International Journal of Bilingual Education and Bilingualism
ISSN: 1367-0050 (Print) 1747-7522 (Online) Journal homepage: http://www.tandfonline.com/loi/rbeb20
Codeswitching Among Multilingual Learners in Primary Schools in South Africa: An Exploratory Study S.O.S. Ncoko , R. Osman & K. Cockcroft To cite this article: S.O.S. Ncoko , R. Osman & K. Cockcroft (2000) Codeswitching Among Multilingual Learners in Primary Schools in South Africa: An Exploratory Study, International Journal of Bilingual Education and Bilingualism, 3:4, 225-241, DOI: 10.1080/13670050008667709 To link to this article: http://dx.doi.org/10.1080/13670050008667709
Published online: 26 Mar 2010.
Submit your article to this journal
Article views: 553
View related articles
Citing articles: 16 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rbeb20 Download by: [The Library, University of Witwatersrand]
Date: 22 January 2016, At: 05:05
Codeswitching Among Multilingual Learners in Primary Schools in South Africa: An Exploratory Study S.O.S. Ncoko
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Gauteng Department of Education, P.O. Box 7710, Johannesburg, 2000, South Africa
R. Osman Department of Education, University of the Witwatersrand, Private Bag 3, Wits 2050, South Africa.
K. Cockcroft Department of Psychology, University of the Witwatersrand, Private Bag 3, Wits 2050, South Africa Codeswitching is an important interactional resource in South Africa’s multilingual and multicultural society. This article discusses the incidence of codeswitching in pr imary schoo ls and examine s the speake rs’ mot iva tions for employing codeswitching. The data is drawn from conversations in both formal (classroom) and informal (playground) situations and the discussion is informed by current theoretical frameworks in codeswitching research. The implications of codeswitching for education in South Africa are considered.
Introduction South Africa has moved from a bilingual past, with English and Afrikaans as the only two official languages, to a multilingual future, with 11 official languages, namely sePedi, seSotho, seTswana, siSwati, tshiVenda, xiTsonga, English, Afrikaans, isiNdebele, isiXhosa and isiZulu. South Africa’s official African languages fit into four major groupings, namely Nguni, seSotho, xiTsonga and tshiVenda. The Nguni languages are made up of isiZulu, isiXhosa, isiNdebele and siSwati. These languages are the most common in the urban areas of South Africa. The Sotho group includes seSotho, sePedi and seTswana. While this group of languages is also common in the urban areas, they are not spoken as frequently as the Nguni languages. xiTsonga and tshiVenda are less popular in the urban areas and speakers of these languages living in urban areas will often learn a Nguni or Sotho language in order to be assimilated into urban townships. XiTsonga and tshiVenda are quite different in their spoken and written forms, unlike the Nguni and Sotho languages which are very similar in these respects. For example, in the Nguni languages the word for ‘language’ translates as ‘ulwimi’ in isiXhosa, ‘ulimi’ in isiZulu, ‘ilimi’ in isiNdebele and ‘lulwimi’ in siSwati. In the Sotho languages, it translates as ‘puo’ in seSotho and seTswana and ‘polelo’ in sePedi; whereas in xiTsonga it translates as ‘ririmi’ and in tshiVenda it translates as ‘luambo’. So, it is relatively easy for speakers of any of the Nguni languages to communicate with one another, and likewise for speak1367-0050/00/04 0225-17 $10.00/0 International Journal of Bilingual Education and Bilingualism
225
©2000 S.O.S. Ncoko et al. Vol. 3, No. 4, 2000
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
226
Bilingual Education and Bilingualism
ers of the Sotho languages. XiTsonga and tshiVenda are regarded as separate language groups, because of the differences between them and the other language groups. Researchers like Herbert (1992) have argued that the similarities within the Nguni and Sotho language groups are so great that the division of these groups into separate languages is artificial and that each group is actually one language with different dialects. However, in Apartheid South Africa these divisions were crucial to the successful implementation of separate development for separate people in separate bantustans or homelands. In keeping with this policy, language under the Apartheid regime, was used explicitly as a divisive tool. The constitution of South Africa (1996) guarantees language freedom to every member of South African society, by stating that ‘all official languages must enjoy parity of esteem and must be treated equitably’ (South African Constitution, 1996: 4). Makoni (1999: 144) states that the Constitution, by recognising 11 official languages, unintentionally creates ‘the linguistic equivalent of bantustans’, where languages are understood as ‘hermetically sealed units’. Rather, the interconnectedness of languages should be recognised, which is particularly evident in codeswitching practices. In addition to the acknowledgement of more than two official languages, South Africa’s first democratic election in 1994 also brought about changes in the demography of schooling. Previously, educational provision was divided along racial lines. Now, these divisions have been blurred with the introduction of ‘mixed’ or integrated schools. Today, many urban South African schools are made up of learners with diverse linguistic and cultural backgrounds thereby markedly changing the linguistic and cultural composition of these schools. The Language in Education Policy (Department of Education, 1997) allows schools to determine their own language policy in consultation with parents and the school community. It states that all children have a right to learn in their mother tongue and that the school must fulfil this right where practical and reasonable (own italics). This in effect means that if the school’s policy is to teach in English and there is a isiZulu learner whose parents would like him to be taught in isiZulu, the school has no obligation to this learner, if this is not practical for the school. In such a school then, the language of learning and teaching will be English. The policy also states that no child should be excluded from a school on the basis of his/her language. Desai (1994) predicted that the clause of ‘being practical and reasonable’ would become a convenient loophole for schools to avoid using all 11 official languages. The tension between language plan on the macro-level and its execution on the micro-level, in this case the school, is evident. The issue of language use in educational settings is further complicated by the power that is associated with English. Many South African parents seem to have the perception that access to English is what their children need in order to succeed in our society and internationally (Granville et al., 1997; Kamwangamalu, 1998; Mawasha, 1995). In many ways they are right, because English has certain material power as it provides entry to middle class jobs and salaries. As a consequence, students often have a negative attitude toward studying their own languages and regard people who speak English as educated (Granville et al., 1997). This has resulted in English being perceived as the ‘High language’ and the other official languages as the ‘Low languages’
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Codeswitching Among Multilingual Learners
227
(Kamwangamalu, 1998: 279–80). So, it is vital that an appreciation of the importance of students’ own languages for education be developed and that these students’ multilingual competencies be acknowledged. In order to address the language-related needs of South Africa, a Language Plan Task Group (LANGTAG) was established in 1995. This was based on the Minister of Arts, Culture, Science and Technology’s observation that ‘a definite tendency towards unilingualism’ was emerging in the country despite the fact that multilingualism is ‘a sociolinguistic reality’ in South Africa and that ‘the Constitution provides for the principle of multilingualism’ (LANGTAG, 1996: i). The LANGTAG final report proposed that the status and power of English, which is regularly referred to as ‘the hegemony of English’, be challenged by developing and expanding the use of African languages. Language as a means of teaching and learning in South African schools is currently a much debated topic. Under this topic, there is heated debate about whether or not codeswitching should be permitted in schools and whether or not it has any educational benefits. A question frequently posed is what happens to learners who are from multilingual backgrounds and find themselves in an English-medium school? Do these learners codeswitch and what are their reasons for doing so? In which situations do the learners codeswitch? Which languages out of the 11 official languages in South Africa are commonly used? Which language(s) do the participants switch to when one interlocutor has a different home language from the other? This paper attempts to address some of these questions by examining the use of codeswitching by multilingual learners in two English-medium schools in Johannesburg, South Africa. The different contexts within which codeswitching occurs and the reasons for such switching are also explored.
Codeswitching, Codemixing and Borrowing The significance of codeswitching, codemixing and borrowing in sociolinguistics, was first noted in Uriel Weinreich’s classic work Languages in Contact (Bokamba, 1988). For the sake of clarity and understanding, the terms codeswitching, codemixing and borrowing, which are frequently used in the literature, will be briefly discussed. The term codeswitching is often used differently by different researchers – some hold that it refers only to intersentential mixing (Kieswetter, 1995), while others use it as a broad term referring to both inter- and intrasentential mixing (Myers-Scotton, 1993a). In this study, the latter definition of codeswitching is preferred. Codemixing involves the mixing of affixes, words, phrases and clauses from more than one language within the same sentence and speech situation. Grammatical rules from all the languages involved are integrated into the discourse. Different types of switches, such as tag, intersentential and intrasentential are involved in the same discourse. For this reason, Poplack (1980) argues that codemixing involves great syntactical risks. Such a mixture of languages signals two or more identities of the speaker simultaneously. Borrowing is defined as ‘… the introduction of single words or short, frozen, idiomatic phrases from one variety into the other. The items in question are incorporated into the grammatical system of the borrowing language. They are treated as part of its lexicon, take
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
228
Bilingual Education and Bilingualism
its morphological characteristics and enter into its syntactic structure’ (Bokamba, 1988: 25). Eastman (1992) asserts that codeswitching and borrowing should not be seen as distinct processes, but as part of a single continuum since they both function similarly morphosyntactically. Myers-Scotton (1988, 1992, 1993b, 1993c), one of the authorities on codeswitching, defines it as the use of two or more languages in the same conversation. These languages are termed either the matrix language or the embedded language. She uses the Matrix Language Frame Model and the Markedness Model to explain the motivation and the functions of codeswitching. According to the Matrix Language Frame Model (MLF), the matrix language provides the morphosyntactic frame for codeswitched utterances, and constitutes the majority of morphemes in the given conversation. The choice of the matrix language is highly influenced by psycholinguistic and sociolinguistic factors. Thus, the matrix language is the dominant language used when codeswitching. There could be one or more embedded languages contributing to the codeswitched utterances. The embedded language is the secondary language used in a codeswitching context (Myers-Scotton, 1988). The MLF Model identifies the semantic and syntactical constraints, determining where the speaker may codeswitch within a sentence. According to the Markedness Model, making a choice to codeswitch at times carries extra-social meaning since the choice of languages used not only conveys the semantic content of the words, but also certain other messages. The motivation to employ a certain code is socio-psychologically driven. A code can either be marked or unmarked. The unmarked code for codeswitching is the normal, expected one for a particular situation. It is neutral and carries no extra-social meaning. A marked code, however, usually carries extra-social meaning in that the speaker’s codeswitching also conveys a meta-message beyond the semantic content of the words (Kieswetter, 1995). Code choice is also seen as being governed by the speakers’ relationship and their goals regarding their social position. All linguistic choices are seen as negotiating some rights and obligations (RO) balances which are based on the norms of the community of the speakers. The RO balances are based on what is expected or unmarked for speakers engaged in a particular conversation. This model is largely speaker-oriented as the speaker tries to negotiate his/her position in a conversational context. A contrasting model is the speech accommodation model (Giles et al., 1987) which is a hearer-oriented model where the speaker alters his/her speech to accommodate the hearer’s position in the conversational context. Kamwangamalu (1998) asserts that codeswitching is a dynamic phenomenon which cannot be explained only in terms of social negotiations of rights and obligations or in terms of power relationships, but should be examined in terms of the social context in which it is used. He suggests that in addition to a ‘we-code’ (the language used with one’s in-group members) and a ‘they-code’ (the language associated with more formal, out-group relations) proposed by Gumperz (1974, 1982, cited in Kamwangamalu, 1998), there should be a ‘code-in-between’, which is used as a neutral strategy which enables the speaker to achieve goals, which may or may not be political. Whether a language may be defined as a ‘we-code’, ‘they-code’ or ‘code-in-between’, depends on the context and social goals one wishes to achieve in a given speech situation. Kieswetter
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Codeswitching Among Multilingual Learners
229
(1995) provides a comprehensive list of social variables that can be negotiated by codeswitching. The list includes variables such as: identity, interpersonal relationships, social positions, group solidarity, ethnic identity, exploring new relationships, status,level of education, authority, neutrality, distancing or intimacy. Taking into consideration these social variables, this research aimed to explore the effects of an English language setting on the occurrence of codeswitching between multilingual learners in a primary school. It attempts to provide ‘classroom-based research’ of the type suggested by Faleni (1993) as well as research into codeswitching within African languages as suggested by Ramsay-Brijball (1999).
Literature Review A review of South African literature on codeswitching reveals studies that explain the functions of codeswitching in a wider social context (Calteaux, 1994; Finlayson & Slabbert, 1997; Herbert & Magwaza, 1994; Kamwangamalu, 1994, 1998; MacKormick in Jeppie & Soudien, 1990; Mesthrie, 1995; Makoni, 1999). There are very few local studies on the role of codeswitching in education (Nontolwane, 1992; Adendorff, 1993; Kieswetter, 1995; Du Plessis, 1995), which was a main focus of this research. None of the existing studies have focused on younger learners in primary schools in South Africa. This review is selective rather than exhaustive and it focuses on both national and international research in the area of codeswitching that is relevant to primary school and teacher education. Adendorff (1993) explored codeswitching among isiZulu-speaking teachers and their learners. The focus of his study was on the functions of codeswitching and the implications for teacher education. He concluded that there is a need for ‘consciousness raising’ among preservice teachers if they are to understand the role of codeswitching as an interactional resource. Du Plessis’ study (1995) was concerned with codeswitching in greetings and partings in school children. It was found that there is very little difference between instances of codeswitching in an informal or a formal situation, in terms of greetings and partings. Canagarajah (1996) examined the use of codeswitching by secondary school teachers in English second-language classrooms in Jaffana, Sri Lanka. Although the use of Tamil is considered inappropriate for English second-language classrooms, this study reveals some useful functions of codeswitching for both learners and teachers in terms of classroom management, transmission of subject matter, and the negotiation of values, identities and roles. ‘Decreasing mother tongue use in the classroom does not automatically increase the quality and quantity of target language use, any more than decreasing one’s consumption of meat automatically increases one’s consumption of cheese’ (Eldridge, 1996: 331). Eldridge’s opinion is that codeswitching seems to be a natural and purposeful phenomenon which facilitates both communication and learning. It does this in three ways. Firstly, codeswitching is a strategy that has benefits for second-language learners, as it provides a natural short-cut to content and knowledge acquisition (Butzkamm, 1998). Secondly, attempts to reduce codeswitching would hinder the acquisition of the second language. Lastly, it is
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
230
Bilingual Education and Bilingualism
proposed that there is a strong relationship between learning style and codeswitching. Sociolinguists often describe the individual’s use of language in codeswitching by classifying the utterances according to their different functions (Ellul, 1978). Nontolwane (1992) investigated the incidence and reasons for codeswitching and codemixing between speakers of isiZulu first language and English second language. The main determinants of codeswitching and codemixing were found to be the topic of conversation, status of the speakers, the speakers’ individual roles, the general context and language choice. Kieswetter’s (1995) study explored codeswitching among African high school learners. Here, codeswitching was identified, described and analysed in terms of social motivation and linguistic structural patterns. These patterns were compared and contrasted. It was found that speech patterns are influenced by the participants’ backgrounds and identity, relationships with each other and the context within which they interact. It is evident that there is a paucity of research about codeswitching among primary school learners in South Africa. Although the studies have been conducted in various contexts, with distinct perspectives and foci there is agreement among researchers that codeswitching indicates skilled performance and language proficiency, and that it is a valuable communicative tool in a multilingual society.
Methodology Subjects Forty multilingual subjects were randomly drawn from two primary schools from the Gauteng Department of Education, Johannesburg, South Africa. The language of learning and teaching in these schools was English while isiZulu and Afrikaans were taught as school subjects. Approximately half of the learners at the two schools spoke English as a second language. Speakers of almost all 11 official languages were represented in the two schools. The ages of the subjects ranged from six to ten years. Initially, children at preprimary schools were also targeted, but as they hardly ever engaged in codeswitching, they were excluded from the study. Procedure Data was collected by tape-recording the children’s informal and formal conversations during lunch breaks and lessons. Although learners were made aware of the project, they were not prepared in any way, nor were they told what was actually expected of them or what the focus of the research was. A hidden tape-recorder was used to tape-record their conversations. This was to avoid conscious codeswitching or, on the other hand, to prevent the avoidance of spontaneous codeswitching. Afterwards, the children and their parents were debriefed and the purpose of the research was explained to them. No-one objected to the data being analysed further, provided that the anonymity of the participants was guaranteed. All the information gathered was later transcribed, translated into English and analysed.
Codeswitching Among Multilingual Learners
231
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Data Analysis A qualitative, ethnographic approach was adopted for the analysis of the transcribed conversations, which was developed from work in sociology, anthropology, sociolinguistics and non-verbal communication (Jacob, 1987). This qualitative tradition assumes that both verbal and non-verbal communication is culturally patterned even though the persons communicating may not be aware of this patterning. The main focus of this tradition is on the patterns of social interaction among members of different cultural groups. Ethnographers of communication base their studies on participant observation data and on audio or video recordings of naturally occurring interactions. The machine-recorded data is then indexed and the segments selected for detailed analysis. The analysis of the transcribed conversations collected from this study was partly based on Myers-Scotton’s (1983) Markedness Theory. This theory emphasises the interactive and negotiated nature of face-to-face interaction. The two criteria, language dominance and context, used by Ellul (1978) to examine codeswitching between Maltese and English, were also applied here. Dominant language according to Ellul (1978) refers to the language that dominates the conversation (English or an African language). Context refers to the setting in which the codeswitching occurs (Ellul, 1978). This context could consciously or unconsciously influence the subject’s codeswitching. For this reason, the subjects’ codeswitching was examined in two different contexts, one formal (classroom) and the other informal (playground). An account of the reasons for codeswitching were based on Kieswetter’s (1995) social factors that typically emerge during codeswitching. A tally was made of the frequency of each type of codeswitching and this is reported in the Results section. Results
Codeswitching in a formal situation Context (formal classroom or informal playground) had a strong influence on the amount and type of codeswitching that took place. English is the language of learning and teaching in the classroom at the majority of urban South African schools. The various educators spoken to acknowledged this as being school policy. In order to avoid the use of African languages in the classroom, most of the classes are arranged in such a way that multi- and bilingual learners are seated well apart in classrooms which are composed of a majority of monolingual learners. Some of the educators mentioned that one of the major attractions for the parents is the language of learning and teaching at a particular school. This is related to the perceived power of English, as discussed earlier (Granville et al., 1997; Kamwangamalu, 1998; Mawasha, 1995). Time was spent in the classrooms to determine whether pupils codeswitch at all and, if they do, what social functions their codeswitching held. In the preprimary classrooms, the conversations of many of the preschoolers were characterised by long pauses of approximately 5 to 10 seconds. The conclusion drawn here was that these pupils still have limited vocabulary and are searching for the right word to express themselves, because these pauses gradually disappeared in Grade 1 and Grade 2. Romaine (1995) found that a certain level of linguistic competence must be reached before children are able to codeswitch.
232
Bilingual Education and Bilingualism
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Reasons for codeswitching in a formal context When codeswitching did occur in the classroom situation, it appeared to play a specific role, such as defiance or solidarity. Defiance Defiance occurs when a pupil uses an impermissible language with the aim of defying the regulations. Learners at the primary schools where the research was undertaken, know that they are not allowed to use an African language at school. In one of the primary schools there was a defiant boy who kept disrupting the lesson and as a form of punishment the teacher asked him to hand out worksheets. For English-speaking learners, he asked in English whether they have the worksheets, and addressed multi- and bilingual learners in different African languages. The teacher quickly reprimanded him because what he was doing was in defiance of the ‘English-only in the class’ school policy. Despite the perceived power of English, African languages can underscore this and reveal their own power when used in defiance of ‘English-as-the-only-language’ rule, especially when the teacher is an English monolingual. This type of codeswitching occurred relatively infrequently in the recordings made during this study. Multi-functional codeswitching: Phatic and solidarity Codeswitching can occur when a speaker tries to change the tone of the conversation. In this case codeswitching has a phatic function, which can be positive or negative. It is positive if it serves to narrow social distance, or if it is indicative of a relationship of solidarity. It is negative if it serves to increase social distance. Example A: Can I use your koki pens? B: No, they dry quickly. A: Oh! Please ngizowavala mangiqeda ukuwa-user (isiZulu). [I will close them after using them.] B: No, I don’t want you to use them. A: Ngiyakucela, toe (isiZulu).[Oh, please, I beg you.] B: No. A: Kulungile (isiZulu) [It’s fine], I am not going to let you use my wax crayons. B: Who cares. The above conversation started in English, which is the permissible language in school. When B responded negatively, A decided to repeat her request in isiZulu on the basis of drawing on the principle solidarity, since both A and B belong to the same ethnic group. A hoped this would make B reconsider her request. B rejected the bid for solidarity by not allowing A to use her koki pens and refusing to switch to isiZulu. At the end of the conversation A switched back to English. The multilingual learners who were part of this study employed this type of codeswitching relatively frequently. Grosjean (1982) gives a similar example where a Spanish–English bilingual student was speaking to a Puerto Rican student in English about giving her a lift to New York. After a while she switched to Spanish to ask how much it will cost. The student driver told her that it will cost 15 dollars. Later, the first student said
Codeswitching Among Multilingual Learners
233
in English that the trip was too expensive. Grosjean interpreted the codeswitching of the Puerto Rican student to Spanish as an attempt to invoke a sense of solidarity, and hopefully get a cheaper deal. Reasons for codeswitching in informal situations
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Because codeswitching occurred much more frequently on the playground, a number of functional taxonomies of codeswitching were investigated and are discussed below.
Unmarked codeswitching It appears that where English is the dominant language setting and where there are multilingual learners, codeswitching is the norm. Grosjean (1982) states that often codeswitching takes place quite unconsciously. The speakers’ main concern is with communicating a message and knowing that the other person will understand them, regardless of whether they speak one, two or more languages. Codeswitching as directive Codeswitching may have a directive function, where the codeswitching specifies the addressee as the recipient of the message. This switch can be defined as either inclusive or exclusive. Inclusive codeswitching The speaker switches to the language that a monolingual joining in knows and understands, and the addressee may even be invited to participate in the conversation. Example A: Uyazi ama-prefects ayakhetha (isiZulu). [You know prefects discriminate.] B: I know, thina abanandaba nathi (isiZulu). [I know, they do not care about us.] They like ama-Grade 1 prep nama-Grade 1 (isiZulu). [They like Grade 1 prep and Grade 1.] A: Thina bayasi-shout-a all the time (isiZulu). [They shout at us all the time.] C: What did you say? A: We are talking about the prefects. They are always in favour of Grade 1 prep and Grade 1’s. B: It’s not fair because they do allow Grade 1’s to play on our playground. Asambe manje (isiZulu). [Let’s go now.] A: Kulungile (isiZulu). [It’s fine.] The two bilingual learners switch between isiZulu and English although English is dominant. When they are joined by the monolingual learner, in order to accommodate her, they switch to English. They are inviting her to comment on the issue. Then to indicate that C is not invited to go with them, A and B switch back to isiZulu. The frequency of occurrence of this type of codeswitching in the sample was moderate.
Exclusive codeswitching This switching is often conscious and contains negative comments about those excluded. It indicates that the speakers share the language and the social distance between them is narrowed at the expense of the others who do not speak
Bilingual Education and Bilingualism
234
the same language. Myers-Scotton (1983) cites a case where parents switch to a different code in order to conceal information from their children. Example
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
A is the English-speaking prefect. B and C are isiZulu-Sotho-English multilingual learners. A: B: C: A: B:
If you do not go out I will lock you in. Uthi uzokhiya (isiZulu). [She says she is going to lock.] Wantina (seSotho). [She irritates me.] What are you saying? O nahanang? O nahana hore o-boss wa rona. Nna ke tla o hijacka ke ye le wena locationing (seSotho). [What do you think? You think that you are our boss. I am going to hijack you and take you to the township.]
This type of exclusive switching was very common at both of the primary schools, where bilingual or multilingual pupils were concealing information from monolingual pupils. In the above example, the two multilingual boys are using seSotho, isiZulu and English in their conversation. In the second line, B switched to isiZulu and he is not saying anything negative about the prefect. However, later on in the conversation, the switch is to seSotho. When the boys were asked why they firstly switch to isiZulu and later to seSotho, the reply was that they could not say horrible things in isiZulu in case the prefect understands it as they are taught isiZulu at the school. Many pupils using exclusive switching gave the same reason. Grosjean (1982) mentions that switching from one language to the other with the aim of withholding information from the third person, who does not know the language of switching and is supposedly part of the conversation, can backfire and lead to embarrassment, as in the example below. Example Two bilingual girls (A and B) sitting together during lunch break. C, a monolingual girl, walks past them. A: B: C:
Bheka lona uhamba kanjani uhamba njengogo (isiZulu). [Look at how this one walks. She walks like a grandmother.] No, njengomkhulu (isiZulu). [No, like a grandfather.] No, wena hamba njengalo old dog (Fanakalo). [No, you walk like an old dog.]
A and B assume that C, being white, cannot understand them as they talk about her in isiZulu. To their embarrassment, C does understand them and issues a sharp retort in a mix of isiZulu, Fanakalo (a pidgin language) and English. (Fanakalo developed in the mining industry and is a combination of English, Afrikaans and the dominant African language groups spoken in the area). In the primary schools where the research was undertaken, exclusive codeswitching usually occurred in order to exclude white learners from the conversations of black learners, as in the previous example. Exclusive codeswitching is not likely to work with a black peer since most, if not all, black learners are exposed to almost all African languages, especially isiZulu, isiXhosa, seSotho, seTswana and sePedi as they are widely spoken in the Gauteng region.
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Codeswitching Among Multilingual Learners
235
Codeswitching as an Exploratory Choice In the case of codeswitching as exploratory choice, strangers explore each other’s language preference. Speakers do not always know what norms apply and therefore do not know what the unmarked rights and obligations for themselves and other participants would be (Myers-Scotton, 1992). Speakers may initially find themselves in an uncertain situation. In this case the speaker will start the conversation in one code choice but will switch to another choice depending on the addressee’s response code. Example A: Heyi! Wena, awazi yini ukuthi i-tuck shop nge-first break ayivulwa (isiZulu). [Hey! You! Don’t you know that the tuckshop is not opened at first break?] B: What are you talking about? I do not understand. A: I am so sorry. Are you from Zaire? B: No, from Nigeria. I can only speak English. A: Okay, I was just saying the tuckshop only opens at second break. In the example above, the speaker initiating the conversation starts the conversation in isiZulu, not knowing that the other interlocutor cannot understand isiZulu. A took it for granted that B, being a black child, would be able to understand isiZulu. IsiZulu is used as the exploratory language. Once A had gathered that B cannot speak any African languages, A switches to English, which becomes the dominant language in the conversation. This type of codeswitching occurred infrequently.
Codeswitching as quotations Codeswitching often occurred when directly quoting what someone has said in a language other than the one used by the speaker(s) at that time. Switching back to the language that was originally used in the conversation indicates or emphasises group identity. Example A: O utlwile se Mrs P a se buileng? (seTswana). [Did you hear what Mrs P said?.] B: Mme e se for the first time a re, she does not want to see papers (seTswana). [It wasn’t for the first time she says that she does not want to see papers.] C: O nketsa gore ke tshege fa a bua, a re, ‘do you understand me, do you understand me.’ O se bua gantsi (seTswana). [She makes me laugh when she says ‘do you understand me, do you understand me?’ She says that many times.] After every quotation the speakers switched back to the group language. This could indicate that while they use a different language for quotations they still belong to the group and still share the same language. Codeswitching in order to quote what someone else had said occurred quite frequently, especially in conversations about English television programmes.
Codeswitching as an indicator of ethnic identity and solidarity This form of switching is used in an intragroup way to indicate group membership and identity (Gxilishe, 1992; Grosjean, 1982; Blom & Gumperz, 1972). It is what Gumperz (1974, 1982, cited in Kamwangamalu, 1998) referred to
Bilingual Education and Bilingualism
236
as a ‘we-code’. Speakers codeswitch between particular languages because they want to be identified with the group that speaks that particular language. Identity construction may have two phases. The first phase involves drawing a boundary between the participants and the rest of the society. This is an anti-social, uncooperative phase (as in exclusionary codeswitching).The second phase involves the establishment of mutual role relationships within the communicative event. This is the social, cooperative phase, where the code is chosen such that it suits every member of the society (Blommaert, 1992).
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Example A: B: C: B: A: D: A: E: D:
You won? Yes, I came one. I’m sure o itumetse (seTswana). [I’m sure you are happy.] Ja, maar (Afrikaans) le wena o tswile number one (seTswana). [Yes, but you also came number one.] Nna ke tswile number four. Every time o a mpheta (seTswana). [I came number four. Every time you are ahead of me.] E came last. (Laughing) She came number hundred and four. Ga ke na sepe (seTswana). [I don’t care.] Jo! Next time uzophuma number hundred and fifty (isiZulu). [Oh! Next time you will come number hundred and fifty.]
In the above example, when commenting about E, who is isiZulu-speaking and part of the group, D and A use English, which is the language common to all of them, because they want to make sure that E, who is not seTswana-speaking, as they are, understands. The extra-social message which the two speakers are sending, could be that E is not performing well because she does not belong to their language or ethnic group. E negotiates ethnic identity by responding in seTswana. D reminds E of her real ethnic identity by switching between isiZulu and English. Language can also be used as a symbol for group membership and solidarity. One tends to feel more positive about and closer to one’s group members than to those of another group. When strangers, meeting for the first time, realise that they have the same language background they usually switch to their mother tongue. In this case they are sending the message that, although they are strangers, they share a similar background (Grosjean, 1982). Example A: B: A: C: A: C:
Match o wa Sunday o ne o bora (seSotho). [Sunday’s match was boring.] Why usho njalo? (isiZulu). [Why do you say that?] Taba ke hore my favourite player Doctor Khumalo ne a sa dlale (seSotho). [It’s because my favourite player, Doctor Khumalo was not playing.] Maar (Afrikaans) nna ke utlwile gore why ne a sa dlale (seSotho). [But I heard why he was not playing.] O maaka. Why? (seSotho). [You are lying. Why?.] They say, wa gopola ka match wa last week Sunday o ne a lemala (sePedi). But ga a tlo tlogela all together. I think next game o tla e bapala (seSotho). [They say do you remember during last week Sunday’s match he was
Codeswitching Among Multilingual Learners
1:
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
A:
237
injured. But he is not stopping all together. I think next game he is going to play.] For leyo nto (isiZulu), fela nna (seSotho) I don’t believe you. I’m going to ask ntate wa ka (sePedi). He reads the paper every day. [For that thing only I don’t believe you. I’m going to ask my dad. He reads the paper every day.] Ntho eo e ka etsahala. Batho bana ba strict very bad (seSotho). [That can happen. These people are very strict.]
The languages used in the above example are sePedi, seSotho, English and isiZulu. The reason for this is that English is the medium of instruction for the school and seSotho is the home language of speakers A and C, who are dominating the conversation. C also uses a few sePedi words. B’s home language is isiZulu. In his first response he switches between English and isiZulu. In all of his other responses there is tag codeswitching, where his sentences are mainly in English with the tag end being in seSotho or isiZulu. B also uses ‘nna’ and ‘I’ tautologically. From this conversation one can surmise that B’s Sotho is not good and he is also aware of this. That is why he opts for tag switching rather than using long seSotho phrases or sentences. He tries to use some seSotho to engender a sense of belonging and group solidarity, which seemed to be a common reason for codeswitching in the sample studied.
Codeswitching as reiterative At times codeswitching can be used to reinforce, emphasise, amplify or even clarify the message that has already been transmitted in one code but which may not have been understood (Gumperz, 1982). Example A: No, no uphumile. You are out. Hhe-e, o tswile (seSotho). [No, no you are out. You are out. No, you are out.] B: No, I’m not. A: You are out, o a bona he o qadile hape (seSotho). That is why I don’t want to play with you. Heyi wena o a bora. That is why ke sa batle go tshameka le wena (seTswana). [You are out, you see, you have started again. That is why I don’t want to play with you. Hey you, you bore me. That is why I do not want to play with you.] In the above example, the speakers switch from one language to the other and frequently phrases are said in one code and repeated in another code, to emphasise and reinforce them. In most of the conversations where such codeswitching occurs, it has a phatic influence since it changes the tone of the conversation. This type of codeswitching occurred relatively infrequently.
Codeswitching as a strategy of neutrality Bilinguals often avoid committing themselves to a single rights and obligations set by avoiding speaking only one code. The speaker recognizes that the use of each of two languages has its value in terms of the costs and rewards which accrue with its use. The speaker decides to choose a middle path regarding these costs and rewards by using two (or more) languages in a single conversation. (Myers-Scotton, 1993a: 147).
Bilingual Education and Bilingualism
238
In such cases, codeswitching can enable the speaker to maintain a position of neutrality. Kamwangamalu (1998) refers to these instances of codeswitching as the ‘code-in-between’. Example A: B:
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
C: D: A: C:
O se ke wa matha (seSotho), ibhola alibalekanga (isiZulu). O a-robber (seSotho). [Do not run when the ball is not away. You are cheating.] No, no o ka etsa jwalo ha o batla (seSotho). [No, no you can do that if you want to.] It’s not fair thina asenzanga kanjalo (isiZulu) lona la-robber (seSotho). [It’s not fair we did not do that, you are cheating.] O ka etsa jwalo as long as o tla khona ho baleha (seSotho). [You can do that as long as you will be able to run.] Uyakhumbula (isiZulu) yesterday lona le hanne (seSotho). [Do you remember yesterday you refused to do that.] Lona everyday le tla (seTswana) le melao ya lona e seleng (seSotho). [Every day you bring your silly rules.]
In this conversation, it was difficult to determine who spoke what language because the participants swiftly switched from one language to the other (seTswana, seSotho, isiZulu or English). In some cases, the switch involved up to three languages in the same sentence. Codeswitching was often used as sentence fillers. On a practical level, this often occurs because of the inability of the speakers to find words to express what they want to say in one code. However, in the example given above, the switching occurs with relatively high frequency words and so this is unlikely to be the case. Rather, speakers A, B, C and D switch between the four languages, whether on a conscious or unconscious level, because they do not want any of the languages to dominate the conversation. Each speaker wants to remain neutral and show no signs of favouring any of the participants by using their home language. This type of codeswitching was not employed often by the sample.
Codeswitching in order to hide one’s identity Language users have certain perceptions about their own language and other languages (Grosjean, 1982). A negative perception of a language can have profound effects on the users of the language, especially if it is not their first language. The learning and use of the negatively perceived language can also be affected. On several occasions learners codeswitched because they were trying to hide their home languages. For example, on one occasion, one Grade 2 boy was standing alone at break. The first author spoke to him in English and asked him why he was not playing soccer with the other boys. The boy responded in English but from his accent it was evident that his home language was xiTsonga. The first author then switched to xiTsonga and continued the conversation. The boy responded in xiTsonga. However, when the boys who were playing soccer ran towards them, the boy immediately switched to seTswana. When the boys ran away from them, he switched to xiTsonga again. He seemed to be codeswitching because he was trying to hide his identity, a reason which did not occur often in the sample.
Codeswitching Among Multilingual Learners
239
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Conclusion Learners use their linguistic abilities to manipulate their conversations according to content and context. The findings from this study indicate that the use of codeswitching by multilingual learners in multiracial primary schools has very specific aims, which may be either conscious or unconscious, on the part of the participants and that its occurrence is very frequent. This supports Makoni’s (1999) notions about the interconnectedness of languages. Makoni (1999) cautions against using a monolingual orientation to understand a multilingual phenomenon – in South Africa’s multilingual situation, codeswitching is the norm and pure Zulu, for example is rare. According to Makoni (1999: 144), mixed codes result from the interconnectedness of languages and attempts to isolate and separate the languages and to view codeswitching as the exception results from the myth that languages are ‘hermetically sealed units’. Since it would be very difficult and possibly detrimental to learners to prevent the use of codeswitching at school, its value for teaching and learning needs to be explored. If well-structured and organised, codeswitching can be used effectively in the classroom (Butzkamm, 1998). It is a linguistic tool which multilingual speakers have available to them without it having been explicitly taught to them, and educators need to be aware of the widespread nature of the phenomenon (Kamwangamalu, 1998, Kieswetter, 1995). The findings from this study can be of use to teachers, teacher educators, policy makers and educational planners in assessing the complex issue of language choice in South African classrooms. The use of codeswitching as a teaching strategy can be effective for both language and content acquisition, since it possesses several communicative functions in the classroom, namely translation, clarification, checking comprehension, giving instructions and procedures as well as acting as a ‘we-code’ (providing a sense of cohesion) (Gumperz in Kamwangamalu, 1998; Guthrie in Martin, 1996). Atkinsons, 1987 in Martin, 1996; Lin, 1988 in Martin, 1996 and Garret et al., 1984 in Martin, 1996, all describe the benefits of using codeswitching in the classroom. These include saving time, maintaining discipline and helping weak learners. It is the authors’ position that there is a place for codeswitching in education, particularly in the multilinguistic setting of South African schools. It should be included in the planning of syllabi, textbooks and any other teaching material for all schools which cater for bilingual and multilingual learners. The positive relationship between multilingualism and educational success is well documented (Bangbose, 1984; Macdonald, 1987; Sure, 1997). While the debate about whether or not to allow codeswitching in South African schools continues, this research may help teachers and teacher educators develop an understanding of what codeswitching entails, the different messages it may convey and its positive implications for teaching and learning. Multilingualism is a meaningful factor in the educational development and transformation of South Africa (Webb, 1998). If, as educators in South Africa, we have the motivation to implement the Language in Education Policy (1997) to the benefit of every learner, and if we wish to make the constitutional provision of language affirmation a reality, then codeswitching should be recognised as crucial to the speech of multilinguals (Butzkamm, 1998; Makoni, 1999) and as a resource that facilitates effective learning (Eldridge, 1996).
240
Bilingual Education and Bilingualism
Correspondence Any correspondence should be directed to Mrs K. Cockcroft, Department of Psychology, University of the Witwatersrand, P.O. Wits, 2050, South Africa (
[email protected]).
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
References Adendorff, R. (1993) Codeswitching amongst Zulu-speaking teachers and their pupils: Its functions and implications for teacher education. Southern African Journal of Applied Language Studies 2 (1), 1–26. Bangbose, A. (1984) Mother tongue medium and scholastic attainment: The Nigerian experience. Prospects 14 (1), 87–93. Blom, J. and Gumperz, J.J. (1972) Social meaning in linguistic structure: Codeswitching in Norway. In J.J. Gumperz and D. Hymes (eds) Directions in Socio-linguistics. New York: Holt, Rinehard and Winston. Blommaert, J. (1992) Code switching and the exclusivity of social identities: Some data from campus Swahili. Journal of Multilingual and Multicultural Development 13 (1–2), 57–70. Bokamba, E. (1988) Code mixing, language variation and linguistic theory: Evidence from Bantu languages. Lingua 76, 21–62. Butzkamm, W. (1998) Code-switching in a bilingual history lesson: The mother tongue as a conversational lubricant. International Journal of Bilingual Education and Bilingualism 1 (2), 81–99. Calteaux, K.A. (1994) A sociolinguistic analysis of a multilingual community. Unpublished thesis, Rand Afrikaans University, Johannesburg. Canagarajah, A.S. (1996) Functions of code switching in ESL classrooms: Socialising bilingualism in Jaffana. Journal of Multilingual and Multicultural Development 13 (1–2), 173–195. The Constitution of the Republic of South Africa (1996) Pretoria: Government Printer. Department of Education (1997) Language in Education Policy. Government Gazette 17997 (16312). Pretoria: Government Printer. Desai, Z. (1994) Praat or speak but don’t thetha: On language rights in South African. Language and Education 8 (12), 19–29. Du Plessis, J.M. (1995) Greetings and parting routines in Nguni: The effects of social distance and sex. Unpublished Masters resea rch report, Univers ity of the Witwatersrand, Johannesburg. Eastman, C.M. (1992) Code switching as an urban language contact phenomenon. Journal of Multilingual and Multicultural Development 13 (1–2), 1–17. Eldridge, J. (1996) Code switching in a Turkish secondary school. ELT Journal 50 (4), 303–311. Ellul, S. (1978)A Case Study in Bilingualism: Codeswitching between Parents and their Preschool Children in Malta. Great Britain: Campfield Press. Faleni, T. (1993) Codeswitching in multilingual classrooms: A teacher’s debate. ELTIC Reporter 17 (1–3), 11–12. Finlayson, R. and Slabbert, S. (1997) ‘We just mix’: Codeswitching in a South African township. International Journal of the Sociology of Language 125, 65–98. Giles, H., Mulac, A., Bradac, J.J. and Johnson, P. (1987)Speech accommodation theory: The first decade and beyond. In M.L. McLaughlin (ed.) Communication Yearbook (10). California: Sage. Granville, S., Janks, H., Joseph, M., Mphahlele, M., Ramani, E., Reed, Y. and Watson, P. (1997) English with or without g(u)ilt: A position paper on language in educational policy for South Africa. Paper presented at the English Teachers connect International Conference, University of the Witwatersrand, Johannesburg. Grosjean, F. (1982) Life with Two Languages: An Introduction to Bilingualism. Cambridge: Harvard University Press. Gxilishe, D.S. (1992) Conversational code switching. South African Journal of African Languages 12 (3), 93–97.
Downloaded by [The Library, University of Witwatersrand] at 05:05 22 January 2016
Codeswitching Among Multilingual Learners
241
Herbert, R.K. (1992) Language and Society in Africa (1–9). Johannesburg: Witwatersrand University Press. Herbert, R.K. and Magwaza, J. (1994) Codeswitching among Black students at Wits. Unpublished paper presented at the African Languages Staff Seminar. Johannesburg, University of the Witwatersrand. Jacob,E. (1987)Qualitative research traditions: A review. Review of EducationalResearch57, 1–50. Jeppie, S. and Soudien, C. (eds) (1990) The Struggle for District Six: Past and Present. Cape Town: Buchu Books. Kamwangamalu, N.M. (1994) SiSwati-Englishcodeswitching: The matrix language principle and linguistic constraints. South African Journal of African Languages 14, 70–77. Kamwangamalu, N.M. (1998) ‘We-codes’, ‘they-codes’, and ‘codes-in-between’: Identities of English and codeswitching in post-apartheid South Africa. Multilingua 17 (2–3), 277–296. Kieswetter, A. (1995) Codeswitching amongst African High School pupils. Unpublished Masters dissertation, University of the Witwatersrand, Johannesburg. LANGTAG (1996) Towards a National Language Plan for South Africa. Final Report of the Language Plan Task Group, presented to the Minister of Arts, Culture, Science and Technology, 8 August 1996. Macdonald, C.A. (1987) Crossing the Threshold. Pretoria: HSRC. Makoni, S. (1999) Shifting discourses in language studies in South Africa. In Kwesi Kwaa Prah (ed.) Knowledge in Black and White. Cape Town: Centre for Advanced Studies of African Society. Martin, P.W. (1996) Code-Switching in the primary classroom: One response to the planned and unplanned language environment in Brunei. Journal of Multilingual and Multicultural Development 17, 128–144. Mawasha, A. (1995) The role of English in education in the new South Africa. Paper presented at the Eleven Official Languages in South Africa: The Implication of South African Language Policy on the Teaching of Languages Conference, Rainbow Communication, 23–25 April 1995. McLaughlin, M.L. (ed.) (1987) Communication Yearbook (10). California: Sage. Mesthrie, R. (ed.) (1995) Language and Social History: Studies in South African Sociolinguistics. Johannesburg: David Phillip. Myers-Scotton, C. (1983) The negotiation of identities in conversation: A theory of markedness and code choice. InternationalJournal of the Sociologyof Language 44, 115–136. Myers-Scotton, C. (1988) Codeswitching as indexical of social negotiations. In M. Heller (ed.) Codeswitching:An Anthropologicaland SociolinguisticPerspective. Berlin: Mouton de Gruyter. Myers-Scotton, C. (1992) Comparing code-switching and borrowing. Journal of Multilingual and Multicultural Development 13 (1–2), 19–39. Myers-Scotton, C. (1993a) Social Motivations for Code-Switching: Evidence from Africa. Oxford: Claredon Press. Myers-Scotton, C. (1993b) Duelling Languages: Grammatical Structure in Codeswitching. Oxford: Claredon Press. Myers-Scotton, C. (1993c) Common and uncommon ground: Social and structural factors in codeswitching. Language in Society 22, 475–503. Nontolwane, G.B.N. (1992) Code-switching and code-mixing in Zulu. Unpublished Masters dissertation, Rand Afrikaans University, Johannesburg. Poplack, S.(1980)Sometimes I’ll start a sentence in Spanish Y terino en Espanol: Towards a typology of codeswitching. Linguistics 18 (7–8), 581–618. Ramsay-Brijball, M. (1999) Understanding Zulu-English codeswitching: A psychosocial perspective. South African Journal of Linguistics 17 (2–3), 161–172. Romaine, S. (1995) Bilingualism. Cambridge: Blackwell Publishers. Sure, K. (1997) Language development and integration in Kenya since independence: The breaking borders. A paper presented at the African–Hispanic and Counters Conference, Durban. Webb, V. (1998) Multilingualism as a developmental resource: Framework for a research programme. Multilingua 17 (2–3), 125–154.