When the Past Is in the Present: The Paradox of Educational Opportunity and Social Inclusion in South Africa and Rwanda

South Africa

In South Africa, racial differentiation was initiated during the colonial and postcolonial period and then formally institutionalized through laws and policies under apartheid, which ended in 1994. In 1948, the Afrikaner Nationalist party gained political power, and it formally institutionalized a national system of racialized discrimination on the basis of the notion of “separateness.” Apartheid created distinct racialized identities through the regulation of movement, education, and labor. Formal racial categorization was codified through the Population Registration Act of 1950, which separated the population into four distinct racial groups—“White,” “Coloured,” “Indian,” and “Black/African”—and issued official identification cards.

After the end of apartheid and the transition to democracy and majority rule in 1994, South Africa adopted a “nonracial” approach to address the legacy of institutionalized racial discrimination. Nonracialism, or the “denial of the ‘reality’ of race” (Neville 1979:221), emerged as a prominent part of the antiapartheid movement and was implemented as the antithesis to a rigid social system that was distinctively marked by its separation of four state-constructed racial groups.

Similarly, education policy in the postapartheid era has followed a nonracialized approach that engenders equal rights and treatment principles, regardless of race or ethnicity. ¹ In South Africa, schools are not mandated to collect disaggregated data, although some schools choose to do so (Soudien 2004). Publicly available educational management information system data are not disaggregated by race.

Some scholars argue that nonracialism elevates equal rights and access at the expense of explicit discussions about the ways in which race continues to operate in South African society to either include or exclude (Soudien and Sayed 2004; Steyn and Foster 2008). The liberal discourse of inclusion, they declare, has allowed a redefinition of racism to include only state-sanctioned and overt acts of discrimination. Consequently, the nonracialism frame has enabled many to ignore the historical consequences of past, de jure racism, while racism in the form of everyday actions persists (Steyn and Foster 2008). For instance, in her work on students in diverse South African schools, Teeger (2015) found that school culture and educators’ nonracialist talk encourage students to ignore interpersonal and structural racism.

Rwanda

In precolonial Rwanda, notions of identity were initially linked to wealth and power (Pottier 2002; Vansina 2004). Under Belgian colonial rule, ethnic identity was firmly institutionalized through the 1933 census, which officially assigned fixed ethnic identities on identity cards. Under colonial rule, Tutsi were systematically favored over Hutu in the realms of education and employment (King 2014; Newbury 1988). The Hutu majority transitioned to power at the end of the colonial rule, and continuing a system of ethnocentrism, they implemented quotas in education and employment on the basis of ethnicity and region, favoring Hutu from the north and discriminating against the Tutsi (King 2014; Prunier 1995). Subsequently, the Tutsi minority faced periods of violence between 1959 and 1964 (Prunier 1995).

Extremist Hutu leaders initiated mass killings, drawing on “Hutu power” ideology that reconstructed Tutsi as the enemy and as an “alien race” (Mamdani 2001:190). Beginning on April 7, 1994, more than 11 percent of the total population (primarily Tutsi and some moderate Hutu) was massacred by ordinary Rwandans during 100 days (Prunier 1995). In the postgenocide period, the Rwandan government, an increasingly authoritarian and minority-ruled state, has adopted a nonethnicized ideology in national policies as a means to quell intergroup tension. To overcome the history of conflict between the two main “ethnic groups,” the Tutsi and the Hutu, the government has promoted the ideal of a unified Rwandan citizenry through an emphasis on reconciliation, a common culture, and a common language (Kinyarwanda) (Buckley-Zistel 2006; Longman and Rutagengwa 2004). ² Additionally, national laws and policies regulate discussion of ethnic identity in public spaces (Longman and Rutagengwa 2004; Purdeková 2008). For instance, the Genocide Ideology Law (2008) officially forbids identification in ethnic terms or division along ethnic lines.

Regardless of the intended goals of national discourse and policies to eradicate ethnic groups as a form of identification, researchers continue to document the persistence of ethnic identities in the present day (Freedman et al. 2004; Longman and Rutagengwa 2004; Russell 2013). Furthermore, several studies have noted the inequities in access to education, employment, and resources that favor Anglophone returnees (Reyntjens 2004; Samuelson and Freedman 2010). Despite the official rhetoric on unity promoted in policy documents, ethnic group boundaries linked to experience during the genocide and language sustain and endure.

South African Study

The South African data come from a larger study of eight schools in four U.S. and South African cities (Carter 2012). ³ We focus in this article specifically on two English-medium high schools in South Africa (grades 8–12). These data were collected over the course of six months from January to June 2004. ⁴ The multiracial school in Coast City (pseudonym), Palmer High School, had a learner population of approximately 1,000 students, 40 percent of whom were White students and 60 percent Black, Coloured, and Indian. ⁵ In Letisi (pseudonym), the multiracial school, Williston High School, had approximately 980 learners, more than half of whom were identified racially as Black. Thirty percent of the learners at Williston were White, 17 percent Indian, and three percent Coloured.

The second author visited both schools daily over a period of six months for roughly 250 hours of ethnographic observations, conducted eight semistructured group interviews with learners at both Palmer and Williston (about four students per group, on average), attended classes and various school activities several times a week, formally interviewed both school principals, informally interviewed a random sample of dozens of learners from different racial groups and educators at both schools across the five grades, and collected myriad school materials and brochures.

Rwandan Study

Data from the Rwandan case are derived from a multilevel and mixed-methods study (see Russell 2013), with quantitative and qualitative school-level data collected in Rwanda over a period of 11 months spanning 2010 to 2012. ⁶ Although the larger study included analysis of key national education policy documents, curricula, textbooks, as well as survey data gathered from secondary students, this article draws primarily on the qualitative data collected from interviews with teachers and students and classroom observations.

The first author conducted 20 interviews with history and civics teachers, 14 individual interviews and 21 group interviews with a total of 104 secondary students, and seven interviews with school directors at seven purposively selected schools in three different geographical locations. Students were randomly selected to represent boys and girls across a range of grades. Interviews were conducted in Kinyarwanda, English, or French depending on the preferences of the teachers and students. Two local Rwandan research assistants assisted in interpretation from Kinyarwanda to English.

The first author also carried out more than 70 hours of direct classroom observations of the same 20 teachers in political education (civics), history, and general paper courses (interdisciplinary social studies for advanced level) in the seven schools. The researcher was able to build trust though repeated visits to the schools and observations before conducting semistructured interviews with teachers and students. Although the researcher never directly asked about sensitive issues such as ethnic identity, teachers and students often brought these topics up as part of discussions around the curriculum and genocide.

For both case studies of the schools in South Africa and Rwanda, the individual and group interviews were recorded, transcribed (translated from Kinyarwanda and French in the Rwandan case), and coded using qualitative software (Atlas T.I. and NVivo). Analysts used an iterative and inductive coding process for both observations and interview transcriptions; initial codes and important themes were identified during a first pass of “open coding” and then refined during a second pass of “focused coding” (Emerson, Fretz, and Shaw 1995). Throughout the coding process, we wrote memos to document emerging ideas and themes across the codes. In what follows, we document the appearance of color-blind ideologies and increased opportunities for schooling, but where everyday interactions and school practices reinscribe racial and ethnic divisions that reproduce unequal school experiences.

South Africa: “The Nonracial Façade”

On a typical school day in the context of a group interview, a multiracial group of high school seniors (“matrics”) is asked about their overall educational experiences over the course of their five years of attendance at Williston, and the first thing out of these 17-year-olds’ mouths is something that indicts the school’s approach to educational inclusion. A 17-year-old high school matric explains how her African peers are sanctioned by the school principal for infractions significantly more than her White schoolmates, and the White learners in the interview group agree:

Williston’s students commented on daily conditions such as these that were hidden from the larger public unless they had the opportunity to go behind school doors and witness them. In this instance, African students lamented the differential treatment and discipline they received for similar infractions as white students. Despite national-level policies for equal treatment and inclusion across racial groups, discriminatory actions persisted at the school level. The disproportionality in discipline of African students was not only discussed by students but also observed. Williston dubbed its outside space where students were sent for “infractions” as the “jail,” and one wintry afternoon as the researcher passed by, she noticed that only African boys and girls were locked behind the gate of the “jail” outside on the school plaza. African youth were also more likely to be sent home for violations of the school’s hair policy—hair either too short (bald heads by the boys) or too “fancy” as described by Principal Billups for the girls.

At first glance, the appearance of a seemingly frivolous cultural matter emerged as a highly racialized and discriminatory one: the banning of specific ethnoracial cultural presentations of self. Williston’s principal, educators, and school governing body stood by a rigid policy that forbade African students from wearing any ethnically related hairstyles; and all of this occurred, ironically, under the larger social umbrella of South African nonracialism.

From Williston’s principal, Caryn Billups, the researcher learned firsthand how school officials maintained symbolic boundaries (Lamont and Molnar 2002) and restricted the inclusion of “ethnic” styles into a frame of respectability. The principal and teachers justified their policy as teaching students how to conform, and to prepare for the workplace.

The principal’s actions are one example of Chisholm’s (2008) classification of an “equality of opportunity” frame in which the educators construct policy such that no group is given “special privilege” unless it overtly speaks to an acceptable civil rights issue. For example, Mrs. Billups and staff had made an exception for a Sikh Indian boy and allowed him to ignore the school’s hair policy by wearing a turban and growing a beard. Williston’s staff justified this allowance as religious tolerance and a respect for religious diversity. In contrast, they relegated nonreligious, ethnic hair forms to mere changeable styles, not upheld by any specific creed or protected by law that could be worn on the weekend, as the principal informed me. Therefore, African learners found themselves bearing the disproportionate burden of adaptation to the White educators’ aesthetic about hair.

In comparison, John Dalton, Palmer High School’s principal, wrestled with the contentious cultural issue of hair, and his treatment of the matter signalled his and other school officials’ qualms about the diffusion of tastes from “Black” cultural repertoires to White learners. On one hand, he and his staff sought to recognize cultural differences, but they also sought to preempt any transmission of cultural styles across racial identity lines through a dual policy for Black and White learners:

Williston Principal Billups took a tough line and merely refused to allow “fancy” heads in her school. She disregarded (1) the disproportionate cultural burden on Black learners and (2) the racialized overtone this policy had taken in a school whose objective was to provide equal educational opportunity in a nonracial manner. Although the Black learners had approached their elected student representatives, the school governing body, and the principal yearly since my first visit there, this principal and her teachers had structured their hardline policy such that no learner would be allowed to present “fancy” heads and thereby avoid the accusation that hair policy explicitly made exceptions for one group and not another. In the process, the discourse surrounding this school’s policy defined ethnic hairstyles such as braids, twists, and locks as not “normal,” though this description was tempered by the euphemism “fancy.” Billups’s “equal treatment” or color-blind approach mandates that all learners assimilate to the dominant White cultural ethos of the school and that no special efforts are made to consider the cultural differences, tastes, styles, and/or difficulties of Black learners (Chisholm 1999).

In the second school, Principal Dalton, self-admittedly struggling with the notion of inclusion, expanded the school’s hair policy to allow Black learners more options for hair styles; he took phenotype into account and, in turn, implicitly acknowledged both biological and cultural conceptualizations of race. At the same time, he bounded the options for White learners to mitigate the chances that Black African learners’ hair tastes and preferences would influence White kids.

Exclusive Language

In a multilingual society such as South Africa, language instruction in schools unwittingly becomes another terrain of boundary maintenance. The South African Ministry of Education mandates that all schools ensure the proficiency of their learners in at least two languages. Primarily, the desegregated schools choose English and Afrikaans, and not the mother tongue of the majority of the nation’s populace (Fiske and Ladd 2004; Mda 1997). Thus, individual school language policies were not immune to political and cultural significations of the in group (the white minority) versus the out groups (mainly Africans).

Although the national language policy allows instruction in mother-tongue languages in school, in practice English is privileged over other African languages. English was the medium of instruction at both Williston and Palmer, and they offered English and Afrikaans primarily, and not the native languages of many of their African students. Williston did not offer Zulu, Xhosa, Pedi, Sotho, Tswana, or Ndebele, while Palmer offered Xhosa as an extracurricular elective to any students desiring to take it. Although some students questioned this logic, the general school approach did not change:

What I think is we shouldn’t be learning Afrikaans in school. We should be learning a Black language, because . . . as a White person, I’d love to learn a Black language to communicate with Black people, because Afrikaans—I—I don’t know if I could go to—the Afrikaans, like, way of living is like kind of dying out. (Williston grade 12 White male)

As a numerical minority in a black-dominant country, this particular White student took an instrumental approach to why he needed to learn an African language, as opposed to Afrikaans, which he perceived as a “dying” language, one only spoken by a small group of persons in South Africa. At one point while the second author visited Williston, she attended a few sessions of a multiweek workshop created by a few of the White teachers desiring to learn Zulu, the language of the majority of Williston’s students. In this workshop, our researcher learned that the motivation for some teachers pertained to their desire to know their students better. In that workshop, a teacher admitted that she had inadvertently misunderstood the motives of one of her students and had given him a demerit because of her inability to understand his language and culture.

Linguistic differences and the political meanings associated with language sustain racial boundaries among learners, in particular. ⁷ On one hand, the linguistic divide among ethnoracial groups has begun to lessen as English has become the major language of instruction in many South African public schools. On the other hand, symbolic boundaries are reinforced as desegregated schools overtly exclude the native language of the majority of learners, in many cases, from the curriculum.

Black learners and their parents are not without agency, however. Each school’s governing body is allowed to set its own language policy, and it is the duty of the provincial department of education and the school governing bodies to provide classes in certain languages if a sufficient number of learners request them. The perceived better mobility context of the desegregated schools masks the hegemonic nature of the schools’ cultural and educational practices. Many Black families consent implicitly to not being taught African languages formally and are willing to forego the preservation of their home languages in exchange for fluency in English, something that we might attribute to the forces of globalization, labor-market demand and English as the lingua franca of the international community.

Despite national policies for equal treatment and inclusion of all groups in education, the data from the case study demonstrate the persistence of both social and symbolic boundaries in terms of differential treatment and discipline, cultural inequality in different groups’ presentation of self, and limited language diversity. Taken together, all of these factors point to a particular paradox: desegregation and ethnoracial diversity do not necessarily correspond to social and cultural inclusion (Carter 2012).

Rwanda: Constructing a New De-ethnicized Identity

In a government-run school in a newly built town less than 30 kilometers from the Ugandan border, the researcher interviews Beatrice, an 18-year-old Anglophone returnee from Uganda, about her experience in school. ⁸ However, when the conversation takes a more personal note, and the researcher asks about the type of person she would like to marry one day, ethnic terms arise in the conversation: “In my opinion, it isn’t necessary to say that such and such person is Hutu, Tutsi, or Twa. . . . But for me, I would prefer to marry someone from my group [in the context of ethnic groups].”

In this example, although Beatrice says that the use of ethnic terms is unnecessary, she at the same time distinguishes between people from her group and those in the “other” group, those who are “called Hutus.” These offhand comments about the “other” group serve to reinforce symbolic boundaries across social groups in the face of what we refer to here as “ethnic blindness.” When asked about tensions at school, she goes on to explain that the cause stems from “ethnic groups”:

In this instance, and in other instances, students do not directly self-identify as Hutu or Tutsi, but they clearly indicate their ethnic affiliation in the way they reference the other group through a form of “othering.” Furthermore, the girl attributes tensions in schools to “ethnic groups,” which demonstrates that the construct of “ethnicity” is still prevalent and used to explain tensions between groups of students. Hence, even though national policies espouse a nonethnic and unified approach, this example illustrates how individual beliefs and behaviors at the school level in fact reinforce ethnic identity.

In an interview with another student at a religious school in the capital city, Fabrice, a Hutu boy, uses ethnic terms to explain why he does not receive the government scholarship for orphans (Fund for Survivors of Genocide [FARG]), even though his father was killed by the Rwandan Patriotic Front during the war:

I don’t receive any aid from FARG. In fact, the primary condition to receive FARG is that you are poor and a Tutsi. This means survivors. So if you are Hutu, you can’t receive aid from FARG. But now, we don’t talk about different ethnicities. (School 7, September 2011)

Hence, other orphans who lost their parents during the civil war but who are not identified by the local government as “Tutsi” do not receive this scholarship. Another student, Richard, age 23, at a private school with a large population of genocide orphans, lamented that although he is an orphan and considers himself to be a “genocide survivor,” he did not qualify for the scholarship. He explained that he felt excluded by other “survivors” because he did not receive the FARG scholarship:

Even here at school, there are some students who form groups, for example students supported by the FARG. Sometimes when they see that you are not supported by FARG, they think that you are not a genocide survivor. So when they are in groups, you can’t come and share ideas with them. They don’t feel at ease with you. (School 5, October 2011)

National government scholarships and supporting organizations for genocide survivors create social boundaries between survivors and nonsurvivors in resource-scarce educational environments where many students hail from disadvantaged backgrounds and have difficulty paying the direct and indirect costs, such as school fees and uniforms (Paxton and Mutesi 2012) Students who received FARG funds or participated in Association des Elèves Rescapés du Génocide (AERG) were often signaled out as being different. The FARG provides funds for students classified as “survivors of the genocide.” ⁹ Although membership in the Tutsi ethnic group does not officially constitute criteria to receive the scholarship, because ethnic identity no longer exists officially, the tacit meaning of a genocide survivor is to be a Tutsi, the ethnic group targeted during the genocide.

The official government discourse on the genocide during the annual commemoration only recognizes Tutsi as genocide survivors, even though moderate Hutu were also killed (Burnet 2009). One student explained that FARG supports only “genocide survivors, or as we say in Kinyarwanda ‘Abahigwaga’—those who were hunted during the genocide” (School 5, October 2011). Students who did not belong to these groups often commented that genocide survivors or FARG students often preferred to associate with one another. Thus, even for students who do not have direct experience or memories of the genocide, the affinity for other students who come from families who have experienced the genocide is strong.

Although there were several clubs in the schools aimed at the promotion of unity and reconciliation that were open to all students for participation (such as the Never Again Rwanda Club and the Unity and Reconciliation Club), AERG, an association for genocide survivors, was more exclusive. Although the association did not explicitly forbid Hutu students from joining, it was understood among the students and teachers that the association was for genocide survivors and sympathizers (Tutsi returnees).

Ethnic identity in this context is linked to experience during the genocide, so it is not seen as divisive or taboo. Hence, although ethnicity is a taboo subject and students never explicitly self-identified as belonging to one group or another (although they did so indirectly), in this particular context around the genocide, ethnicity was invoked openly. In a meeting of AERG, members queried students who wanted to attend the club and asked them to share their personal histories. This type of explicit ethnic identification would not likely be expressed in a public context; however, because AERG meetings are private and attended only by student members, they are not subject to scrutiny by the school administration. In this instance, membership in AERG functions both as a social boundary that demarcates membership and also a symbolic boundary that creates the terms of a (Tutsi) survivor identity.

In Rwanda, institutional structures and organizational practices and policies influence the school culture and restrict the extent to which students are free to discuss sensitive issues linked to ethnicity and the genocide. During group interviews, students often echoed a sentiment in line with the curriculum and textbooks that emphasized a narrative of postgenocide reconciliation and unity, whereby all Rwandans are the same. However, despite the mantra of “one Rwanda” and the taboo around the emphasis on ethnic identity, students continued to invoke ethnic classifications indirectly in conversation when discussing the genocide, language, and groups at school. Although one cannot explicitly ask about ethnicity or self-identify as belonging to any particular group because of fear of the genocide ideology law, there are other markers that indicate one’s ethnicity, such as whether someone is a “returnee” or a “survivor,” a euphemism for “Tutsi.”

Exclusive Language, Too

In a private school near the capital, Karen, a Ugandan history teacher, described the groups that formed in class when she assigned group work to the researcher:

So when you are grouping them you’re sort of mixing them. You mix the Francophones and the Anglophones. Obviously, when you tell them to group themselves, the Anglophones will all sit together because they know each other. . . . When you tell them to sit in a group of ten, you will find ten Anglophones sitting around one table. (School 5, October 2011)

In this example, the teacher notes the boundaries that students invoked about language. Language in Rwanda, as in South Africa, serves as a signifier of new symbolic boundaries linked to identity. For instance, Francophone students often referred to “Anglophones” as a distinct group or spoke about “cultural differences” between groups, expressing preference for others who spoke French or were similar to them. Because ethnicity is no longer a salient marker of identity, at least officially, other markers such as experience during the genocide and language have emerged as stronger identifiers. With the shift in the medium of instruction from French to English, the distinction between Anglophones and Francophones has been amplified. The change in the medium of instruction from French to English in 2008 was presented by the government in practical and nonpolitical terms, namely, for integration into the East African Community and the global economy. However, some teachers and students perceived the policy change as political and as benefitting Anglophone Tutsi returnees associated with the ruling class (Samuelson and Freedman 2010).

When applied within the school context, the new national language policy, although seemingly apolitical, serves to reinforce symbolic boundaries and privileges one group over another, fostering divergent identities (Bourdieu 1991; Carter 2012). For Bourdieu (1991), language exemplifies “relations of symbolic power in which the power relations between speakers or their respective groups are actualized” (p. 37). In the case of Rwanda, the power of the Anglophone Tutsi elite is realized in the classroom, as well as at the political level. However, the switch to English is justified by the government in globalized terms, as English is increasingly the global lingua franca and the most important foreign language in primary and secondary schools worldwide (Cha 2007).