Int. J. of Educational Development 19 (1999) 351366 www.elsevier.

com/locate/ijedudev

Multilingualism in democratic South Africa: the overestimation of language policy

Vic Webb*
University of Pretoria, CentRePoL, Pretoria, 0002 South Africa

Abstract The article focuses on the increasing mismatch between South Africas stated ofcial policy on language and its gradually evolving linguistic realities. Whereas the countrys institutional documents (the Constitution and the national policies being developed) proclaim linguistic pluralism to be the national objective the country seems to be regressing to its pre-apartheid situation of monolingual practicea situation of English only. This phenomenon is illustrated with reference to language-in-education, particularly the issues of language of learning and teaching (medium of instruction), language study and language and certication. The article then debates three possible reasons for the inability of the Government of the country to take positive steps towards policy implementation, concluding with the view that language planning is not enough, and that the lack of a meaningful political will in the political (and educational) leadership of the country is the chief obstacle to giving substance to multilingualism in South Africa. 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Language policy; Multilingualism; Language-in-education; Language of learning and teaching/medium of instruction; Sociolinguistics; Language planning; Language management; Africa; South Africa; African languages; Bantu languages; English; Afrikaans

1. The problem: the mismatch between dream and reality1 The language stipulations of the South African Constitution (both the interim and the nal, 1996

* Corresponding author 1 In order to understand the problem to be discussed in this contribution, it is probably necessary, particularly for non-South African readers, to provide background information on the history of linguicism in South Africa, showing, especially, the use of language for control (divide-and-rule) and oppression. Obviously, however, space does not make this possible, See Webb (in prep.), chapters one and two, for more information.

constitution) were, probably, a surprise to most language planners internationally, in that 11 languages2 were declared to be ofcial languages at the national level. Despite the possibility that this decision was taken on the grounds of political expediency, it does form a logical part of the whole philosophy which underlies the constitution and its Bill of Rights, namely national integration and pluralism (unity within the diversity). The statu2 These are: Afrikaans, English, Ndebele, Northern Sotho, Southern Sotho, Swazi, Tsonga, Tswana, Venda, Xhosa and Zulu. See Webb (in prep., chapter two) for a sociolinguistic prole of these languages.

0738-0593/99/$ - see front matter 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 8 - 0 5 9 3 ( 9 9 ) 0 0 0 3 3 - 4

352

V. Webb / Int. J. of Educational Development 19 (1999) 351366

tory acceptance of multilingualism as a basic ingredient in the realisation of equity and the processes of democratisation and national integration, and the 11-language decision is thus a bold (and possibly unique) initiative to address the manifold challenges of a complexly multilingual and culturally diverse country. The South African Governments acceptance of the multilingual and culturally heterogeneous character of South Africa did not, of course, stop with the acceptance of a number of constitutional stipulations. Their commitment to realising their political philosophy is also expressed in additional, supportive institutions prescribed by the Constitution (such as the Commission for the Promotion and Protection of the Rights of Cultural, Religious and Linguistic Communities) and the further language planning decisions they took. The latter include: the appointment of the Language Task Group (LANGTAG) in 1995, which was asked to propose a framework within which a comprehensive national language policy could be developed, and the acceptance of its report in 1996 the decision to establish a national language management body, the Pan South African Language Board (PANSALB), and their support for the creation of lower-order language management structures (provincial language councils and language-specic committees in each of the nine provinces, and national language bodies) the development of multilingually-oriented policies in a number of state departments as well as the national and the provincial education departments, and the work done by the Department of Arts, Culture, Science and Technology, which has been tasked with language planning, which includes: expanding the former State Language Services into a National Language Service, creating the Telephone Interpreting System, resourcing an expansive Language Awareness Campaign and presenting seminars and workshops directed at the development of language policy. However, despite the Constitution, all these very imposing language planning decisions and the equ-

ally signicant supporting structures, very little seems to have changed in the language practice of South African communities. In fact, there are various signs that South Africa may be regressing to where it was before the apartheid era, and that it is becoming more and more monolingual in its public life. Consider, for example, the following proposals put forward by parliamentarians: 1. The proposal (since abandoned) by the Speaker of Parliament and the Leader of the Council of Provinces that English be the main language of parliamentary reporting (and therefore that the agendas of meetings, all resolutions adopted at meetings, and all parliamentary speeches be recorded in English in Hansard) 2. The proposal by the portfolio committee for Defence that English be the only language of the South African National Defence Force, and that English be the only language of orders, training, general communication, control and co-ordination (February, 1998) 3. The serious discussion of using only English as the language of records in the courts (February, 1998) Consider also the following decisions by parastatal bodies: 1. The decision by the Post Ofce to use only English for internal business, with English as the language of meetings, the minutes of meetings, memoranda, notices, letters, and even of advertisements of employment opportunities (April 1998) 2. The national telecommunication rms similar decision in May 1998 3. The strong bias towards English in the programmes of the South African Broadcasting Corporations television section, and 4. The decision by South African Airways to use only English on its internal ights (since being reconsidered) Similar decisions were made in the private sector: 1. The Landbank (now called the Land and Agricultural Bank of South Africa), who decided in 1997 that English would be its ofcial language, despite the fact that the majority of its staff

V. Webb / Int. J. of Educational Development 19 (1999) 351366

353

2. 3.

4.

5.

members were Afrikaans-speaking, and that 80% of its clients were also Afrikaans-speaking The Council of Real Estate Agents, who made a similar decision in 1997 The ABSA banking group (which was and still is primarily Afrikaans), who decided in 1997 to use English as language of internal business The editorial board of De Rebus, the ofcial journal of the Society of Lawyers, to switch to English The decision of the ofcial journal of nurses, Nursing World, to use only English as their ofcial language in 1997

If one compares these decisions with the Governments language planning intentions listed at the beginning of this paper, there seems to be a signicant mismatch between policy and ofcial practice (language choice, preferences and attitudes) in South Africa.3 This mismatch is not restricted to South Africa of course. The rest of Africa provides many similar examples. Except for the (temporary) successes in Tanzania (with their impressive promotion of Swahili) and Somalia (with the implementation of Somali as language of learning and teaching even in secondary schools), language policy implementation has not been very successful. Bamgbose (1991: 111121), for example, lists failures in the (former) Congo, Ghana, Kenya, Nigeria, Senegal, Sierra Leone and Zambia. These failures cannot be ascribed to a lack of dialogue, since language policy conferences have been held in Africa from soon after African states started becoming politically independent.4 (See the
3 There are, of course, signs of support for multilingualism: the major Bantu languages are used quite widely on national television, some ofcial documents are available in languages other then English, and there are serious projects directed at exploring the potential of multilingualism for formal educational purposes. Furthermore: the incidence of individual multilingualism is high, especially among black South Africans (with the majority of the rest either bilingual, with prociency in Afrikaans and English, or procient only in English). The conict between dream and reality in this contribution refers, therefore, to ofcial (government) language behaviour. 4 There were, of course, earlier meetings, like the meeting of the executive committee of the International African Institute in Rome in 1930, which issued a strong statement in support

list quoted by Bamgbose, 1991: 125127, which includes conferences in Nigeria in 1960, 1976 and 1977, Cameroon in 1970, Tanzania in 1971, and Ghana in 1968 and 1975.) Nor can the failures be attributed to the disinterestedness of African governments, since African Ministers of Education as well as the Heads of State have formally declared their support for policies directed at the greater utilisation of the African languages on a number of occasions (e.g. in 1976); and in 1986, more than 10 years ago, the OAU even adopted a Language Plan of Action for Africa. The failure of language policy implementation can also not be blamed on a lack of sociolinguistic information, since a number of valuable language reports have been available for many years, such as the ve reports on countries in eastern Africa. Finally, the ineffectiveness of language planning can also not be blamed on a lack of language development agencies. Cameroon, for example, has at least six major institutions involved with language planning activities, and has language planning committees for 50 of the local indigenous languages. (Chumbow, 1997) Despite all these formal decisions and all the supporting work by linguists, very little has changed on the African language political scene in any meaningful way. Before going on to a more detailed discussion of the mismatch between dream and reality in the domain of language and education in South Africa, it is necessary to emphasise that the criticism to be expressed is not based on the assumption that the present Government should have been able to undo 350 years of linguicism in four years, or that it is possible to effect the sociolinguistic transformation and reconstruction of South Africa within the same short period of time. However: given the clear negative signs mentioned above, and given the similar experiences in other African states, it is, I think, appropriate to give a critical evaluation of the language policy and planning situation in the country, particularly at this very early stage.

of the extended use of the African languages in education (Reference by Kashoki, 1993: 8)

354

Table 1 Comparison of home language (HL) of pupils with choice of rst language of learning and teaching (LoL/T) in all schools in ve provinces as percentages in 1997a Afrikaans English Pedi Sotho Tswana Ndebele Swazi Xhosa Zulu Tsonga Venda

Province (and total no. of pupils) 8.3 2 35.2 5.8 0.8 0.6 10.8 3 56.7 32.8 1.0 3.7 0.7 0.3 0.4 0.3 3.5 1.6 1.5 0.5 0.06 15.6 4 1.6 15 8.9 2.5 2.2 2.4 6.7 1.6 1.5 0.5 1.5 0.2 18.6 6

Gauteng (1.45M.)

V. Webb / Int. J. of Educational Development 19 (1999) 351366

HL LoL KwaZulu/Natal (2.7M.) HL LoL Mpumalanga (0.87M.) HL LoL Northern Prov. HLc LoL Western Cape (0.91M.) HL LoL

17.5 19 1.2 1.6 6 6.5 2.6 1.6 60.1 66

13 66 10.6 64.3 1.4 61 0.4 50.8 19.6 28

20 8.5 85 34b 16.3 16.8 0.8 0.6

3.6

1.2

22.7 6.9

11.8 6.3

a Ones interpretation of these statistics should reect an awareness of the following problems: that it is not certain what, exactly, is meant by rst choice of language of learning and teaching, that the second and third choice of language of learning and teaching is not available, that there is no distinction between primary schools and secondary schools, or the different grades, and that there is no distinction between the rural urban divide. b Relatively speaking the percentages for Zulu in KwaZulu/Natal and Pedi in the Northern Province are low. c Home language according to the 1991 census. Source: Education Ministry Information Service (EMIS), Education Department.

V. Webb / Int. J. of Educational Development 19 (1999) 351366

355

Table 2 Choice of language of learning and teaching in DET schools in the late 1980s Choice of language of learning An African language, i.e. a rst language English, a second language (primary schools) English, a second language (secondary schools) Afrikaans and English, both second languages Afrikaans, a second language (Source: DET Report for 1988) Pupil numbers 2 632 603 1 338 833 1 496 529 13 313 1278 Percentages 48.03 24.43 27.3 00.24 00.23

Table 3 Choice of language of learning and teaching in DET schools in 1991 in percentages and number of schools Choice of language of learning Percentages Number out of the 4912 schools who responded 368 1080 74 2652 185 658

An African language English from grade 1 Afrikaans from grade 1 Initially an African language but with a gradual transfer to English Initially an African language but with a gradual transfer to Afrikaans Initially an African language but with a sudden transfer to English Source: Heugh 1993 (adapted)

7.5 22 1.5 54 3.7 13.4

2. The position in language-in-education The mismatch between policy and practice is also found in the domain of language-in-education. This is true for the teaching and study of the African languages in primary and secondary education, but is particularly apparent in the case of the preferred language of learning and teaching. Consider, for example, the statistics shown in Table 1, for the choice of language of learning and teaching in South Africa for 1997. The dominance of English is clearly shown in the totally disproportionate relationship between English as home language and English as language of learning and teaching, and the inverse, equally disproportionate relationships in the case of the African languages. In the late eighties the position regarding the language of learning and teaching in the former Department of Education and Training (DET) schools (the schools for black pupils) is shown in Table 2. A survey of the choice of language of learning

and teaching in DET schools in 1991 is shown in Table 3. The three sets of gures are not wholly comparable, for obvious reasons. However, it is clear that there has been a decline in the choice of an African language as rst language of learning and teaching, and an accompanying increase in the choice of English: whereas an average of about 25% of the pupils in DET schools were taught in English in the late nineteen-eighties, more than 60% of the schools in the four provinces outside the Western Cape selected English as rst language of learning and teaching in 1997. This phenomenon, the increasing dominance of English in a country in which it is the second (or third) language of 90% of the inhabitants, is, surely, an unequivocal denial of the Constitution and its underlying philosophy. The South African Governments policy of multilingualism and pluralism seems to be moving in the wrong direction. Elsewhere in Africa the situation is not much different.

356

V. Webb / Int. J. of Educational Development 19 (1999) 351366

Oladejo (1993: 9395) discusses the contravention of language-in-education policy decisions in Nigeria. Despite a clause in the 1977 National Policy on Education, that the Government will see to it that the medium of instruction in the primary school is initially the mother tongue or the language of the immediate community, he writes, it is common knowledge that, in general, only English is used right from the primary level through to the tertiary. Oladejos conclusion is that it is no exaggeration to say that the Nigerian attempt to formulate and implement a coherent bilingual education policy has so far been an exercise in futility. A rather poignant illustration of the problem in Nigeria is supplied by the so-called Six-Year Primary Project of the University of Ife in Nigeria, discussed by Bamgbose (1984; 1998). This project was designed to compare the use of Yoruba as a medium of instruction for the full six-year duration of primary education for all subjects (except English) with the use of Yoruba as a medium for the rst three years and a change-over to English as a medium thereafter. (1984: 88) Provision was made for an experimental group and four other control groups for comparative purposes. The results of the project were quite denitive: The experimental group performed signicantly better than the other groups in Yoruba (naturally), in the subjects science, social and cultural studies and mathematics, and, also, in English. However, despite the strength of these ndings, and despite the fact that the Minister of Education at the time was himself a linguist, these ndings had no impact on the thinking of policy-makers in Nigeria. From these scenarios it is clear, as is generally known, that there is a strong preference in African countries for the ex-colonial languages, English, French and Portuguese. It is also generally accepted that this preference is educationally detrimental since the ex-colonial languages are not known well enough5 to allow for the full develop5 The poor knowledge of English among black pupils in rural areas is shown in the following ndings in a research project in KwaZulu (black) primary schools in 1984 (Odendaal):

ment of knowledge and cognitive skills. The effect of pupils inadequate competence in the language of learning and teaching in South Africa is reected in the statistics in Table 4. The poor educational performance of South African pupils is also reected in the percentage pass with endorsement (the requirement for acceptance for university study), which was as follows for 1997 in the ve provinces shown in Table 5. The average pass-marks in Mathematics and Science could not be ascertained. However, the average pass-marks for English and the African
Table 4 The comparative pass-rate for Science and Mathematics as subjects in the Std. 10 examination in 1993 as percentages Science Black candidates Coloured, Indian and white candidates 53 97 Mathematics 27 85

dard are you? That 21% of the teachers she interviewed were of the opinion that their pupils could not understand them when they used English, and That 83.5% of them thought that their pupils could not understand the textbooks which were used. One can, of course, argue that in the new educational context black pupils will be able to improve their competence in English drastically. This may happen, but it will not necessarily be the case. The Indian sociolinguist Debi Pattanayak, for instance, pointed out (in 1994) that attempts to teach English to the Indian population have been made for more than 200 years, yet only 24% of the Indian population know English today. There are still 400 million illiterates in India. In South Africa conditions for acquiring an adequate knowledge of English are generally not favourable. Besides the fact that the present education set-up is poor (overcrowded classrooms, absence of the basic facilities (desks, books), absence of a culture of literacy and learning), pupils (in the rural areas in particular) have very little personal exposure to English and probably also do not feel a serious need for it in their daily lives. In addition, the knowledge which black teachers have of English is often also inadequate. In this regard Odendaal found that 69.4% of the teachers in KwaZulu had obtained between 40% and 49% for English in their nal school examinations. Finally, from all reports the actual teaching practice in English classrooms is unlikely to change matters, since teachers generally seem to use a kind of grammar method, and seem unable to teach communicatively.

That Std. 3 pupils addressed individually in English could not understand questions such as: Where is your home? Have you come far? What does your father do? and In what stan-

V. Webb / Int. J. of Educational Development 19 (1999) 351366

357

Table 5 Percentage pass with endorsement in ve provinces Gauteng 17.3 KwaZulu/Natal 18.2 Mpumalanga 9.3 Northern Prov. 5.7 Western Cape 23.2 All provinces 12.5

Source: Education Ministry Information Service (EMIS), Education Department.

languages in the DET schools for 1993 were: 41% for English and 42% for the African languages. Given these examination results and the presumed role of English, the language of learning and teaching, as a factor in the poor performances, it is clear that serious attention must be given to the apparent inability of African governments, including the South African Government, to implement a policy of multilingualism in education6 in a meaningful way, that is, a policy which will lead to an appropriate and a far more meaningful role for the African languages in educational development.7 Why cant the South African Government (begin to) give substance to its philosophy of linguistic pluralism? This, then, is what the rest of this contribution wants to do: It wants to list some of the possible reasons for the conict between dream and reality, and it wants to discuss strategies for bridging the gap. It concludes with the statement that language policy in and of itself is not enough.

3. Reasons for the inability to realise language policy objectives There are several obvious reasons for the mismatch between policy and practice, such as the under-availability of nancial, human and educational resources. These factors, though applicable, are not insurmountable. They do not prevent the authorities from giving a clear, precise expression of their intentions or from starting the process in seriousness. It seems necessary to look deeper. In the relevant literature a number of possible explanations are given, for example by Bokamba (1993); Bamgbose (1991); Kashoki (1993); and the Working Group on Educational Research and Policy Analysis (1997, see especially pp. 4246). I would like to focus on three possible explanations for the situation, namely the sociolinguistic character of South Africa, the inadequate language policies, and the apparent lack of political will. 3.1. The sociolinguistic character of South Africa

In the policy documents of the South African Education Department additive multilingualism is listed as a central aim, yet the realities expressed by Tables 13 show that there has been very little movement in the direction of a multilingual situation. 7 It is, of course, not the use of the African languages as such (often inappropriately referred to as mother-tongue instruction), that will make the difference, but the use of a language which pupils know (very) well. There is ample proof that the cognitive, psychological and social development of an individual can only occur fully in a language that the individual knows well (See the Working Group on Educational Research and Policy Analysis, 1997: 3741). Such a language is also the language of creativity, which is an essential component in personal development.

The sociolinguistic complexity of South Africa is a well-known topic, and need not be discussed in any detail. The main features which impact directly on any language policy are the countrys linguistic diversity, the extreme politicisation of its languages, the exceptional social meaning of English and the low estimation of the African languages. In Mazruis words: Africa is an acute case of linguistic dependency. (Broader discussion in Webb, 1996, and in preparation.) The effect of these factors is naturally of great importance, and each needs to be addressed (for example by attempting to change attitudes to the

358

V. Webb / Int. J. of Educational Development 19 (1999) 351366

African languages8). All I would like to do is to emphasise the need for language policy developers to take direct note of the sociolinguistic diversity in the country, and to realise that different policy options need to be developed for different sociolinguistic situation types. In this regard attention must rst be drawn to an article by Nancy Hornberger (1987), which discusses the difculty of implementing a policy directed at the utilisation of indigenous languages in primary schools despite its demonstrable educational advantages. Hornberger demonstrates that a language-in-education policy which seems to have obvious advantages may be inappropriate if the sociolinguistic context is not adequately taken into account. Hornberger reports on a project on bilingual education in Puno in rural Peru in which Quechua (which had been kept on the margins of the national education system) was introduced as language of learning and teaching alongside Spanish in a dual-medium system. An important aspect of the project was that it was directed at the maintenance approach to bilingual education (rather than a transitional approach), with Quechua used as language of learning and teaching throughout the primary school. An additive approach to bilingualism was therefore followed rather than a subtractive one. Hornberger lists a number of successes with the project, including the development of the vernacular and the production of written material in it, the promotion of cultural integration, overcoming cultural discrimination, the reduction of illiteracy, and the better use of educational opportunities. Furthermore, there was also greater pupil participation in class-room talk, more meaningful reading behaviour, concepts and logic were stressed in arithmetic classes rather than step-following and memorisation, and generally there was more content-orientation and skills development. Clearly, therefore, a maintenance-directed bilingual education in this

rural Quechua community had clear educational advantages. However, despite the educational successes, the policy as such was not successful. There was considerable community resistance and a large number of schools eventually withdrew from the project. Though some of the reasons for the withdrawal were logistic (for example the transfer of teachers and their replacement by unequipped teachers), and others had to do with wrong decisions (the inclusion in the project of urban schools where Spanish was known), a major reason for the policy failure was the strong prejudice against the use of the vernacular languages in education. Hornberger points out that the community traditionally saw the school as a non-Quechua institution, in which Quechua did not belong, but where Spanish was appropriate. The Quechua case provides a number of pointers for language-in-education policy in South Africa. First, it will be difcult to introduce languages as languages of learning and teaching which have a low status and a low economic value, which have restricted usefulness, and whose speakers are decreasingly isolated (i.e. urbanised). A maintenance directed dual-medium approach may therefore not prove workable even though the conditions may seem favourable to it. Second, language-in-education policy must be supported by large-scale information campaigns, and thirdly, such policies should provide for the development of different policy implementation options for different situations: different courses for different horses, as it were. A single generalised policy statement is not wise.9 In the South African context, I would suggest that different policy options should be considered for each of the following sociolinguistic category types: 1. The larger urban areas (Johannesburg, Pretoria,

8 Language attitudes cannot be changed externally and directly, of course. They are basically attitudes towards people and as such they can only change if the social standing of the communities themselves increases. For this to happen the communities must be perceived to have become successful.

9 Recently, there has been a strong movement towards accepting a dual-medium model in South Africa (Heugh, 1998). Though this model is probably applicable in some urban contexts, it is not applicable in rural contexts where only the home language/community language is heard.

V. Webb / Int. J. of Educational Development 19 (1999) 351366

359

Durban, Pietermaritzburg, Port Elizabeth, Bloemfontein and Cape Town) where English and/or Afrikaans have strong dominating presences where a second language (Afrikaans, English or both) is the language of economic and social power where there is a high level of exposure to the second language and a strong incentive to learn it where the African languages generally do not have a strong sociolinguistic presence functionally or prestige-wise, but are nevertheless still signicantly present, albeit in different degrees (e.g. from ve in Pretoria and Johannesburg, to one in the other larger urban centres) where there are urban varieties of particularly the African languages (but also of Afrikaans and English), and where strong urban identities have developed and there is a resistance against traditional ways of life. 2. Smaller urban areas (Pietersburg, Potchefstroom, Rustenburg, Kroonstad, Ladysmith, Newcastle, Grahamstown, Kimberley, Mmabatho, Stellenbosch) and larger towns in rural areas (Potgietersrust, Middelburg, Bethlehem, Vryheid, Kokstad, Queenstown, Worcester) where Afrikaans generally has a stronger presence than English where people are not exposed to English as a second language to a signicant extent where the African languages may be far stronger statistically speaking where children are generally not signicantly literate, and where teachers are often not very welltrained or exceptionally procient. 3. The deep rural areas (Nongoma, Thohoyandou, Botshabelo, Engcobo, Bisho) where both Afrikaans and English are largely absent (even in the electronic media) where one African language is usually dominant where there is little subsantial (nonsymbolic) incentive to learn either English or Afrikaans

where children grow up in a largely illiterate environment where there is still a strong attachment to non-western cultural patterns, and where teachers are generally poorly trained and not adequately procient Clearly, a single policy on the language of learning and teaching cannot be made applicable in all three sociolinguistic category types. 3.2. Inadequate language-in-education policies The second possible reason for the mismatch between policy and practice, is that the proposed language-in-education policies may not be adequate as statements of policy. The national language-in-education policy of South Africa is described in the South African Constitution of 1996, the South African Schools Act of 1996, a language-in-education policy document and a proposed language-in-education implementation plan. The latter two documents were announced in July 1997 and were subsequently discussed at a national forum in May 1998. The South African Constitution stipulates (Section 29) that (2) Everyone has the right to receive education in the ofcial language or languages of their choice in public educational institutions where that education is reasonably practicable. In order to ensure the effective access to, and implementation of this right, the state must consider all reasonable educational alternatives including single medium institutions, taking into account (a) equity (b) practicability and (c) the need to redress the results of past racially discriminatory laws and practices The Schools Act determines, in summary, that: 1. The provinces must formulate their own language-in-education policies (subject, of course, to national policy) 2. The governing body of a (public) school determines the schools language policy (subject to any relevant provincial acts)

360

V. Webb / Int. J. of Educational Development 19 (1999) 351366

3. The use of language policy for racial discrimination is prohibited, and 4. The use of a language prociency test for the purpose of admission to the school is also prohibited The provisions of the 1997/98 language-in-education policy documents of the central Department of Education can be summarised as follows: 3.2.1. Aims The Education Department declares itself committed to three aims, namely (a) the promotion of multilingualism (which means developing the 11 ofcial languages, establishing respect for all the languages of the country, and developing the skills needed for communication across the barriers of colour, language and region), (b) countering the disadvantages resulting from the mismatches between home languages and languages of learning and teaching, and (c) building a non-racial nation and contributing to citizens full participation in society and in the economy. 3.2.2. The language of learning and teaching policy In this regard the policy makes four statements: 1. It accepts the principle of additive multilingualism 2. It accepts any ofcial language as possible language of learning and teaching 3. It stipulates that learners (in practice: parents and school authorities) select their language of learning and teaching on entering a school, accepting that their choice may differ from the language policy of the school concerned, and 4. It describes the bases upon which disputes about the language of learning and teaching must be handled (the principles of rights, equity, redress and practicability), as well as the way in which such disputes must be resolved (consultation with the Member of the Executive Council for Education for the province, the Pan South African Language Board and an appeal for arbitration) In the case of a language of learning and teaching being requested which is not available in a parti-

cular school, the policy states that such a request will be regarded as reasonably practicable if it is supported by the learner/teacher ratio and costeffectiveness, that is, if it is requested by 40 learners in grades 1 to 6, or 35 in grades 7 to 12. The policy documents do not recommend any specic language of learning and teaching model, but do suggest that two such models are considered practical: the use of a rst language as language of learning and teaching (with the requirement that an additional language also be studied), and a structured bilingual approach, by which they presumably mean dual medium schools (since they say that this model implies a two-way immersionparagraph 4.1.5.) 3.2.3. Language study The Education Departments policy on the study of languages is formulated as follows: 4.4.2 From Grade 3 onwards, all learners shall offer their language of learning and teaching and at least one additional approved language as subjects; 4.4.4.2 From Grade 5 onwards, one language must be passed; and, 4.4.4.3 From Grade 10 to Grade 12 two languages must be passed.

3.2.4. Multilingualism The policy expresses quite a strong position on multilingualism. It obligates schools to promote multilingualism by requiring of them that they stipulate how they will do it, and it suggests that they (i) use more than one language of learning and teaching and/or (ii) offer additional languages as fully-edged subjects and/or (iii) apply special immersion or language maintenance programmes (particularly in cases where learners home languages are not used as language of learning and teaching). Schools are urged to share human resources. 3.2.5. Policy implementation assessment The policy documents also contain directives for assessing policy implementation. The Department proposes that language managers be appointed in school districts and that these ofcials monitor the

V. Webb / Int. J. of Educational Development 19 (1999) 351366

361

language policy implementation of schools actively, for instance with questionnaires to stakeholders, and with interviews. They also propose that the Pan South African Language Board be approached to assist in the implementation assessment process. It is clear, I think, that the language-in-education policy and the plan of implementation contain a number of positive features, also from the point of view of the Governments philosophy of pluralism. It is also clear, however, that they possess several rather serious shortcomings. This is particularly clear if one evaluates them from the perspective of what is usually understood by the notion policy. Human (1998: 1502) characterises policies as precise and detailed statements. They must, he says, explicitly state the goals they wish to achieve, clearly indicate how these goals are to be achieved and specify the tasks which need to be performed. They must also be based on appropriate factual information (preferably collected after a strategic analysis), they must be based on a thorough understanding of all the relevant variables, and they must specify the resources required to realise their aims, the skills needed and the necessary structures. They must therefore specify what must be done, how, by whom, with what and when.10 Policies are mission statements and must contain a detailed list of tasks to be performed in order to realise the vision the political leaders have for the country as a whole. (Human, 1998: 150152.) (See also the Working Group on Educational Research and Policy Analysis, 1997: xivxv.) Given this conception of policy, given the aims the Education Department has set for itself regarding language-in-education, and given the sociolinguistic character of the country, the proposed policy is not adequate, despite its positive features. This
10 Human (1998: 56) illustrates this aspect of policy development with reference to raising the literacy level in a society. A policy aimed at such a goal needs to specify who the target community is, how many candidates will be accommodated, how many instructors will be involved, where and when the training will take place, for what language the training will provide, what training material needs to be available/developed, how long the literacy programmes are to last, how their success will be determined and what budget is required for the programme.

can be illustrated with reference to a number of issues: 3.2.6. Language policy development by governing bodies Though it lists the principles to be followed in formulating school language policy, it gives no explicit direction on how the governing bodies are to go about developing their language policies. Whilst the philosophy of individual choice and the devolution of decision-making accords nicely with a democratic approach to policy-making, it is essential that decision-makers be enabled to make informed choices. The two documents, however, give no indication that the Department intends doing this. Given the politicised nature of the language situation this is an essential requirement, and a guiding document or localised information sessions are necessary. 3.2.7. Language of learning and teaching policy Given the strong stand on multilingualism in the policy documents one would expect a clear and comprehensive directive on the language of learning and teaching. This is not the case. On the contrary, the actual policy statement on the language of learning and teaching is unexpectedly general and inexplicit. It reads, simply, that The language(s) of learning and teaching in a public school must be (an) ofcial language(s) (paragraph 4.5). There is also an unexplained ambivalence concerning the notion of multilingualism, expressed in the shifting use of the terms multilingualism and bilingualism. Compare, for instance: to promote multilingualism (4.1.1) vs. additive approach to bilingualism (4.1.5) vs. to establish additive multilingualism as an approach to language in education (4.3.2) and the promotion of bi- or multilingualism (5.1.1.1.2). This ambivalence creates the impression that the Department is not sure whether a multilingual approach is feasible, thus reecting negatively on the philosophy of multilingualism. Given the sociolinguistic realities of the country and the strong support for multilingualism one would have expected a stronger stand on the language of learning and teaching. In its present form the

362

V. Webb / Int. J. of Educational Development 19 (1999) 351366

policy will most probably contribute very little to changing the realities expressed by the statistics of Table 1. In fact, its weakness could even lead to increased subtractive bilingualism. 3.2.8. Languages of study The policy position on the study of languages is equally disappointing, since it requires the successful formal study of only one language up till the ninth year of school (when the majority of school children exit from the system), and thereafter only two languages (one of which has to be the language of learning). One is tempted to see it as an explicit contradiction of its own stated aim of promoting multilingualism11 and as a measure which, in fact, strengthens the monolingual practice which the Government gradually seems to be allowing to become reality.12 3.2.9. General shortcomings Besides these specic shortcomings the policy documents are also decient in the following ways: 1. They give no direction on the language(s) to be used by school authorities in their communication with educators, learners and parents, which means that an opportunity to provide credibility for the African languages as working languages in ofcial contexts could be lost 2. They pay no attention to the dire need for additional language learning programmes in the language of learning and teaching for learners whose knowledge of the selected language of learning and teaching may be inadequate 3. Though the documents raise specic issues (inadequacies of the African languages: par. 5,

11 It is reasonable, I think, to dene multilingualism as a phenomena involving at least three languages. White, coloured and Indian South Africans, who usually know only two languages at most, cannot be regarded as multilingual, at least in the (South) African context. (Multilingualism obviously involves more than just the number of languages people know. It also has a qualitative dimension, as discussed in Webb, 1998.) 12 Unconrmed information from KwaZulu/Natal schools in the Durban area, where Afrikaans is almost non-existent, indicates that the (two) languages of study selected by secondary school pupils are Afrikaans and English. The reason for this selection relates, most probably, to their market value.

p. 11 of the second document) they make no reference to vital issues such as: changing the negative attitudes to the African languages, providing specialist teachers of English in every school, countering the hegemony of English, the provision of educational material in these languages, improving language teaching methods (in rst language teaching, second language teaching and foreign language teaching) 4. They display a disturbing lack of linguistic insight into the nature of human languages. They state, for example, that the African languages are somehow semantically and syntactically inadequatepresumably in technical usage. It is also not careful enough in its reference to the need to develop and expand academic, scientic and technological vocabulary. The view they express is an oversimplied view which unintentionally (?) lends support to the common objection against the African languages, namely that they cannot be used as languages of learning and teaching in the higher grades because they do not have the necessary technical terminology, and hinting that this may not happen soon because it takes such a long time to create ones own technical terms. 5. The documents demonstrate a confusion in the minds of their developers about the language planning philosophy which is appropriate to language-in-education policy development. The preamble of the policy document suggests that the Education Department wishes to approach the language issue from the perspective of language as a resource: it speaks of additive multilingualism, the equality of languages and the promotion of the African languages. However, the vagueness and indecisiveness of the policy statement itself suggests that in reality the Education Department has thought about the language issue from the perspective of language-asa-problem, or, at best, as a (restricted) right. The African languages, they seem to believe, are only considered important for personal, cultural and, maybe, religious use. 6. They fail to commit the Department to supplying human and material resources for the implementation of the policy, and so pave the

V. Webb / Int. J. of Educational Development 19 (1999) 351366

363

way for a possible failure in policy implementationas was demonstrated by the inability of the Pan South African Language Board to achieve anything of any signicance over the two years of its existence. The South African language-in-education policies in larger or lesser degree display the same weaknesses as those listed by Bamgbose (1991): 111) as typical of the language policies in African countries: avoidance, vagueness, arbitrariness, uctuation, and declaration without implementation. Avoidance in the policy is found in the number of escape clauses, such as reasonably practicable, reasonable alternatives, equity, and practicability. Vagueness is found in the fact that some of its basic elements (like multilingualism) can be interpreted in different ways. The stipulations are too exible, with too few details of exactly what should be done. Equally, there is no clarity on implementation procedures, in particular: Who has to do what? Which body? When? How? A negative reading of the documents can easily lead one to deduce that the policy is a declaration without serious intent. Lack of specicity, says Bamgbose (p. 117), effectively gives governments an alibi for non-implementation. It is necessary, I think, to keep in mind that the South African language political situation, as elsewhere in Africa, is not situated on an equal playing eld. In order to achieve the goals set out in the South African Constitution strong steps need to be taken. And these steps need to be prescribed in the language policies of the country. In this regard it is helpful to take note of the views of Wielemans and Berkhout (in press) about the role of policies in transforming public life. They emphasise that policies are not merely simple sets of linear events (for example problem identication, possible solutions, implementation strategies and evaluation procedures) nor a collection of authoritative statements indicating future directions. Policies are, in reality, instruments in the exercise of power and are directed at establishing certain values in a community (p. 2). Policies should not only be seen as things but also as processes and outcomes, not only as text but also as discourse and effect.

The criticism of the Education Departments language-in-education policy expressed in this contribution should not, of course, be interpreted as a plea for Government prescriptiveness or strong-arm strategies. Nor does it wish to suggest that policies should have the power of laws. As indicated in the concluding paragraph of this contribution school communities have to be allowed free choices and the policy implementation process has to be consultative by nature. The criticism does, however, want to suggest that policy documents must be far more explicit, with clear descriptions of the available options and their consequences, with a full exposition of the information needed by school communities to make decisions which will be in their own interests in the long run. The politicised nature of the language situation in the country requires a far stronger leadership role from the Department so that parents and schools can make informed decisions. 3.3. Lack of political will The tendencies towards increasing monolingualism in South Africa, listed at the beginning of this contribution, suggest that there may be a lack of meaningful political will among the political leadership of this country to seriously implement the national ideals expressed in the Constitution, and that the publicised policy of pluralism may be little more than a symbolic gesture or a strategy to obtain public support, without being intended to lead to any real change in the South African society. It may even be that the elite closure mentality which existed in Kenya according to MyersScotton, also exists here. One must not forget that political policy-makers are apt to serve their own interests, and that policies are often designed to control society and to maintain power. (See Weinstein 1990 and Bokamba 1993.)13
13 One of the ndings of the Working Group on Educational Research and Policy Analysis was that language attitudes was a basic factor in the non-implementation of language-in-education policy (p. 46). This is undoubtedly true, but I would argue that these attitudes are themselves a consequence of factors which lie even deeper, and have to do with the lack of political will among the communities themselves, as well as among the political leadership.

364

V. Webb / Int. J. of Educational Development 19 (1999) 351366

In studying the implementation/non-implementation of policy, it is necessary to pay direct attention to the political agendas of politicians, and to keep Kashokis views (1993: 12) in mind: political will (is) the primary ingredient in utilising the factor of language in any meaningful way in the process of national development. In this regard it is useful to take note of the example of Afrikaans. It is true that the political history of Afrikaans provides ample examples of how language should not be used in a political context (for example its appropriation by its white speakers, and its use for political manipulation and gain).14 But it does, also, give a useful illustration of the central role of political will in the promotion and development of a language. The promotion of Afrikaans did not begin with sociolinguistic fact-nding, the determination of the social status of the language, the development of language planning skills, or any of the other tasks which language planning scholars usually propose as necessary pre-planning steps (see Chumbow, p. 10). The rise of Afrikaans began in the eastern Cape in the mid-19th century as part of a political movement (of separatism from the colonial government), and in the western Cape as part of a religious movement (of translating the bible into a language which members of the working class could understand). These movements were not led by government leaders. On the contrary, the English-speaking Government as well as the Dutch-speaking cultural leaders were opposed to the use of Afrikaans in public life. The process was driven by community leaders, such as newspaper journalists, teachers and a few church leaders. One of the rst (necessary) outcomes of the
14 It is necessary, of course, to consider the full picture of the hegemonisation of Afrikaans. The eventual rise of Afrikaans as a language of power is directly related to the military and economic power of the white descendants of the Dutch colonists. Later, in the second half of the 19th century in the two Boer republics, and from 1948 nationally, the white speakers of Afrikaans also obtained political power, which automatically led to the hegemonisation of Afrikaans. (Ironically, the direct association of Afrikaans with the policy of apartheid from the early fties onwards contained within it the germ of its own eventual demise as a primary language of ofcial life.)

movement was the development of a sense of socio-political identity, which developed into a driving force and a will to promote the language, in and of itself. The consequence of the development of this political will was that a written form of the language was developed, that norms and standards were established, that language promotion agencies were created, that literature was published, and, eventually, that Afrikaans-dominant economic institutions came into being. The government of the time was not involved in these actions. The Afrikaans-speaking community took the task upon themselves to develop the language as an educational, economic and political tool. The same type of process is necessary in the case of the African languages.

4. Conclusion The task which a democratic government in South Africa has to perform, namely the social and structural transformation of the country, is enormous. Performing such a task requires a fundamental change in the attitudes of people, their values, norms, beliefs and aspirations, it requires rebalancing the existing asymmetric power relations in the country, and bringing about the educational, economic and political development of the people of the country. This task is made even more difcult by the social, political and economic forces operative in South Africa (such as the centralization of executive power, urbanisation and globalisation). One cannot, of course, expect signicant changes within four years of government. However, the Government is the prime agent in the transformation process, and it must give a clear and unambiguous lead in a programme directed at establishing the social and institutional power of the African languages. Strong, pro-active measures will be required. I would like to end this contribution with a list of the types of strategies which I think will be required: 1. Explicit policies and plans, with a clear specication of the appropriate alternative options and the sociolinguistic situation types in which they are appropriate

V. Webb / Int. J. of Educational Development 19 (1999) 351366

365

Policies can be effective instruments of power, but to be effective in this way, they need to be specic, and need to provide clear indications about the different possible options and the contexts in which each option is appropriate (for example, options such as the single-language model, the dual medium model with all its possible congurations,15 the parallel medium options, the multilingual model, etc.) In each case the situational appropriateness of each model should be described, so that school communities can make an informed choice. 2. Strong government structures and institutions, with the necessary resources The Pan South African Language Board, which has been in existence for almost two years, has achieved almost nothing of signicance up till now. (In fact, it has only just been able to establish itself administratively, with its own ofces and own administrative facilities). The reason for its ineffectiveness lies with its lack of particularly material resources. Maybe this lack of material resources is an indication of the Governments lack of seriousness about the language issue. If language policies are to succeed state structures need to be appropriately empowered. 3. A pro-active policy formulation and implementation approach is necessary. In a politically and linguistically complex situation as we have in South Africa, with all its deeply embedded imbalances, a vigorous approach is needed. Non-negotiable in this regard is the comprehensive provision of information campaigns and awareness campaigns in all the school communities of the country.

4. The process of national and local consultation obviously needs to be continued. The major reason for such a strategy is that it will lead to informed and involved school communities. In a post-modernist, complex society, policies cannot be imposed, and their implementation must be handled in a dynamic and interactive way. It is reasonable to accept, though, that teachers (the central instruments in policy implementation) who have been allowed to become part of the process of policy implementation, will facilitate the realisation of educational goals. 5. The need to establish a primary reward system It is essential to support the promotion of the African languages by providing incentives for their use in more than just the primary domains of life, such as personal interaction, religion and cultural activities. School communities should be rewarded for opting to use these languages as languages of ofcial work and as languages of learning and teaching, teachers should be rewarded for signicant contributions to the use of the African languages in these functions, public administrators should receive benets (appointment, promotion) for acquiring African languages, and so on. The basic aim of this paper was to evaluate the role of policy in social transformation. In this regard Id like to refer to a statement by the Cameroonian sociolinguist, Sam Chumbow, that the absence of a comprehensive language policy in Cameroon is the main reason for the neglect of the indigenous languages of the country, and that language planning is not only a major condition for development, it is also the most fundamental guarantee for a true evolution in the direction of development and democracy in countries of the South. I would like to differ from Chumbow in some degree, stressing that (a) language planning does indeed have a role to play in the government of a people, but (b) that its role must not be over-estimated. (c) It is likely that the major condition for development and the most fundamental guarantee for a true evolution in the direction of development and democracy rather lies with the

15 A dual-medium approach could be implemented rst, by using the rst language as the primary language of learning and teaching, with a second language a subject with a specialised teacher, followed by the gradual transition to using the second language as co-language of learning and teaching in the higher grades of the primary school, and progressing to the use of both languages as language of learning and teaching for different subjects.

366

V. Webb / Int. J. of Educational Development 19 (1999) 351366

presence or absence of a political will among the people of a country and its leadership, and that policies, including language-in-education policies, are direct reections of the presence or absence of this political will.

References
Bamgbose, A., 1984. Mother-tongue medium and scholastic attainment in Nigeria. Prospects 14 (1), 8793. Bamgbose, A., 1991. Language and nation. The national language question in sub-Saharan Africa. Edinburgh University Press, Edinburgh. Bamgbose, A., 1998. Language as a resource: An African Perspective. In: Ambrose, M., Read, J., Webb, V. (compilers), Workshop on the Role of the African Languages in Democratic South Africa. CentRePoL, University of Pretoria, Pretoria. Bokamba, E., 1993. The politics of language planning in Africa: critical choices for the 21st Century. Conference paper, Linguistics Society of Southern Africa, University of Pretoria. Chumbow, B.S., 1997. Language Planning and National Development: The case of Cameroon. Manuscript. Heugh, K., 1993. Not so straight for English. Bua! 8 (2), 31. Heugh, K., 1998. The new language in education policy: South African Perspectives on Implementation. Paper presented at the Education Departments conference on language policy and implementation, Pretoria. Hornberger, N.H., 1987. Bilingual education success, but policy failure. Language in Society, 16, 205226. Cambridge University Press. Human, P., 1998. Yenza. OUP, Cape Town.

Odendaal, M., 1986. Die milieu binne die skool. [The environment inside the school.] Conference paper, Human Sciences Research Council, Pretoria, February. Oladejo, J.A., 1993. How not to embark on a bilingual education policy in a developing nation: the case of Nigeria. Journal of Multilingual and Multicultural Studies Development 14 (1/2), 91102. Kashoki, M.E., 1993. Some thoughts on future language policy for South Africa vis-a-vis the Language Plan of Action for Africa. Paper presented at the LSSA Conference, University of Pretoria. Webb, V., 1996. Language and politics in South Africa. In: Dua, Hans Raj (Ed.). Language and politics. International Journal of the Sociology of Language, vol. 119. Berlin: Mouton de Gruyter. Webb, V., 1998. Mutilingualism as a developmental resource: Framework for a research program. In: Nkonko M. Kamwamgamalu (Ed.), Aspects of multilingualism in post-apartheid South Africa. Special issue of Multilingua, 17 (2/3). Mouton de Gruyter, Berlin. pp. 125154. Webb, V., in preparation. Language in South Africa. The quest for a future. The role of language in national reconstruction and development. Weinstein, B., (Ed.), 1990. Language Policy and Political Development. Ablex, Norwood, New Jersey. Wielemans, W., Berkhout, S.J., in press. Towards understanding education policy: an integrative approach. To appear in Educational Policy. Working Group on Educational Research and Policy Analysis, 1997. Languages of Instruction. Policy Implications for Education in Africa. International Development Research Centre. Data sources: Education Ministry Information Service (EMIS), Pretoria; the Education Foundation, Johannesburg, and the Centre for Education Policy Development, Evaluation and Management, Johannesburg.