. 2013. Interview with Serena Olgiati. 4 November, South Africa
• Magona, S. 1998. Mother to Mother. Boston: Beacon Press. Mbeki, T. 1996. I am an African. In: Fassin, D. ed. 2007. When Bodies Remember: Experiences and Politics of AIDS in South Africa. California: University of California Press, pp.252-253.
• Moffett, H. 2006. ‘These Women, They Force Us to Rape Them’: Rape as Narrative of Social Control in Post-Apartheid South Africa. Journal of Southern African Studies.  32(1), pp.129-144.
• Mueller, J. 2000. The banality of “ethnic war”. International Security. 25(1), pp.42-70.
• Orford, M. 2014. Oscar Pistorius trial: the imaginary black stranger at heart of the defence. The Guardian. [Online]. 4 March. [Accessed 27 December 2014]. Available from:
• Paton, A. 2003. Cry the Beloved Country. New York: Scribner.
• Penrose, J. 2002. Nations, States and homelands: territory and territoriality in nationalist thought. Nations and Nationalism. 8(3), pp.277-297.
• Ruane, J and Todd, J. 2004. The roots of intense ethnic conflict may not in fact be ethnic: Categories, communities and path dependence. Archives Europeennes de Sociologie. 45(2), pp.209-232.
• Said, E. 2003. Orientalism.  London: Penguin Books.
• Schönteich, M and Louw, A. 2001. Crime in South Africa: a country and cities profile, occasional paper No. 49. Pretoria: Institute for Security Studies.
• Simonsen, S.G. 2005. Addressing Ethnic Divisions in Post-Conflict Institution-Building: Lessons from Recent Cases. Security Dialogue. 36(2), pp.297-318.
• South African Police Service. 2014. Murder in RSA for April to March 2004/2005 to 2013/14. [Statistical release]. [Accessed 27 December 2014]. Available from:
• Statistics South Africa. 2012. Victims of crime survey. [Statistical release]. [Accessed 27 December 2014]. Available from:
• Steenkamp, C. 2009. Violence and Post-War Reconstruction: Managing Insecurity in the Aftermath of Peace Accords. London: I.B. Tauris & Company.
• Steyn, M. 2001. “Whiteness Just Isn’t What It Used To Be”: White Identity in a Changing South Africa. Albany: State University of New York Press.
• Todd, J. 2005. Social Transformation, Collective Categories, and Identity Change. Theory and Society. 34(4), pp.429-463.
• Wimmer, A. 2008. The Making and Unmaking of Ethnic Boundaries: A Multilevel Process Theory. American Journal of Sociology. 113(4), pp.970-1022.

Last updated : 28-Apr-2015

This article was produced for South African History Online on 28-Apr-2015