The Dilemma of the Mother Tongue: Prospects for Bilingual Education in South Africa Felix Banda Linguistics Department, University of the Western Cape, Private Bag X17, Belville 7535, South Africa The paper examines the position of the mother tongue in the proposed additive bilingual programmes in South Africa. It is argued that since the teaching and use of African languages, particularly as media of instruction, is less than adequate, their use in additive bilingual programmes, the official South African language in education policy, is unlikely to be successful. The paper is divided into three parts. First, the sociolinguistic, cultural, and political factors are examined. Here, apartheid education and its legacy are discussed. Second, using data from De Klerk (1996) and Smit (1996), language use and attitudes of South Africans to language and education are discussed. In this section, the pressures on standard African tongues by urban varieties and English are highlighted in relation to the pedagogical and theoretical underpinnings of additive bilingualism. Third, the practical possibilities for the implementation of mother-tongue medium of instruction given the 11 official languages, as well as the implementation of a viable bilingual education programme in which mother tongues play a decisive role are discussed. Finally, it is concluded that government’s additive bilingualism policy is unlikely to succeed as long as role models, learners and their parents see little utility in languages other than in English. Most academics and researchers in South Africa are unanimous that the mother tongues should be at the core of bilingual programmes (Alexander, 1995; Heugh, 1995; Luckett, 1995; Pluddemann, 1996). The present situation in which English is the preferred medium of instruction is thought to impede learning, and that it only leads to poor mastery of both English and the mother tongues. In fact, the poor matric results and the general lack of academic skills and intellectual growth among blacks at high school and tertiary levels, have most often been attributed to the use of English as a second-language medium of instruction in South Africa (cf. Pluddemann, 1996; Heugh, 1995; Luckett, 1995). Research also shows that increasingly black and coloured parents (the disadvantaged group during apartheid), are demanding English medium of instruction for their children (Pluddemann, 1996). This begs the questions: why, then, do black and coloured children and their parents still insist on English medium of instruction? That South Africa is a multilingual country with more than 20 languages (cf. Lemmer, 1993) is no longer in dispute. There is little doubt that the apartheid nationalist government took its belief in preserving ethnic identities too seriously to the extent of attempting to create monolingual societies out of multilingual South Africa. Apart from separating people according to race and language group, the apartheid regime also insisted on mother-tongue education. However, recent literature shows that typically, a South African uses at least two languages depending on context. Blacks, particularly those in urban areas, have 0790-8318/00/01 0051-16 $10.00/0 LANGUAGE, CULTURE AND CURRICULUM © 2000 F. Banda Vol. 13, No. 1, 2000 51 52 Language, Culture and Curriculum been known to use three or more languages (cf. De Klerk, 1996; Winkler, 1997; Slabbert & Finlayson, 1998). What, then, is the utility of the ‘mother-tongue’ concept given such multilingual reality? And can additive bilingualism as a language in education policy work efficaciously given the multiplicity of languages spoken in South Africa? The paper answers the three questions in three sections. The first section traces the history of language in education policy in South Africa. The second section deals with language use and attitudes, and the possible effects on government’s language in education policy of additive bilingualism. The third section looks beyond the pedagogical and theoretical concerns to examine other possible obstacles to change away from English medium of instruction, and to successful implementation of additive bilingual programmes in South Africa. The Sociolinguistic Situation South Africa has eleven official languages with the following percentage distribution if the 1991 census is anything to by (cf. De Klerk 1996): Zulu 21.9%, Xhosa 17%, Afrikaans 15%, Northern Sotho 9.6%, English 9%, Tswana 8.6%, Sesotho 6.7%, Tsonga 4.3%, Swati 2.6%, Venda 2.2% and Ndebele 1.5%. In line with the apartheid group areas act people were grouped into what were perceived as homogeneous ‘homelands’, each with its own language and, supposedly, culture. However, data on linguistic distribution suggests that the distribution of languages in South Africa is so diverse that all the eleven official languages are represented in all the nine provinces. For instance, Kwazulu-natal was designated a Zulu area, but Gauteng and Northern and parts of Mpumalanga provinces have more than a million Zulus each (De Klerk, 1996). Similarly, significant numbers of Northern Sothos, Tswanas, Tsongas, etc. are found in the other provinces, not just in Gauteng Province. In addition, with the demise of apartheid one can only assume that there is freer migration of South Africans than before between the various provinces, so that all things being equal the linguistic distribution might turn out to be even more divergent in the next decade or so. Thus despite the apartheid government’s policy of separating South Africans according to their ethnic and language orientation, the language distribution still shows a remarkable diversity. In turn, any bilingual education programme in South Africa will have to take into consideration the distribution of languages. Apartheid and its legacy The amalgamation of the former Boer republics with the Cape Colony and Natal in the Union Act of 1910 gave Afrikaans (hitherto Dutch) the boost that it needed. Afrikaans was made equal and joint official language with English (cf. Lanham, 1996). Although economic power was still with the English, there was an increase in English-Dutch bilingualism as Afrikaners started entering the urban job market and in government as civil servants. But Afrikaner nationalists were hostile towards playing second fiddle to English and ‘the concept of the taalstryd (language struggle) was cultivated and language loyalty became the biggest social division in white South African society’. (Lanham, 1996: 22). Whites were effectively split into two camps: English- and Afrikaans-speaking. The Dilemma of the Mother Tongue 53 The division widened with the alignment of the Afrikaner National party with the Nazis during the Second World War. At the same time use of English among black South Africans was on the increase thanks mostly to availability of state and mission schools, and more interaction with English speakers in the ever-growing cities and the work places. However, all this was to change with the coming to power in 1948 of the Afrikaner National Party. According to Lanham (1996: 26) the Afrikaner National Party proclaimed ‘confrontation with all things English … and attempts [were] made to enforce Afrikaans over English in every sphere of public life’. Afrikaner nationalists perceived education as a weapon through which to advance Afrikaans and reduce the influence of English in South Africa. Black education suffered the most under the changes that followed. Particularly, with the Bantu Education Act of 1953, Afrikaner nationalists swiftly implemented mother-tongue education and systematically reduced the role of English while increasing that of Afrikaans. The switch in medium of instruction was delayed until the secondary school (Lanham, 1996; Branford, 1996). At that point blacks were to learn through both English and Afrikaans as media of instruction. Although there was strong opposition from both white and black communities to such draconian education measures on an unwilling black people, the policy was enforced. In effect, the Afrikaner nationalist government went on a deliberate campaign uprooting white English mother-tongue teachers from Bantu Education, thereby denying black children authentic models of English and well trained, experienced teachers. As Lanham (1996: 27) succinctly puts it: Social segregation, the removal of white teachers from classrooms including mission schools, and the denial of entry to so-called white English universities, confined the black child’s encounter with English to the classroom with teachers, themselves products of deprived learning experiences with little gained in knowledge of teaching methods or competence in English from training colleges, which had suffered the same way as the schools. Thus, the lack of consultation with stakeholder, the removal of experienced and trained teachers, lack of funding, and generally the lack of interest in black education culminated into the collapse of Bantu education. The demise of Bantu education is partly responsible for blacks associating mother-tongue education on which it was based with mediocrity and failure. Opposition to forced Afrikaans medium of instruction in black schools culminated in the now famous Soweto Revolt of 1976 in which the apartheid police massacred protesting school children. This in part explains why blacks still appear to have a negative attitude towards Afrikaans. Pluddemann (1996) reports that black parents prefer their children to be educated through the English medium of instruction (EMOI), rather than Afrikaans medium of instruction. The Sunday Times (25 January 1998) carried a story of black parents, who rather than have their children learn through Afrikaans at a nearby school, decided to keep them at home until such time that the school introduced an EMOI stream. It is therefore interesting to note that the current language in education policy is additive bilingualism. Government has legislated that a child’s mother tongue Language, Culture and Curriculum 54 will be central throughout his/her education, and the second or third languages will supplement efforts in the L1 (cf. South African School Act, 1996). For blacks this means an African language being the significant language of education, and effectively having education in a language other than English or Afrikaans. As has been implied already, the problem is not how best to integrate African tongues into the envisaged additive bilingual programmes, but rather how to change blacks’ attitudes towards mother-tongue medium of instruction, particularly considering the apartheid legacy. This becomes even more evident in the next section when we analyse data from Smit (1996) and De Klerk (1996) on language use, preferred medium of instruction, attitude to English and degree of multilingualism in South Africa. The two studies on which this paper builds were done in Grahamstown, Easter Cape, South Africa. Subjects were high school and university students. Language Use and Attitudes It has been suggested by studies on language use in Africa that for political, socioeconomic and educational considerations, those whose first language is an African language are likely to be at least bilingual (cf. Banda, 1995; 1996). If Table 1 is anything to go by, this appears true of South Africa, despite attempts during the apartheid regime to divide people into unique linguistic communities, each with its own ‘mother-tongue’ education. Those South Africans who speak an African language tend to be multilingual, while English and Afrikaans L1 speakers tend to be either monolingual or at most bilingual. Table 1 Declared spoken and written multilingual competence (%) (n = 2975) Number of languages English Afrikaans Xhosa SA Black 1 27 0 1 0 Spoken 2 62 73 52 30 3+ 11 27 47 70 1 21 1 1 2 Reading 2 68 69 31 28 3+ 11 30 68 70 Note: SA Black refers to ZuluTswana and Sotho. Source: Adapted from De Klerk (1996: 116). With regard to language use outside the classroom, Table 2 shows that English is the language of choice, though Xhosa speakers appear to use more of their mother tongue. I should point out that the table does not give an indication of incidents of code-switching or whether ‘outside the classroom’ extends to the neighbourhood and home environments. However, the data’s significance should be measured in terms of informants’ perceptions of language usage, which also offer us insights into their attitudes towards the languages concerned. What is worth noting here is the fact that Xhosa L1 speakers’ extensive use of the mother tongue outside the classroom can be construed to suggest that they have a strong sense of identity and that for them English is more a ‘classroom’ language than the other language groups (cf. Saravanan, 1993; Banda, 1996). The Dilemma of the Mother Tongue 55 Table 2 Language use outside the classroom (%) (n = 2975) Afrikaans Xhosa SA Black Always Eng. L1 35 5 5 15 11 2 Mostly Eng. L1 57 25 35 68 61 38 Sometimes Eng. L1 8 45 40 12 28 52 Seldom Eng. L1 0 25 2 4 0 8 Never Eng. L1 0 0 0 1 0 0 Source: Adapted from De Klerk 1996: 121 This does necessarily mean that Xhosa speakers would prefer Xhosa as a medium of instruction. This is evident in Table 3. On the contrary, Xhosa L1 speakers, like other informants were unanimous in their choice of English as the medium of instruction. It is also worth noting the apparently changing role of languages in South Africa, as Afrikaans, a language whose status and appeal the Afrikaner Nationalist government had tried to match with English, finds itself as one of ten ‘other languages’ in the English-plus-ten-other official languages combination. Table 3 Preferred medium of instruction (%) (n = 2975) English Afrikaans Xhosa SA Black English 94 83 74 96 English/Afrikaans 2 13 3 1 English/Xhosa 4 4 22 3 Source: Adapted from De Klerk (1996: 117) Of interest in the study by Smit (1996) are issues relating to language, social identity and attitudes to a variety of South African Englishes. The attitudes were measured through respondents rating of three accents: mother-tongue English (MtE), Black South African English (BlE) and Afrikaans English (AfE) on five attributes: competence, career-mindedness, personal integration, social attractiveness and self-awareness. It is noteworthy that legally, segregation according to language, race or colour was no longer the case at the time of the research, but in practice ‘the majority of all pupils in Grahamstownstill attended those schools the apartheid system had reserved for them’ (Smit, 1996: 85). Thus, respondents could be categorised into five school systems popularly known by the following acronyms: … DET: Department of Education and Training (EMOI), mainly black pupils; … DEC: Department of Education and Culture (Afrikaans MOI), mainly coloured pupils; … CED-A: Cape Education Department (Afrikaans MOI), mainly white Afrikaans L1 pupils; … CED-E: Cape Education Department (English medium of instruction), mainly white English L1 pupils; and … PRV: Private Schools (EMOI), mainly white English L1 pupils. Language, Culture and Curriculum 56 Table 4 shows that on average, there is no clear-cut preference for any particular accent, although the trend is towards MtE. Note that ‘black’ English (BlE) and Afrikaans English (AfE) are evaluated worse than ‘white’ English. It is also interesting to note that the general trend towards ‘white’ English (MtE) appears to be largely influenced by preferences related to attributes of competence and career mindedness. These qualities are usually associated with personal achievement in formal and professional spheres. However, during apartheid, these attributes were largely reserved for whites. In other words, for blacks, knowledge of English did not usually translate into socioeconomic mobility for the nature of apartheid was such that non-whites could not achieve the same status or power levels as enjoyed by whites with equivalent socioeconomic qualities and educational qualifications. Table 4 Ratings on a variety of Englishes (n = 282) Competence Career mindedness Personal integration Social attractiveness Self-awareness Average mean BlE 2.44 2.08 1.85 1.75 1.72 1.97 AfE 2.37 2.05 1.87 1.97 1.92 2.04 MtE 1.95 1.87 1.73 1.95 1.95 1.89 Average mean 2.07 1.83 1.67 1.71 1.71 1.80 Note: The lower the mean the more positively evaluated. Source: Adapted from Smit (1996: 107). Therefore, it is not surprising that taking into account the ratings of each ethnic group with regard to the three varieties of English, non-whites prefer their own variety of English rather than MtE. This is demonstrated in the ratings of the five types of school on the three varieties of English regarding socioeconomic status as can be seen from Table 5. It is significant that only white Afrikaans L1 speakers (CED-A schools) would prefer MtE to their own AfE. In essence the fairest conclusion one can draw from Table 5 is that, generally, each language group appears to favour its own variety of English. Table 5 Ratings on social status (n = 282) DET DEC CED-A CED-E PRV BlE 1.83 2.51 3.06 3.05 3.38 Note: The lower the mean the more positively evaluated. Source: Adapted from Smit (1996: 108). AfE 1.89 1.67 2.09 3.03 3.32 MtE 2.40 1.93 1.85 1.51 1.72 The Dilemma of the Mother Tongue 57 In a nutshell, blacks’ preference for EMOI could be due to their quest to attain personal achievement in formal and professional spheres so as to have attributes of status and power denied them during apartheid. In summary: … Monolingualism is more prevalent among English L1 speakers; bilingualism is more prevalent among Afrikaans while those whose L1 is an African language tend to be multilingual. … Some English and Afrikaans L1 speakers would not mind an English/African language or Afrikaans/African language MOI. … All language groups prefer an EMOI, but each prefers its own variety of English. Only white Afrikaans L1 speakers would prefer MtE. … On average English is the language of choice outside the classroom, particularly in a multilingual context. But Xhosa L1 speakers use a lot more of their L1 than other groups. … There is a possibility that among Xhosa L1 speakers, English can be taken to be more of a ‘classroom’ language than a language for interpersonal communication outside the classroom. … ‘Black’ English and Afrikaans English are evaluated as being less desirable than ‘white’ English with regard to attributes of competence and career mindedness. … The supremacy of MtE (‘white’ English) over other languages is not confirmed. But ‘white’ English appears to be associated with attributes of competence and career mindedness. The issues Government documents (cf. Curriculum 2005, 1997) on language in education policy appear to emphasise integration in diversity, rather than coercion and segregation as was the case during the apartheid days. By integration, it does not by any means imply movement towards the imposition of one language or several lingua francae. On the contrary, the diversity of South Africa’s linguistic and cultural patrimonies, solidly anchored in historical developments leading to South Africa’s new constitution, makes it absurd seriously to envisage a language situation that could lead to language death and shift (unless one includes the Khoi-San languages). Ironically, it is the apartheid legacy that appears set to preserve the linguistic communities in South Africa, be they small or large, for some time to come. As argued elsewhere, people in South Africa are still more or less living in those areas set aside for them during apartheid. Moreover, in general, though accepting a significant role for English in the new South Africa, South Africans are largely closely tied in with feelings of regional or ethnic identity, and as a consequence few would even envisage sacrificing their linguistic or cultural identity on the alter of South Africa’s integration. Blacks, for instance, want to use English as a tool for socioeconomic mobility, while preserving their linguistic and cultural identity. Thus, it is not at all surprising that the findings of the Grahamstown’s research show respondents from different linguistic backgrounds preferred their own variety of English. This brings into sharp focus Ndebele’s (1987: 13) pronouncement that: ‘South African English must be open to the possibilities of its becoming a new language. This may result from the proximity of English to indigenous African languages’. 58 Language, Culture and Curriculum Therefore, integration in South Africa’s highly diversified population should be interpreted as interventions to ensure and to propagate multilingual proficiency. This means that language management programmes should ensure that English and other languages in South Africa play a facilitative role and not a displacive role. However, one hurdle in integrating South Africa’s educational set up emanates from the inherited legacy pertaining to institutional inequalities and differences between so-called former white schools and so-called former coloured or black schools. The inequalities relate to the delivery and quality of education, to procedures and practices that weigh against so-called former coloured and black schools. In this regard, government has tried to improve the delivery and quality of education in the former black and coloured schools through increased funding, retraining of teachers and the introduction of a new curriculum. At the same time schools have been desegregated through the dismantling of apartheid laws and enactment of a new constitution. Still, there have been problems associated with the implementation of new government policy. For instance, there are reports of white parents and principals (through the Parents-Teachers Association) in some private schools resisting full integration of ‘their’ schools into the new South Africa, by subtly raising school fees under the guise of maintaining ‘high’ standards, but effectively shutting out blacks and coloureds who cannot afford such fees. The other problem relates to the choice of variety of English and the diversity of mother tongues found in South Africa, and the type of classroom practice to engender this diversity of languages. A comprehensive bilingual programme in South Africa would have to give an indication of which variety of English and which African language(s) should be used in a particular region. Such a programme may also have to indicate, at least in a general way, the possible classroom practice to be adopted to ensure and maintain the multilingual nature of South Africa. Following government’s language in education policy of additive bilingualism, several models of bilingual education have been proposed for South Africa (cf. Luckett, 1995; Heugh, 1995; Alexander, 1995). They are all based on mother-tongue education, preferably as soon as a child enters school. Of particular interest, Luckett (1995: 77) argues for ‘affirmative action for African languages … to be developed to carry cognitively demanding content and concepts’. In as far as the additive bilingualism argument is concerned, Luckett (1995) cites research on the Threshold Project done in South Africa in the early 1990s which showed that many black pupils had difficulty adjusting to the switch from mother-tongue education to English-medium of instruction in Standard 3. The research, according to Luckett (1995), revealed that pupils could not transfer knowledge acquired in the L1 to English; at the same time they could not transfer what they had learnt in English to the first language. Drawing on Cummins’ (1981) distinction between basic interpersonal communicative skills (BICS) and cognitive academic language proficiency (CALP), she attributes this failure by pupils to transfer acquired knowledge to failure to achieve CALP in either language. However, she goes on to propose that pupils should use their mother tongue until they reach CALP: ‘Only once pupils can operate in CALP in their first language should they begin to operate in a second language at cognitively-demanding levels’ (Luckett, 1995: 76). Clearly, the problem does not The Dilemma of the Mother Tongue 59 appear so much with the medium of instruction, as with classroom practice. If anything the research seems to suggest that both English and the mother tongues suffered from poor classroom practice which led to inadequate acquisition of both English and the mother tongues. There is also the question of the pedagogical implication of using the CALP argument in the South African context to support the medium of instruction in mainly one language – the mother tongue. According to Cummins (1981: 24), it requires 5–7 years for a child to achieve CALP that enables him perform well on academic skills. This might be taken to mean that a child should be introduced to English earlier rather than later (cf. Lemmer, 1993). On the other hand, it can be assumed that a South African child has already achieved BICS in their mother tongues by the time they enter school. This is where Cummins (1981) Common Underlying Proficiency (CUP) comes in. CUP appears to support early mother-tongue education in that it assumes that efficiency in the mother-tongue transfers to the L2. But if the Threshold Project research (Luckett, 1995) is anything to go by, there is need for improved mother-tongue education and EMOI education. Therefore, given the South African context, it makes more pedagogical sense to argue for improving both mother-tongue and EMOI education, than to use the CALP and CUP arguments for mother-tongue education. In fact, the concept of ‘mother tongue’ is increasingly becoming fuzzy and untenable considering rural-urban migration and the changing South African language context after the end of apartheid and legislated compartmentalisation of people according to race and language group (cf. Calteaux, 1995: Winkler, 1997: Slabbert & Finlayson, 1998). Regional language education is more apt. I will elaborate on this later. Current school situation and practice In spite of the government’s policy of additive bilingualism, schools have generally continued teaching as before. This means that primary and secondary schools are teaching through Afrikaans or English. It can be argued that some schools that were Afrikaans MOI only have opened an additional EMOI stream. Also, some former Afrikaans only and English only ‘white’ schools are now offering African languages such as Xhosa and Zulu as optional subjects. The token use of Xhosa and Zulu does not necessarily mean full integration of languages and language groups. In practice, working-class children from black and coloured families cannot afford the high fees of these former ‘whites only’ schools. Almost without exception, school application forms and other official documents in South Africa have a ‘Home Language’ section. Ironically, despite the overwhelming evidence of multilingualism and government’s additive bilingual policy, it is still assumed and expected that South Africans are typically monolingual using a ‘Home Language’ in their neighbourhoods. The argument is sometimes used by a school to deny a child entry on account of his/her ‘Home Language’ if it is deemed to differ from the medium of instruction of the school. In such a situation a child is ‘encouraged’ to enrol in another school. Although laws have been promulgated to transform and integrate higher education, there is still no university offering instruction through the medium of an African language. Moreover, despite talk, there has been minimal curriculum 60 Language, Culture and Curriculum transformation apart from the fact that English is increasingly becoming the de facto MOI in institutions of higher learning. Those institutions that used to offer instruction in Afrikaans only are being forced to open English streams as well. The reason is that to get government subsidy, they need black students to swell their numbers as well as to show that transformation is taking place. But black students as already seen prefer EMOI and have had their secondary school education in English (or a mixture of English and an African language). Thus the majority of South African schools and tertiary institutions have a monolingual orientation despite the government’s policy of additive bilingualism. The situation is unlikely to change in the near future. EMOI and its Implications It is interesting to note that academics (backed by research albeit most of it done outside South Africa) appear unanimous that there should be mother-tongue education throughout schooling. This is in sharp contrast to the findings in the present study which suggest that learners from different linguistic backgrounds would prefer at least an English MOI throughout schooling. Recent studies have also shown that black parents and increasingly coloured parents (Afrikaans L1 speakers) prefer EMOI for their children (Pluddemann, 1996). What is at issue here is whether those who prefer EMOI should be forced to have mother-tongue education; or indeed whether learners and their parents have the right to choose, or can make an informed judgement as far as medium of instruction is concerned. However, if there was anything to be learnt from the unpopular Bantu education it is that people should not be forced to learn through a language they do not want. Another issue relates to the phrase ‘mother-tongue education’. It has already been shown how variegated the linguistic situation in South Africa is despite more than 40 years of attempts by the apartheid regime to keep linguistic groups apart. In addition, there are at least 25 tongues spoken in South Africa (Lemmer, 1993). The 11 official languages are, therefore, just a proportion of the full range of tongues. Thus, the argument that there will ever be a situation when all South Africans will be taught in their mother tongues is clearly presumptuous. As already argued, it makes more sense to talk about regional language education. In addition, recent research (ELTIC, 1997) has shown that although South Africans have generally welcomed the designation of official status of the major African languages through the 1996 Constitution, their use in education and workplace still lacks motivation. According to Slabbert and Finlayson (1998: 4), the feeling is ‘if you can speak it why should you learn it?’ As a result, parents would rather have their children ‘learn something useful and new such as English rather than a Bantu language’ (p. 4). Language varieties It is also noteworthy that the standard forms of African languages used in schools are based on the rural or regional standard forms. The younger generation is increasingly finding little or no appeal in the standard African languages (cf. Calteaux, 1995) which they often associate with ‘rural’ languages. Studies on the emerging urban forms of African languages done elsewhere in Africa (e.g. The Dilemma of the Mother Tongue 61 Banda, 1996 for Zambia) have shown that the younger generation finds more status and prestige in the urban varieties than the rural-based standardised languages. In this regard, the urban forms may turn out to be a threat to standard African languages in South Africa (cf. Slabbert & Finlayson, 1998). Teachers of African languages are in a dilemma about what to do with the situation. They are aware that the urban varieties can be used to make African languages popular, but at the same time they are wary about becoming unwitting agents of the demise of the standard form (cf. Slabbert & Finlayson, 1998). The gravity of the dilemma is echoed in the words of a Zulu teacher who is against the use of urban varieties, and had this to say: The children don’t speak standard Zulu. It is a serious problem. These [urban] languages kill Zulu. Do you want to kill Zulu? We are to enforce and maintain the standards. I am as a Zulu teacher totally against it. If these languages are successful, Zulu writers will be regarded as outdated. (Slabbert & Finlayson, 1998: 4) Another teacher commented, ‘It disturbs them [the students in the learning of the standards]. When they go out of the classroom it is the end of the standard language’ (Slabbert & Finlayson, 1998: 4). Clearly, these concerns have to be taken into account when drafting future language in education policies. My argument is that everybody is preoccupied with the threat of domination of English on African languages ignoring the danger posed by the urban varieties of these languages on the standardised forms. In time, these mushrooming urban forms of African languages may prove to be a more decisive threat to standard African languages than English. After all, English remains a foreign language in the minds of most black South Africans. The same cannot be said of urban varieties of African languages. In other words, before universal mother-tongue education in African languages is introduced, it would be necessary to address the growing gulf between the popular urban varieties and the standard form. This is also true if African languages are to be used effectively in additive bilingualism programmes. It is often very tempting to cite studies done elsewhere to support arguments for mother-tongue education for all in the new nation states. In South Africa, Cummins (1981) is widely cited in support of mother-tongue education. Even government’s postulations on additive bilingualism appear to be anchored in Cummins (1981). The problem here is whether such studies are applicable to a different context, particularly considering that South Africa, for example, has a multiplicity of mother tongues. The Canadian situation under which Cummins’ additive bilingualism evolved only had French and English to contend with. Europeans have their own bilingual programmes and did not have to copy from Canada (cf. Baetens-Beardsmore, 1992). The European bilingual programmes differ from country to country. The argument is that South Africa’s language in education policy should take into account the peculiarities of the country. In South Africa, like elsewhere in Africa, the choice of a particular ‘mother tongue’ has implications and ramifications that go beyond pedagogical-didactic considerations, as it effectively means elevation (‘favouritism’) of one ethnic group over the other(s), and thus increasing the possibility of ‘tribal’-language conflicts. Africa has already seen too many wars triggered by language-related 62 Language, Culture and Curriculum problems. Thus, bilingual programmes in South Africa need to take into account the question of language and ethnicity. Considering that black education collapsed with the enforcement of mother-tongue education during apartheid (cf. Lanham, 1996; Branford, 1996), it is not surprising that black parents associate poor education with their mother tongues, and educational excellence with English. It is easy to see why this is so. Those who went through EMOI in mission schools which accepted black students, or those in exile in countries such as Swaziland, Zambia and England are perceived to have got a better education than those who underwent Bantu education (Lanham, 1996), which at most prepared them to be houseboys or maids for their white masters. Therefore, it is not surprising that those advocating mother-tongue education in South Africa (cf. Alexander, 1995; Luckett, 1995; Heugh, 1995) are at pains to explain that their proposals should not be construed as calling for a return to apartheid education. Studies have shown that in South Africa varieties of English range from Indian English, black English, white Afrikaans English to white ‘mother-tongue’ English (cf. Schmied, 1991). This means there is also a question of which variety of English should be used as MOI. As already shown, data used in the present study suggest that apart from white Afrikaans, the other language groups would prefer their own variety of English. The implication of using different varieties of English in South Africa’s education system has not been fully investigated or appreciated, though it is common knowledge that in black and coloured schools teachers do not use MtE as far as pronunciation (and to some extent vocabulary and grammar) are concerned. I want to suggest that although the presence of black, coloured, Indian etc. Englishes has been noted, they have neither been studied and documented well enough, nor standardised for them to be of pedagogical value. In addition, these varieties, being mutually intelligible, do not warrant their own standard different from MtE. Models of bilingualism From the foregoing, the dilemma of choosing either mother tongues or English as media of instruction is all too apparent. On the one hand, there is research suggesting that mother-tongue instruction is superior to second-language medium of instruction (cf. Schmied, 1991). At the same time, the standard form of African languages is under threat from urban varieties. Yet there remains pressure, particularly from black learners and parents, for English medium of instruction (cf. Pluddemann, 1996; Smit, 1996; De Klerk, 1996). Thus, any move towards mother-tongue education would have to be handled cautiously. For instance, it would make little sense to strictly enforce mother-tongue education in South Africa given the prevailing situation where there is little or no educational material in African languages, and also when employers in South Africa use either English or Afrikaans. In this regard, given the current situation, black learners who trained in their mother tongues only would be both prejudiced against and disempowered economically and educationally. For comparison and illustration of the above argument, a good case is India. Krishnamurti (1990: 19) reports that India’s Education Commission’s (1954–1964) recommendation that mother tongues (regional languages) be used as a medium of instruction up to tertiary education, had the unintended effect of The Dilemma of the Mother Tongue 63 creating two streams of students, ‘those with the English medium having a definite advantage over the regional language medium students, both in employment and postgraduate education’. Consequently, English gained in stature as the demand for mother-tongue education plummeted to levels lower than before the proclamation (Krishnamurti, 1990). Therefore, for black South Africans in particular, what is required is to reconcile their scepticism about mother-tongue education with the need for excellence in classroom practice. At the same time, society has to be educated to the fact that in a multilingual society each language could have a role to play. Moreover, bilingual programmes must take into consideration ‘the attitudes and expectations of pupils, parents, and teachers, as well as the official proclamation on language policy’ (Pakir, 1993: 220). Clearly, in South Africa, pupils and parents favour EMOI, while government language policy and academics insist on education in which the mother tongues of all South Africans are safeguarded. Hence, the additive bilingualism policy. I have tried to argue that the government’s language in education policy, which should be part of a more general language planning policy, should ensure that each language plays a facilitative role rather than a displacive role. Currently, the mother-tongue concept and its use and acceptance in a bilingual programme appears to be taken for granted, resulting in English becoming an unequal partner. But a bilingual programme in South Africa will have to evolve out of its geo-political, socioeconomic and educational needs. Such a bilingual programme should not be ‘imported’ from Canada, the USA or Europe. Once transplanted, imported models can lead to chaos and the disadvantaging of certain groups (cf. Pakir, 1993). Moreover, as Baetens-Beardsmore (1992: 182) notes, ‘no single model of bilingual education is universally applicable and no single existing model should be transplanted to a totally different context’. The implication here is that a comprehensive bilingual model has to take into consideration not just the people’s views on the proposed languages of instruction, but also the socioeconomic and educational factors existing in a country. In short, benefits can only accrue from such a bilingual programme if each language is allowed and enabled to contribute its best to the education system (cf. Tickoo, 1993). In South Africa’s case, English, or more accurately, attitudes to English are proving to be a stumbling block to implementation of bilingual education. Government might do well to try to change some of these entrenched attitudes before embarking on something that might have negative repercussions. As to the actual classroom practice, the new Curriculum 2005, with its emphasis on innovation and creative classroom interaction, offers many possibilities, particularly for students coming from diverse linguistic backgrounds. But there is clearly a dearth in research on how Curriculum 2005 can be effectively integrated into government’s language in education policy of additive bilingualism. Conclusion Language in education policy should be viewed as part of a more general language planning policy. In this connection, it is ironic that while government policy, on paper at any rate, promotes the use of all South African languages, 64 Language, Culture and Curriculum politicians and other role models are more likely to use English than any other language at public functions. Lanham (1996) laments the fact that former President Mandela’s speeches have almost exclusively been in English. In 1999 Home Affairs Minister Buthelezi, a Zulu chief, and King Zwelithini of the Zulu addressed assembled Zulus on Chaka Commemoration Day in English. One would expect such an important day among the Zulus to be commemorated in Zulu (perhaps with English translation for commercial television and radio broadcasts). It is also not uncommon for opposition leaders from the National Party and the Afrikaner Freedom Front lamenting in English on South African television, the diminishing influence and use of Afrikaans in government and public life. In an article entitled ‘Language Problem Wasn’t Technical Defect’, The Cape Times (11 February 1998) had a story in which the then Western Cape Premier, a white Afrikaans L1 speaker from the National Party, was angry when he was asked a question in Xhosa in the provincial legislature. This is despite the fact that Xhosa is supposed to have similar status to English and Afrikaans in the province. His first excuse was that he did not understand Xhosa, but when told to use earphones, he claimed that they were not working despite the fact that other members were listening through earphones for translations. However, the Premier did not seem to mind when the question was repeated to him in English. Xhosa, English and Afrikaans are the official languages of the Western Cape Province legislature. The argument is that role-model speakers are unwittingly strengthening the position of English over the same languages they are purporting to promote. Contrast the South African situation with the Tanzanian one. After Nyerere proclaimed Ujamaa (self-reliance) through kiSwahili, the entire country interpreted (or misinterpreted) it to mean that they would be faced with kiSwahili in every aspect of their lives. What is interesting here is that the political, civic and government leadership were in the forefront in providing the motivation by adopting kiSwahili as the language for the offices and public life (Mushi, 1996). In contrast, in South Africa political, civic and government leaders appear to prefer using English, at least when they appear in South African media, particularly television, thereby defeating their own objective of language pluralism. The argument here is that as long as policy makers and those in places of influence appear to send mixed signals to ordinary people, the general change of attitude required for an effective bilingual language in education policy will remain illusive. Particularly, as long as the actions of policy makers and those in the corridors of power suggest that a particular language (in this case English) offers opportunity for education, job opportunities and accessibility to communication, economic, political and industrial success, then language policies enacted to promote other languages will be futile. It is hardly encouraging to say to black learners, ‘learn through your mother tongue because it is the language of your ancestors and it is the language of your culture’. Such an argument is becoming a less attractive prospect in this global economy than the need for status and socioeconomic mobility that is perceived to be offered by English. Given such a situation, it might require a great deal of motivation to make black children see the functional sense of learning through the medium of African languages, such as Zulu or Xhosa. Moreover, due to urbanisation, industrialisation and migration The Dilemma of the Mother Tongue 65 brought about by the money economy, the concept of ‘mother tongue’ is becoming blurred and needs to be redefined. This is particularly true considering the growing gulf between the standardised rural ‘classroom’ tongues used in education and the more versatile urban forms of the same languages that are in use in the ever-growing cities of South Africa. Recent studies show that the urban forms of African languages provide a formidable challenge to English (Calteaux, 1995; Slabbert & Finlayson, 1998). But their utility in education has either been ignored or dismissed under the guise of safeguarding the standard mother tongues. As long as the supposed mother tongues of South Africans do not form mutually inclusive partnerships (particularly in relation to English) in the education process of learners, prospects for effective additive bilingual programmes will remain elusive. Therefore, given the socioeconomic, cultural and geo-political situation obtaining in South Africa, English is set to be the language and preferred medium of education in South African schools and tertiary institutions for some time to come. Correspondence Any correspondence should be directed to Dr Felix Banda, Linguistics Department, University of the Western Cape, Private Bag X17, Belville 7535, South Africa (fbanda@uwc.ac.za). References Alexander, N. (1995) Models of multilingual schooling in South Africa. In K. Heugh, A. Siegruhn and P. Pluddemann (eds) Multilingual Education for South Africa (pp. 79–82). Johannesburg: Heinemann. Baetens-Beardsmore, H. (1992) On European models of bilingual education. In G. Jones and C. Ozog (eds) Papers Presented at the Conference on Bilingualism and National Development (1), 165–187. Brunei Darussalam, 9–12 December 1991. Brunei: Pusat Tecknologi Pendidikan, Universiti Brunei Darussalam. Banda, F. (1995) The ‘meaning-to-grammar hypothesis’ of Zambian English: Implications for classroom language instruction in multilingual context. Unpublished PhD thesis, Free University Brussels. Banda, F. (1996) In search of the lost tongue: Prospects for mother-tongue education in Zambia. Language, Culture and Curriculum 9 (2), 109–119. Branford, W. (1996) On English in South Africa. In V. de Klerk (ed.) Focus On South Africa (pp. 35–51). Amsterdam: John Benjamins. Calteaux, K. (1995) Standard and non-standard African language varieties in urban areas of South Africa. Main report for the STANON Research Programme, Pretoria, HSRC. Cummins, J. (1981) Empirical and theoretical underpinnings of bilingual education. Journal of Education 163 (1), 16–29. Curriculum 2005 (1997) Document on language, literacy and communication. Pretoria: Ministry of Education. De Klerk, V. (1996) Use of and attitudes to English in a multilingual university. English World-Wide 17 (1), 111–127. ELTIC (1997) Report. On multilingual education in South Africa. Pretoria: ESL Forum. Heugh, K. (1995) The multilingual school: modified dual medium. In K. Heugh, A. Siegruhn and P. Pluddemann (eds) Multilingual Education for South Africa (pp. 83–88). Johannesburg: Heinemann. Krishnamurti, B. (1990) The regional language vis-à-vis English as the medium of instruction in higher education: The Indian dilemma. In P. Pattanayak (ed.) Multilingualism in India (pp. 15–24). Clevedon: Multilingual Matters. Lanham, L. (1996) A history of English in South Africa. In V. de Klerk (ed.) Focus On South Africa (pp. 19–34). Amsterdam: John Benjamins Publishing Company. 66 Language, Culture and Curriculum Lemmer, E.M. (1993) On the black child and proficiency in the medium of instruction. In J. le Roux (ed.) The Black Child in Crisis (pp. 146–170). Pretoria: CTP Book Printers. Luckett, K. (1995) National additive bilingualism: Towards a language plan for South African education. In K. Heugh, A. Siegruhn and P. Pluddemann (eds) Multilingual Education for South Africa (pp. 73–78). Johannesburg: Heinemann. Mushi, S.L.P (1996) On problems linked to second language teaching in Tanzania. Language, Culture and Curriculum 9 (2), 133–147. Ndebele, N. (1987) The English language and social change in South Africa. English Academy Review 4, 1–16. Pakir, A. (1993) Making bilingualism work: Developments in bilingual education in ASEAN. Language, Culture and Curriculum 6 (3), 209–223. Pluddemann, P. (1996) Response to multilingualism: Language support in a Western Cape primary school. Unpublished Mphil. thesis, University of the Western Cape. Saravanan, V. (1993) Language and social identity amongst Tamil–English bilinguals in Singapore. Language, Culture and Curriculum 6 (3), 275–289. Schmied, J. (1991) English in Africa. London: Longman. Slabbert, S. and Finlayson, R. (1998) The future of the standard African languages in the multilingual South African classroom. Unpublished paper, University of the Witwatersrand. Smit, U. (1996) On the status, roles and attitudes to English in South Africa. English World-Wide 17 (1), 77–109. The South African Constitution (1996) Act No. 108. The South African Schools Act No. 84 15 November (1996) Government Gazette No. 17579. Vol. 1867. Tickoo, M.L. (1993) When is a language worth teaching? Native languages and English in India. Language, Culture and Curriculum 6 (3), 225–239 Winkler, G. (1997) The myth of the mother-tongue: Evidence from Maryvale College, Johannesburg. Southern African Journal of Applied Language Studies 5 (1), 29–41.