35.1 Introduction

It has always been felt by African educationists that the African child’s major learning problem is linguistic. Instruction is given in a language that is not normally used in his immediate environment, a language which neither the learner nor the teacher understands and uses well enough (Obanya 1980, 88).

If the African child’s major learning problem is linguistic, then all the attention of African policy-makers and aid to the education sector from donors should be devoted to a strengthening of the African languages as languages of instruction, especially in basic education. My own experience after having taught in Africa for 4 years and having visited hundreds of classrooms both in east and West Africa is that Obanya is completely right; the African child’s major learning problem is linguistic. Children are being branded as unintelligent when they lack knowledge of the language used in instruction, a language they often hardly hear and seldom use outside of the classroom. The concept “education for all” becomes a completely empty concept if the linguistic environment of the basic learners is not taken into account (Brock-Utne 2000, 2001, 2012a, b, 2013a, b; Klaus 2001; Rugemalira 2013).

Yet there is hardly another socio-cultural topic one can begin discussing with Africans that leads to so heated debates and stirs up so many emotions as that of the language of instruction in African schools. It is difficult to discuss this topic as a strictly educational question phrased for instance as: “Through which medium of instruction would children learn subject matter best?”; “If the aim is to master a ‘world’ language, would it be better to have that language as a language of instruction at the earliest time possible or to develop the vernacular or a commonly spoken national language further first?”; or “What does it mean for the learning potential, the development of self-respect and identity that the language one normally communicates in does not seem to be deemed fit for a language of instruction in school?” Brock-Utne 2013a, b; Brock-Utne and Qorro 2015).

When it comes to the choice of language of instruction in African schools socio-cultural politics, economic interests, sociolinguistics and education are so closely interrelated that it is difficult to sort out the arguments. It is an area with strong donor pressure, mostly from the former colonial masters, who wish to retain and strengthen their own languages. There are strong economic interests from publishing companies overseas who see that they will have easier access to the African text-book market when the Euro languages are used. There are also faulty, but widely-held beliefs among lay-people when it comes to the language of instruction. In a five-year research project, LOITASAFootnote 1 (Language of Instruction in Tanzania and South Africa) which I am conducting together with partners in Tanzania and South Africa we have come across many of these beliefs.

Some years back I was sitting for several hours in the back of a classroom in a secondary school in Tanzania. I observed students who did not understand what the teacher was saying when he spoke English, and who had to ask each other what the teacher said and sometimes ask the teacher to express himself in Kiswahili, a language they all spoke very well. When I spoke in Kiswahili to one of these students afterwards and mentioned that I had noticed that he did not understand the language of instruction, he admitted that my observation was correct. He did have great difficulties following the teacher, especially if the teacher did not switch to Kiswahili during the lesson when he saw that the students did not understand. When I asked him if it would not have been much better for him had the lesson been given in Kiswahili throughout, he admitted that it certainly would have been much easier. He would then have been able to understand. However, when I then asked him whether he thought the language of instruction should be changed, he said that he did not think so because English was the language of technology and modernisation. English was the global language without which one could not get a good job. He believed that he had to learn English and could not see another way than having it as a language of instruction. (Brock-Utne 2013a, 2017).

We shall return to this argument. In this chapter I wish to revisit and critically discuss some of the commonly heard arguments against the strengthening of the African languages as languages of instruction. Some of the arguments are promoted by Western donors and Western academics, others by Africans. One argument often heard is that there is such a multitude of languages in Africa that it would be impossible to choose which language to use. It is therefore better to retain the colonial languages. Another argument is that it is too costly to publish textbooks in these languages. Some African parents, school-children and lay public claim that children need to study in an ex-colonial language as early as possible in order to get the best possible command of that language. This is supposed to further personal development, the earning potential of the child and the development of his family, society and country. There is a tendency that even in a country like Tanzania, where more than 95% of the population are fluent in the national language Kiswahili, and where Kiswahili is the official medium of education all through primary school, that the new private schools in Dar es Salaam advertise that they are English medium primary schools. These are schools where parents who are somewhat better off send their children and where school fees are charged. These schools are better resourced and teachers are better paid. Kiswahili is not only the national language of Tanzania but also one of the two official languages, Rugemalira (2013) notes that if a national language shares its official position with a ‘foreign’ language, then a justification for the situation ought to be provided in a language policy that also sets out the roadmap for the eventual removal of that foreign language from its usurped position. He states that there has been no such policy in Tanzania.

Towards the end of the chapter I shall discuss the coping strategies African teachers use in their classrooms.

35.2 The Myth of the Many Languages of Africa

In 2001 I had the pleasure of being invited to a conference held to mark the creation of a centre for African languages in Bamako, Mali. There were only three researchers from outside of Africa, all knowledgeable in African languages. The rest of the participants were sociolinguists and linguists from all over Africa, east and west, north and south. One of the keynote speakers was the renowned sociolinguist and sociologist Kwesi Kwaa Prah. Prah is originally from Ghana but has for many years worked as a professor at the University of Western Cape in South Africa. He was for many year the Director of the Centre for Advanced Studies of African Society (CASAS) in Cape Town and is still connected to the centre which is now (2020) located at the University of the Western Cape.

Professor Prah began his keynote speech by quoting some of the Western linguists with their different estimates of the numbers of African languages. While for instance David Westley (1992) claims that at least 1400 languages are spoken in Africa in 51 countries, Barbarba Grimes (1992) assesses the number of languages in Africa to be 1995Footnote 2. He asked: What is this? Don’t we know how many languages we have in Africa? Who has classified them? Who has put them into writing, for what purpose? According to what system? To what effect? He went on to say that he would now read aloud a list of African languages and he wanted everybody present to raise their hand if they heard a language mentioned by him that they could communicate comfortably in. This language need not be our first language, but a language we understood well and felt comfortable using. When he had read out a list of 12–15 core languages (a core language is a cluster of mutually intelligible speech forms which in essence constitute dialects of the same language) all of the participants in the conference had their hands up. These core languages included Nguni; Sesotho/ Setswana; Kiswahili; Dholuo; Eastern Inter-Lacustrine; Runyakitara; Somali/Rendile/ Oromo/Borana; Fulful; Mandenkan; Hausa; Yoruba; Ibo and Amharic. He characterised these languages as the first order languages of prominence. Below these, there may be about six which are not so large, in terms of speakers, but which have significant numbers of users. The work of the Centre for Advanced Studies of African Society (CASAS) over the past 5 years has revealed that as first, second and third language speakers about 85% of Africans speak no more than 12–15 core languages. (see e.g., Prah 2000; Prah 2002). This is actually fewer languages than the number of core languages spoken in the much smaller continent of Europe.

The truth is that the demographics of language and linguistic diversity in Africa are not really different from what obtains in other parts of the world. The myth of the multitude of languages seems to fit well into a description of Africa as the dark and backward continent. It is also a convenient excuse for donors backed by strong publishing interests in the West to use when they insist that one of the colonial languages has to be used as language of instruction.Footnote 3

What may be different in Africa from in other parts of the world is that the identification of linguistic units in Africa tends to be loose. The identification of language communities in Africa has been approached in a way, which favours the recognition of practically all dialects, and phonological variations as separate languages. When in 1995 I made a study for the Namibian Ministry of Education on the situation of the African languages after independence (Brock-Utne 1995, 1997), to my great surprise I discovered that the two main “languages” in the north of Namibia, Oshindonga and Oshikwanyama actually are the same language. The reason why there are two written forms of the language has to do with rivalry between Finnish and German missionaries and later the creation of separate language committees which suited the divide and rule policy of the apartheid government.

Roy-Campbell (1998) describes how faulty transcriptions, some arising from inaccurate associations by missionaries occurred across the African continent, resulting in a multitude of dialects of the same language and different languages being created from what was one language. The difficulty of putting a definite figure to the number of African languages on the continent can be attributed to this process, as contention has arisen over whether certain language forms are indeed languages or dialects. Sinfree Makoni (2000) likewise writes about the crucial role of missionaries in the specification of speech forms subsequently regarded as African languages. African missionary converts played the role of laboratory assistants. They provided the vocabulary and the missionaries the orthography and the grammar. Makoni claims that the grammar books made did not aid any meaningful communication between English and Shona speakers in Zimbabwe. The phrases for translation and the vocabulary used reflected settler and missionary ideology. The phrases were useful for talking about Africans but not for engaging with them in any egalitarian communication. He further writes about the way in which different missionary stations magnified differences between dialects, obscuring the homogeneity of the real situation. To quote him: “Missionaries were not sin-free in their creation of African vernaculars (Makoni 2000, p. 158)”. He concludes a chapter on the Missionary influence on African vernaculars in general and on Shona in particular with these words:

It is generally well known that the spread of “European languages” was one of the consequences of European imperialism. What is less well-known, however, is the effect of the work of missionaries in the construction of African languages. The written African languages which they created were “new” in many respects (Makoni 2000, p. 164).

In a keynote speech he gave at the opening of the LOITASA project, Kwesi Kwaa Prah (2003) also spoke of the harm done to African languages through the missionary settlements, as well as missionary rivalry and evangelical zeal. Missionaries lacking a proper understanding of the language transcribed any speech form they heard into a written language resembling the way similar sounds were transcribed in their own indigenous language. By this approach dialects like Cockney, Tyneside, broad Yorkshire, etc. in Britain could easily be made into languages in themselves. This fragmentation approach is still popular with the Summer Institute of Linguistics (SIL), a leading group in the work of rendering African languages into script, in order to translate the Bible into African languages. One can agree with Kwesi Kwaa Prah’s claim that the rendition of African languages into scripts for purposes of the development of Africa cannot at the same time proceed with the fragmentation of languages as is being conducted by the SIL. In effect, the SIL is building and destroying at the same time. To quote Prah (2002, p. 13):

When one asks why this is the case, the reason that comes easily to the fore is that the object of such endeavours at rendering African languages into script is not in the first instance to help in the development of Africa, but rather simply to translate the Bible into African speech forms and to evangelise and convert Africans into Christians. Unless one assumes that converting Africans to Christianity represents development. All other considerations are for such purposes insignificant.

Kwesi Kwaa Prah (2002) notes that those who write about the multitude of languages in Africa have, in most instances, never looked at African societies outside the framework of colonial boundaries. It is necessary for African linguists to work across national boundaries because practically all African languages are cross-border speech forms which defy the colonially inherited borders. When the colonial powers divided up Africa between themselves in Berlin in 1884 they never considered the language borders of Africa. Working within the framework of African neo-colonial borders creates many more problems than working across borders.

Prah claims that the sentimental glories of neo-colonial flags and national anthems maintain the fragmentation process of African languages. For the sake of flag and so-called national identity, Kamuzu Banda of Malawi refused to accept the reality of the fact that ciNyanja and ciCerwa are the same language. Sometimes these tensions are perceptible in the same country and represent attempts to own and control linguistic turf. In Ghana, 25 years after the harmonisation of Akan to produce a unified Akan orthography, writers still persisted in using the pre-unification orthography that separated mutually intelligible dialects like Akuapim, Asante, Fanti, Akim, and Brong.

The approach of CASAS is to organise the technical work on the harmonisation of orthography and the development of common spelling systems of African languages. When this has been successfully done, workshops are organised so that the new system is being taught to writers and teachers who then produce materials using the new orthographies. Many such workshops have already been organised by CASAS. The target of CASAS in the short run is to complete work within the next few years on the 12–15 core languages. The logic of this work is that once this approach runs its course, it should be possible to produce materials for formal education, adult literacy, and everyday media usage for large readerships which on the economies of scale make it possible to produce and work in these languages. According to Prah (2002, p. 15) “it is the empowerment of Africans with the use of their native languages, which would make the difference between whether Africa develops, or not”. In a book edited by Prah and myself (Prah and Brock-Utne 2009) the twelve contributors probe the thinking and practice of language of instruction policies in contemporary Africa. The thrust of the discourses in this volume is to move away from the colonially inherited positions.

The argument about the many languages in Africa is often being used to strengthen the ex-colonial languages. The claim is being made that if for instance the majority African language of the area were chosen as the language of instruction it would disadvantage children speaking a minority language, therefore everybody should instead use an ex-colonial language. If one looks closer into this claim, one will often find that children from minority groups will find the African majority language much easier to use as a language of instruction than the ex-colonial language. For instance in Zambia children from Bemba-speaking families showed greater communicative competence in Nyanja than in English. Likewise children from Nyanja-speaking families showed greater communicative competence in Bemba than in English.

35.3 It Is Too Expensive to Publish in African Languages

This is another argument we often hear and which is being used to promote the ex-colonial languages and publishing companies in the West. When it comes to Africa, it certainly would be too expensive to publish textbooks all through primary and secondary school as well as in tertiary and adult education in between thousand and two thousand languages, which is, as mentioned, the number some Western linguists give for the number of languages in Africa. But through inexpensive desk top publishing techniques it should be possible in Africa, as it has been in Papua New Guinea (Klaus 2001), to have African children study through their mother tongue in the first years of schooling. At the same time they should be taught a regional, cross-border African language which comes close to their mother tongue which they can use as the language of instruction at higher levels of learning. Through the harmonisation process that e.g. CASAS is working on it should be possible to concentrate on 12–15 languages that are understood by at least 85% of the African population and to have these languages being used as languages of instruction at the highest level of teaching. This would also mean an intellectual revival of Africa since it is only when text-books are published on a large scale that publishing companies have money to publish fictional and especially non-fictional books.

When economists try to figure out how much it will cost to publish text-books in African languages, they also have to figure out how much it costs to have African children sit year after year in school, often repeating a class without learning anything. The African continent abounds with examples of the low pass rate and high attrition rate in schools. I concur with the socio-linguist Zaline Makini Roy-Campbell (2000, p. 124) who has done extensive work in Tanzania and Zimbabwe when she writes:

What is often ignored is the cost to the nation of the continued use of European languages which contributes to the marginalisation of the majority of the population. One cannot overstate the damage being effected upon the psyche of African children being forced to access knowledge through a language in which they lack adequate proficiency and upon the nation which produces a majority of semi-literates who are competent neither in their own language nor in the educational language.

The Cameroonian sociolinguist Maurice Tadadjeu (1989) in his well argued and interesting book Voie Africaine argues for a three language model for Africa whereby everybody first learns to master his/her mother tongue, then learns a regional African language that can be used as a language of instruction in secondary and tertiary education and then learns an international language as a subject, a foreign language.

35.4 The Ex-Colonial Languages as the Languages of Modernisation, of Science and Technology

We shall return to the argument from the secondary school student I interviewed in Tanzania.

He looked at English as a language he had to master to get anywhere in the world. In actual fact there are not many Tanzanians who need English in their daily lives as all communication outside of the classroom is either in vernacular languages or in Kiswahili. Kiswahili is the language spoken in Parliament, in the lower courts, in the radio and television, in the banks, the post office and in the Ministries. There are more newspapers in Kiswahili than in English and they sell much better. But let us assume that this student would belong to those who would need a good command of English in his future career. I shall argue here that it would be better both for his knowledge acquisition in general as for his learning of English if the normal language of instruction would have been Kiswahili and he would have learnt English as a subject, as a foreign language by teachers who were English language teachers, had both a good command of English and of the children’s first language, knew the methodology of teaching English and were interested in language acquisition. One of the Tanzanian participants in the LOITASA project group, dr.Martha Qorro (2002), is herself a Senior Lecturer in English in the Department of Linguistics and Foreign Languages at the University of Dar es Salaam. The reason why she is a great promoter of the use of Kiswahili as the language of instruction in Tanzanian secondary schools has to do with the fact that she, as an English teacher, has seen that children neither learn English (they learn bad and incorrect English) nor subject matter. The English language has become a barrier to knowledge.

In the English language newspaper the Guardian in Tanzania the editors started a Kiswahili medium debate in the spring of 2002. In an editorial of 30th April 2002 the editor openly warns Kiswahili medium advocates. On May 29th Martha Qorro gave a substantial answer to the editor based on her own observations and research. Here are some quotes from her answer:

In terms of language use in public secondary schools in Tanzania most students and the majority of teachers do not understand English. For example, the headmaster of one of the secondary schools once admitted that, of the 45 teachers in his school only 3 understood English well and used it correctly. This in effect means that the other 42 teachers used incorrect English in their teaching. This is not an isolated case. Those who have been working closely with secondary school classroom situations will agree with me that this situation prevails in most public secondary schools in Tanzania.

Dr. Qorro claims that it is the prevailing situation in the secondary schools in Tanzania, where most teachers teach in incorrect English, that forces her to argue for the change of medium of instruction to Kiswahili. She feels confident that students can, in fact, learn English better than is currently the case when it is taught well as a subject, and eliminated as the medium of instruction. In her own words:

The use of English as a medium actually defeats the whole purpose of teaching English language. For example, let us suppose that, in the school mentioned above the 3 teachers who use English correctly are the teachers of English language, and the other 42 are teachers of subjects other than English. Is it not the case that the efforts of the 3 teachers of English are likely to be eroded by the 42 teachers who use incorrect English in teaching their subjects? If we want to improve the teaching and learning of English in Tanzania secondary schools, I believe, that has to include the elimination of incorrect English to which students have been exposed from the time they began learning it (Qorro 2002).

In her article Martha Qorro argues for the elimination of incorrect English by not using it as a medium of instruction. She knows that many people are put off by this suggestion because of the belief that by using it as a medium of instruction students would master English better. Though she agrees that mastering English is important she feels that the best way to do this is through improved teaching of English language as a subject and not to the use of English as a medium. And then she adds:

Not everyone who recommends a change of medium of instruction to Kiswahili is a Kiswahili Professor. I for one am not a Kiswahili Professor, I have been teaching English for the last 25 years, and to me a change to Kiswahili medium means:

  • Eliminating the huge amount of incorrect English to which our secondary school students are exposed.

  • Enhancing students’ understanding of the contents of their subjects and hence creating grounds on which they can build their learning of English and other languages.

  • Eliminating the false dependence on English medium as a way of teaching/ learning English, addressing and evaluating the problems of teaching English.

  • Impressing on all those concerned that English language teaching is a specialised field just like History, Geography, Physics, Mathematics, etc. It is thus unreasonable and sometimes insulting to teachers of English when it is assumed that teachers of all subjects can assist in the teaching of English.

To the young secondary school student I met in Tanzania it would be correct to reply yes it is important for you to learn English and learn it well, but there is reason to believe that you will learn the language better if you study it as a foreign language, as a subject. Using Kiswahili as the language of instruction will help you learn science and other subjects better than you do now.

Ferguson (2000) points to several research studies showing that those students who learn in their own language do better in school. He refers to a study by Prophet and Dow (1994) from Botswana. A set of science concepts was taught to an experimental group in Setswana and to a control group in English. They then tested understanding of these concepts and found that Form I students taught in Setswana had developed a significantly better understanding of the concepts than those taught in English. A similar study with the same results has been carried out in Tanzania. Secondary school students taught science concepts in Kiswahili did far better than those who had been taught in English (Mwinsheikhe 2001, 2002).

35.5 Inside the African Classroom

How do African teachers, who often do not master the language of instruction themselves very well, behave in the classroom? What teaching methods do they use? What coping strategies do they employ? The chorus teaching you often hear in African classrooms owes itself much to the fact that the teacher does not have a vocabulary large enough to employ an interactive teaching method. It is difficult to use an interactive teaching strategy when you do not command the language well. Observations that I have made in Tanzania both in secondary school classrooms and when I have taught university students show that if a teacher attempts to engage her/his students in group work or group discussions the groups will immediately switch into Kiswahili. Most of the time the teacher will either use what Heller and Martin-Jones (2001, p. 13) have called “safe talk” or will code-mix or code-switch in the classroom.

35.6 Safe Talk

Heller and Martin-Jones (2001, p. 3) define “safe talk” as:

Classroom talk that allows participation without any risk of loss of face for the teacher and the learners and maintains an appearance of “doing the lesson”, while in fact little learning is actually taking place... This particular style of interaction arises from teachers’ attempts to cope with the problem of using a former colonial language, which is remote from the learners’ experiences outside school, as the main medium of instruction.

Rubagumya (2003) found in a study he made of the new English medium primary schools in Tanzania the main manifestation of “safe talk” in an encouragement of chorus answers from pupils, repeating phrases or words after the teacher and copying notes from the blackboard. He found very little encouragement of pupils to freely express their ideas without the teacher’s control.

The two examples below are taken from Rubagumya’s research and illustrate “safe-talk” as observed in many classrooms of the sample schools.

  • T : So you have positive fifty-five plus positive what now?

  • PP: (chorus) ten

  • T : Positive ten. What do you get then?

  • P : (one student answers) Positive sixty-five

  • T : Sixty-five positive. How many got that? Only one… any question?… no ques tion. Do this exercise

  • [Maths std. 3 school D2]

In this example, the teacher is going through the exercise he had given pupils earlier. After making the corrections, he asks how many pupils got the right answer. Only one out of a class of 35 pupils had the right answer. The teacher then asks whether pupils have any questions. After a very brief moment (about 2 s) he decides that pupils have understood, and he proceeds to give them another exercise. Here both the teacher and the pupils are practising “safe talk”. Since only one pupil got the right answer, we would have expected several pupils to ask some questions. But they hesitate because they don’t want to lose face. The teacher on his part waits for only about 2 s and proceeds with the next task. He doesn’t want to encourage pupils to ask questions because either this might expose his lack of fluency in English, or because he is trying to cover the syllabus. Either way, this is ‘safe talk’.

  • T : number twelve… let us go together... one two three

  • PP : (chorus) The doctor and his wife has gone out

  • T : The doctor and his wife has gone out... Kevin?

  • Kevin : The doctor and his wife have gone out

  • T : The doctor and his wife have gone out... is he correct?

  • PP : (Chorus) YEES!

  • (English std. 2 school A3)

Here the teacher is trying to correct the pupils when they say “the doctor and his wife has”. Kevin gets the right answer “the doctor and his wife have”. Once the other pupils confirm this as correct in a chorus, the teacher does not care to explain why the right form of the verb is have and not has. There is no way he can find out from the chorus answer whether every pupil understands the difference between have and has, but accepting the chorus answer is “safe” both for him and for his pupils.

35.7 Code-Mixing and Code-Switching in the African Classroom

The young student I talked to in Tanzania told about the difficulties he had understanding the lesson if his teacher did not translate the English words for him from time to time. This is what most Tanzanian teachers teaching in secondary schools do. In their classrooms they use strategies we term code-mixing, code-switching or regular translations. When the word code is used here it simply means different languages.

In the research project sponsored by the Norwegian Research Council my research assistant/collaborator and I have decided to use the following definitions of code-switching and code-mixing:

Code-switching refers to a switch in language that takes place between sentences, also called an intersentential change, code-mixing refers to a switch in language that takes place within the same sentence also called an intrasentential change.

Code-mixing is generally looked at more negatively than code-switching. Code-mixing often indicates a lack of language competence in either language concerned. Code-switching does not necessarily indicate a deficiency on the part of the speaker, but may result from complex bilingual skills (Myers-Scotton 1993). Code-switching is a strategy a teacher even with good command of English (if that is the language of instruction) may use when s/he sees that his/her students do not understand. It is a strategy often used by teachers who are knowledgeable in the first language of students. From observations I have made so farand by analysing observations made by other researchers it seems to me that the strategy code-mixing is mostly being used by teachers who are not language teachers and do not have a good command of the language of instruction.

35.8 Examples of Code-Mixing

In the example below the geography teacher mixes in English words in his sentences but lets the important words be said in Kiswahili. The following excerpt is taken from classroom observations made in a Form I geography lesson:

  • T: These are used for grinding materials. It looks like what?

  • S: Kinu (pestle)

  • T: Kinu and what?

  • S: Mtwangio (mortar)

  • T: It looks like kinu and mtwangio and it works like kinu and mtwangio.

In this example the teacher is satisfied with the answer from the student which shows that the student has the right concepts. The fact that these concepts are expressed in Kiswahili does not seem to bother the subject matter teacher, who does nothing to expand the vocabulary of the student within the English language. From the excerpt we do not even know whether the teacher knows the correct terms in English. Even if s/he does, s/he does not bother to make his/her students partake of this knowledge. Had the teacher insisted on an answer in English, s/he would most likely have been met by silence.

Observations that Osaki made in science teaching in secondary schools in Tanzania have made him reach the following conclusion:

Students either talk very little in class and copy textual information from the chalkboard, or attempt discussion in a mixed language (i.e., English and Kiswahili) and then copy notes on the chalkboard in English... teachers who insist on using English only end up talking to themselves with very little student input. (Osaki 1991).

As all educators know, student input is essential for learning. In an experiment one of my doctoral students Halima Mwinsheikhe (2001, 2002) conducted as part of the research on my project and in the connection with her Master thesis she had teachers teach some biology lessons solely through the medium of English, and later had the same teachers teach some other biology lessons solely through the medium of Kiswahili. She tells that during the experimental lessons one could easily see that teachers who taught by using English only were exerting a great effort not to succumb to the temptation of code mixing or switching. They seemed to be very tense and their verbal expressions were rather “dry”. Those who taught in Kiswahili were much more relaxed and confident. Those who taught through the medium of Kiswahili also seemed to enjoy teaching. They found it easy to make the lessons lively by introducing some jokes.

It is not only when teachers are to teach students that the language of communication becomes a problem. Halima Mwinsheikhe (2003) tells that after her study for the master degree and her return to Tanzania she felt compelled to probe further into the issue of Kiswahili/English as LOI for science in secondary schools. Whenever she found herself among teachers and/or students she observed and sought information/opinions regarding this issue. She tells how in May 2002 she co-facilitated a training workshop for science teachers of the SESS (Science Education in Secondary Schools) project together with an American Peace Corp. The main objective was to train the teachers on the use of participatory methods to teach/learn some topics on Reproductive Health. She relates:

The intention was to conduct the workshop in English. However, it became evident that the low level of participation, and the dull workshop atmosphere prevailing was partly due to teachers’ problem with the English language. This is not a very shocking observation considering that some of these teachers were students some four years ago. The workshop co-ordinator and I agreed to use both Kiswahili and English. The problem was immediately solved. Since we started with this mixture, the working atmosphere was good, lively and conducive to learning. The other workshop co-ordinator was well aware of the language problem in secondary classrooms in Tanzania. …An interesting observation is that my co-facilitator, an American, who had been in Tanzania for only 18 months, used Kiswahili rather well in teaching a science subject intended for secondary schools!

Halima Mwinsheikhe sees the observations made during this particular workshop as a cause for concern because in the final analysis the language problem of the teachers involved will impact on students during teaching/learning experiences. The implication is that teachers will most likely opt to use Kiswahili to surmount the existing language barrier. And yet at the end of the day students will be required to write their test/examinations in English.

35.9 Examples of Code-Switching

The examples of code-switching and code-mixing reported here will be taken from Tanzania and South Africa, the countries in which my research project was located. The same practice has, however, been observed in classrooms in Uganda, Swaziland, Namibia and Burundi (see e.g., Ndayipfukamiye 1993). In Tanzania Kiswahili is used as the language of instruction through primary school while English is supposed to be used as language of instruction in secondary school and institutions of higher learning (except in some Teacher Colleges for primary school teachers where the language of instruction is Kiswahili). Despite of what may be regarded as a very progressive language in education policy in South Africa, which in principle enables learners or their guardians to choose any of the 11 official languages as the language of instruction; English is used as the medium of instruction from grade 4 in primary school onwards. The transition to English is only a policy decided by individual schools and reflects the actual 1979 apartheid language policy. When one reads the official government policy carefully, one sees that this policy does not state that a change of language of instruction needs to take place in the fourth or fifth grade in primary school or, for that matter, at all. According to this policy the whole of primary school as well as secondary school could be conducted in African languages as the languages of instruction (Brock-Utne and Holmarsdottir 2003).

In connection with the South African part of my research project a reading comprehension task was given to 278 students in 6 different classrooms in three schools in Cape Town. The overall results showed that students who received instruction only in isiXhosa and the task in isiXhosa performed far better than those who received the same task in English only. In addition, the isiXhosa group also out performed the group that was given the task in English, but where instruction was given in both English and isiXhosa i.e., code-switching method. All sessions were videotaped and analyzed as part of the doctoral thesis written by Holmarsdottir (2005).

Teachers in African classrooms know that they are not allowed to code-mix or code-switch, yet most of them still do. Halima Mwinsheikhe, who has worked as a biology teacher in Tanzanian secondary schools for many years, admits:

I personally was compelled to switch to Kiswahili by a sense of helplessness born of the inability to make students understand the subject matter by using English (Mwinsheikhe 2001, p. 16).

In the following passage the science teacher changes languages completely as he sees that his students do not understand. His own English is not easy to understand. He expresses himself much clearer and better in Kiswahili. For him the important thing is to get the subject matter across. He is a teacher of science, not of English.

  • T: When you go home put some water in a jar, leave it direct on sun rays and observe the decrease of the amount of water, have you understood?

  • Ss:(silence)

  • T: Nasema, chukua chombo, uweke maji na kiache kwenye jua, maji yatakuaje? (I say take a container with water and leave it out in the sun, what will happen to the water?)

  • Ss:Yatapungua (it will decrease)

  • T: Kwa nini? (Why?)

  • Ss:Yatafyonzwa na mionzi ya jua (it will evaporate by the sun’s rays)

In the example above the teacher, after his initial try in English and the following silence from the students, switches completely to Kiswahili.

In South Africa it is also assumed that by using English in all content subjects students will in turn become more proficient in English. Also here teachers of students from the black majority population generally code-switch or code-mix during most lessons. In this case research shows that although officially the language of instruction is English the actual language used most in the upper primary school classrooms in the western Cape where we conduct our research is isiXhosa.. The following example highlights the use of both English and isiXhosa (code-switching) in a South African classroom.

During a mathematicsFootnote 4 lesson at the grade four level the teacher was explaining to the students 20 + 19, which she had written on the board. At first the teacher made an attempt to explain the lesson in English, but quickly switched to isiXhosa after realising that the students were not following along. During the explanation of this lesson in addition the teacher proceeded as followsFootnote 5:

  • T: We are now going to do the addition together and I will explain and you will follow along. We are breaking up the numbers. Do you understand?

  • Ss:(Silence, no one responds).

  • T: Siyacalula ngoku, siyawaqhekeza la manani. Sithatha bani phaya (We are simplifying now, we are breaking these numbers. What do we take from there)?

  • Ss:Utwo (two).

  • T: Sithathe bani phaya (And what do we take from there)?

  • Ss:Uone (one).

  • T: Utwo ujika abe ngubani (two changes into what number)?

  • Ss:Abe ngu-20 (becomes 20).

  • T: Right, u-1 lo ujika abe ngubani (Right, this 1 becomes what)?

  • Ss:Abe ngu-10 (Becomes 10).

The entire lesson was carried out in isiXhosa except for the initial attempt to use English only. The teacher switched languages after receiving no response from the students when she initially used English only. The remaining mathematics lesson then continued in isiXhosa with only some minor code-mixing taking place like the insertion of words like “right”, “okay”, “understand” and so on. The book the teacher was working from was in English.

The mathematics lesson described above is not an isolated case and in fact many of the lessons observed during the fieldwork in the South African part of the project were conducted mainly through the medium of isiXhosa. However, at the end of the day students are expected to use English for all the writing that is done in the subjects, except for the subject isiXhosa.

They are also expected to answer all exam questions in English.

National examiners working for the National Examination Board of Tanzania have told me of the many times they have seen students answer examination questions correctly, but in Kiswahili. The examiners were instructed to give such students zero points because the answers were supposed to be in English.

35.10 Conclusion

The situation that African teachers are forced into is tragic. Their own limited command of the language of instruction, as well as the great difficulties their students have understanding what the teacher is saying when s/he expresses him or herself in the ex-colonial language, force them to use teaching strategies I have here characterised as safe talk, code-mixing and code-switching. This gives the teachers a bad conscience since they know that they are not supposed to code-switch or code-mix but to use the ex-colonial language throughout the lesson. They also know that at exam day students who code-switch will be punished.

The ideal situation when it comes to classroom learning in Africa to me seems to be the three language model so well argued for by Maurice Tadadjeu (1989). As mentioned this language model for Africa would mean that the students first learn to master their mother tongue, then learn a regional African language, that can be used as a language of instruction in secondary and tertiary education, and then learn an international language as a subject, a foreign language. This would mean that the local language would be used as a language of instruction during the first grades while lessons in the regional language would also be given. The regional language would gradually become the language of instruction through secondary and tertiary education. The so-called international language would be taught as a subject from the time the regional language takes over as the language of instruction.

While we are waiting for the ideal situation to happen teachers must be allowed to code-switch because this speech behaviour is sometimes the only possible communicative resource there is for the management of learning. Learners should be awarded full points for a correct answer on exam questions whether they express themselves in the local, regional or foreign language.