This chapter highlights the significance of this research, which is contextually located in WCED primary schools located in disadvantaged communities in SA. It reviews:
· The challenges faced in formulating LoLT policy in SA;
· SA LoLT policy as legislated as opposed to its practice;
· What options are available to meet this challenge as well as the debate around these alternatives; and
· The repercussions of LoLT policy in SA on factors affecting learner performance.
Successful education for the majority of learners in SA depends on their becoming fluent in a language that is not their first language (L1). This is because all SA learners must write their school-leaving examinations in either English or Afrikaans, although English is more frequently chosen (Taylor & Coetzee, 2013, p. 33). In 2007, 83,4% of learners wrote their school-leaving examinations in a second language (L2) (DBE, 2010, p. 13). During this year, 81,4% of SA learners wrote these examinations in English, compared to 12,8% that wrote them in Afrikaans (DBE, 2010, p. 16).
This raises questions about what might be the most advantageous language education policy options. For example:
· Is learners’ L1 to be used in school, or is schooling to be conducted in English or Afrikaans?
· If learners’ L1 is to be used, should it be used as a LoLT or taught as a subject?
· If learners’ L1 is to be used as a LoLT, is this to be throughout schooling, or should there be a transition from L1 to L2 as the LoLT at some point? If so, at what point and how should such a transition take place?
Various LoLT policy options embody possible answers to these questions.
LoLT policy options can be divided into two often opposing groups: weak and strong bilingual models.
Weak bilingual models include: subtractive bilingual models, where the aim is to move learners from L1 to L2 LoLT as soon as possible; as well as transitional models that aim to move learners from L1 to L2 LoLT in a transition (Heugh, 2006, p. 60). These models are referred to as ‘weak bilingual models’ as they have one language as a target for use as a LoLT, this language being a learner’s L2.
At the subtractive extreme of weak bilingual models is the submersion approach, where a L2 as a LoLT is used without any transition and without learners having had any prior exposure to it in school.
Should a transition from L1 to L2 LoLT be present and occur within the first 1 to 4 years of starting school, the model is an early-exit transitional model (Ouane & Glanz, 2011, p. 11). If transition occurs after 5 to 6 years of having started schooling, the model is a late-exit transitional model (Heugh, 2006, p. 114). Very-late-exit transitional models also exist. Here, the transition from L1 to L2 LoLT occurs after 8 to 9 years of having started schooling (Heugh, 2006, p. 120).
Strong bilingual models are additive bilingual models. The target of LoLT is either the L1 with L2 taught as a subject, or the use of both L1 and L2 as LoLTs in a dual-medium style (Heugh, 2006, p. 61).
Within strong bilingual models the L1 is not removed as a LoLT and the target is a strong learner proficiency in both L1 and a L2 (Heugh, 2006, p. 114).
The LoLT policy debate is one that draws on two conflicting pools of theoretical thought.
Proponents of a ‘stronger’ bilingual LoLT policy hold the view that a later transition to English LoLT is needed, as non-English L1 learners cannot understand English from day 1 of school and are not able to learn if it is used (Taylor & Coetzee, 2013, p. 3). This view is that L1 LoLT should be used until the learner has reached a level of academic proficiency in the L2. Consensus exists between L2 acquisition experts that this is the most pedagogically sound approach (Heugh, 2013: 220).
Those in favour of a ‘weaker’ bilingual LoLT policy are of the opinion that using L1 as LoLT during early schooling delays and distracts learners’ learning of the L2 (Taylor & Coetzee, 2013, p. 4). This thinking is that L2 learning must begin as soon as possible to maximise time spent learning the L2 (Imhoff, 1990), and language learning is easier at a young age (Pinnock & Vijayakumar, 2009, p. 12).
LoLT policy in SA is based on the Constitution of the Republic of South Africa (RSA, 1996a) and the South African Schools Act (SASA) (RSA, 1996b). From these, the then Department of Education promulgated the Language in Education Policy (LiEP) (DoE, 1997), expanded upon in the Revised National Curriculum Statement (DoE, 2002), and the Curriculum and Assessment Policy Statement (CAPS) (DBE, 2010). In 2013, the now Department of Basic Education published a draft and yet-to-be-finalised policy document, The Incremental Introduction of African Languages in South African Schools (DBE, 2013). This legislation currently sets the primary legal context for which LoLT policy is developed in SA.
The Constitution legislates the official and equal status of eleven languages in SA as well as the constitutional right of all South Africans to receive education in the official language(s) of their choice in public schools within reasonable practical limitations. Informed by this, SASA (RSA, 1996b) prescribes that school governing bodies (SGBs) determine the LoLT policies of schools subject to prevailing law. LiEP (DoE, 1997) adds voice to the Constitution and SASA’s views on LoLT policy. LiEP requires that schools formulate LoLT policy that:
· Promotes multilingualism (the ability to speak two or more languages) in the form of strong bilingual dual-medium LoLT implementation by “using more than one LoLT, and/or by offering additional languages as fully-fledged subjects” (DoE, 1997, p. 3); and
· Assists Provincial Education Departments to honour the constitutional right of learners to determine their LoLT.
LiEP (DoE, 1997) recommends schools adopt an “additive approach to bilingualism” (DoE, 1997, p. 1) as a way of achieving multilingualism where “the underlying principle is to maintain home language(s) while providing access to the effective acquisition of additional language(s)” (DoE, 1997, p. 1). As such, “an additive approach to bilingualism is to be seen as the normal orientation of our [DoE] language-in-education policy” as “most learners benefit cognitively and emotionally from the type of structured bilingual education found in dual-medium schooling” (DoE, 1997, p. 1).
In practice, SA primary schools are obliged to ensure that all learners learn:
· Their home language and at least one L2 which must include English, as prescribed by CAPS (DBE, 2011), as subjects from Grade 1; and
· An African language as a subject for a minimum of three years by the end of the General Education and Training band of schooling (Grades R up to and including Grade 9).
The schools must then determine all remaining aspects of LoLT policy via decisions made by SGBs (Taylor & Coetzee, 2013, p. 4).
In contrast to the additive approach to bilingualism as DBE’s normal orientation, how LoLT policy pans out for the majority of disadvantaged primary schools in SA is that a transitional approach is taken in the form of a late-exit transitional LoLT model where the L1 is used as the LoLT up to but not including Grade 4, while English is taught as a subject in this time. Then, from Grade 4 onwards, a transition occurs, and English is used as LoLT while the L1 is taught as a subject. Within a minority of disadvantaged SA schools a straight-for-English approach is taken, where English is LoLT from Grade 1, while the L1 is taught only as a subject. This is a type of submersion LoLT policy.
As such, and contrary to the prevailing language in education legislation, the models of LoLT policy prevalent in disadvantaged SA primary schools are subtractive, not additive, in their approach to bilingualism. DBE (2010) data illustrates this trend. In 2007, 70% Grade 4 African-language L1 learners in SA were taught in English (up from less than 30% in Grade 3) (Pluddeman, 2015, p. 192).
As such, a gap in implementation exists between the emphasis placed by the Constitution and LiEP on achieving multilingualism, and how LoLT policy is formulated and implemented by schools. According to Heugh (2013, p. 1), this is as a result of a disjuncture between constitutional and other government policies, which results in “an assimilatory drive towards English” as the LoLT in SA schools.
With regard to where responsibility for the implementation of LoLT policy in SA schools rests, this is only vaguely referred to within government policy documents as “the school”. The extent to which LoLT policy is implemented within SA schools is difficult to ascertain as doing so is reliant on reports from education role-players who often share some responsibility for the LoLT implementation — and are therefore unwilling to disclose information on the matter.
The high prevalence of teachers not using the LoLT when code switching and code mixing (switching between two languages when speaking) during lessons within L2 LoLT classrooms in Africa provides an indication of the limited extent to which LoLT policy is implemented in similar SA contexts (Benson, 2005, p. 2; Brock-Utne & Alidou, 2006, p. 116; Bush et al., 2009, p. 3; de Wet, 2002, p. 121).
LoLT policy implementation holds far-reaching effects on learners both inside and outside of the classroom. The research on these effects discussed below refers to LoLT policy contexts that are comparable to those found in disadvantaged SA schools.
The majority of SA learners must accept schooling that is conducted in a L2 that they cannot understand to some degree. Because of this tendency of SA LoLT policy to presuppose learner (and teacher) competence in the LoLT, learners often do not understand what their teachers are saying (Brock-Utne & Alidou, 2006, p. 102) and are not able to access the curriculum because of miscommunication of a systematic nature found to exist between teachers and learners (Erickson, 1987).
Similarly, most teachers in disadvantaged SA schools must teach in a language that is not their L1 and in which they are not fully proficient. As such, communication of the curriculum is jeopardised because neither teachers’ nor learners’ L2 skills are adequately developed (Probyn, 2008, p. 217). Learners’ and/or teachers’ inability to use the LoLT is therefore viewed as the central pedagogic problem in L2 LoLT schooling contexts as a result of the curriculum inaccessibility caused by inefficiency in ‘pedagogic transmission’, that is, learning (Smits, Huisman & Kruijff, 2008, p. 7).
Learners perform poorly in subjects taught in an unfamiliar LoLT, particularly in those that require decontextualised use of language such as Mathematics and Science (Heugh, 2006, p. 71). This is strikingly apparent in SA. During the most recent Trends in International Mathematics and Science Study (TIMSS) held in 2015, SA learner performance was amongst the very worst of the 57 countries that participated (The Economist, 2017).
A difference between the language that learners speak at home and LoLT negatively affects educational outcomes (Smits, Huisman & Kruijff, 2008, p. 6). Within disadvantaged L2 LoLT contexts, learner performance and school attendance are lower, while repetition and drop-out rates are higher compared to similar contexts where learners are taught in their L1 (Pinnock & Vijayakumar, 2009, p. 26).
Learners learn better when they are taught in their L1 because they can understand their teachers better (Bender et al., 2005, p. 2). L1 LoLT use is also conducive to the effective teaching of early literacy because the language barrier introduced when learners are required to decode a language that they do not understand before making connections between spoken and written language is avoided (Smits, Huisman, & Kruijff, 2008, p. 8).
Table 2.1, below, developed from international research conducted by Thomas & Collier (1997, p. 76) shows anticipated learner performance of Grade 12 learners in the subject English following different LoLT models. Score A illustrates the anticipated score of a learner who begins schooling in L2. Score B shows the predicted score of a learner who begins learning in a L2 at the start of Grade 4. Score C indicates the predicted effect on the Grade12 English score of beginning to learn in the L2 from the start of Grade 8. Score D indicates the predicted learner score for learners who follow a type of dual-medium LoLT model, while Score E shows the anticipated score for those learners that learnt in the L1 throughout school with L2 as a subject.
Table 2.1: Anticipated learner performance in English at Grade 12 (Thomas & Collier, 1997, p. 76)
Compared to learners taught in their L1, learners taught in a L2 suffer a pedagogic processing disadvantage (Cooper, 1999, p. 254) because learning new ideas and information is most efficient when what is being taught is located within learners’ existing frame of reference, accessed easily via, and inextricably linked to, the L1, in a learning process where learners’ prior knowledge is built upon to create more advanced knowledge during a process of assimilation (Bransford, Brown & Cocking, 1999). Learners’ L1 is the primary vehicle upon which the success of their learning and cognitive development depends, because teaching in learners’ L1 better supports learners' ability to learn than teaching in a L2 does (Traoré, 2001).
Longitudinal experimental research on the effect of extended L1 LoLT use on learner performance in disadvantaged contexts gives evidence to the above and has shown that learners who are taught in their L1 for the first six years of school perform significantly better throughout their schooling (Brock-Utne & Alidou, 2006, p. 106) compared to those learners who were taught in their L1 only for the first three years of their schooling. Comparable studies have found similar results (Alidou, 1997; Bergman et al., 2002; Chekaraou, 2004; UNESCO, 2003). Research also shows that the use of L1 LoLT is more conducive to the development of literacy in learners, generally, than a L2 LoLT is (UNESCO, 2016, p. 3).
Learners who are taught in their L1 during the early years of school not only learn more quickly but also develop stronger reading and writing skills than learners in comparable contexts who are taught in a L2 (Smits, Huisman, & Kruijff, 2008, p. 8). Learners are also able to learn and use a L2 for learning far more easily after they have fully learnt their L1, which takes about 12 years from birth to accomplish (Benson, 2005), compared to older learners or adults who have not fully developed their L1 (McLaughlin, 1992).
Understanding the effect of L1 on learning and L2 acquisition, research brings to light problematic implications for a learner who has had the mastery of their L1 curbed before it is fully developed:
· If a learner has not first learnt ideas and words in his or her L1, he/she will be less unable to later transfer his or her understanding of these ideas and words for use in a L2 (Daby, 2015, p. 7).
· A learner is likely to be unable to understand or use L2 words that relate to academic or abstract concepts if such concepts have not first been properly learnt in the learner’s L1 (Pinnock & Vijayakumar, 2009, p. 13).
· Any word that a learner learns to use but that does not understand is an exercise in memorisation, not the development of that learner’s literacy. The use of a L2 that a learner does not understand in learning to read and write is counterproductive (Pinnock & Vijayakumar, 2009, p. 13).
· If a learner’s L1 is not used at school, that learner will be unable use his or her L1 for academic and abstract purposes later because only through use can a language develop.
The relationship between LoLT and the development of literacy in learners[1] emphasises the link between language development and literacy. Such research indicates that without L1 development, progress in gaining L1 and L2 literacy in learners is undermined in that a sufficient level of L1 development is required to attain literacy in any language, regardless of whether a L1 or a L2 (Mackenzie & Walker, 2013, p. 6). As an effect of subtractive bilingualism, this results in learners achieving only what scholars describe as semilingualism, the incomplete linguistic development of both the L1 and the L2 (Hibbert, 1995, p. 7).
The extent to which a learner must be proficient in a L2 before being able to use it adequately as a LoLT is another prominent factor that should be taken into account when deciding upon the appropriate age for a learner to make the transition from learning in his or her L1 to learning in a L2, an unavoidable reality for most disadvantaged learners in SA.
While the ‘common sense’ notion is that the sooner a learner is exposed to a L2 at school, the better they will learn it, this is not the case and, to the contrary, is likely to have the opposite effect (Heugh, 2006, p. 64). It is not only unreasonable to expect that a learner learns in a L2 LoLT when they have not yet developed proficiency in using it, but expecting this from a learner is also likely to undermine the development of both the learner’s literacy in both the L1 and L2 (Mizza, 2014, p. 102).
An example of a LoLT model that requires learners to make the transition from L1 to L2 LoLT before they are adequately prepared to do so is the early-exit LoLT model, found in most disadvantaged SA primary schools. This model is criticised by those who favour for the following reasons a stronger bilingual approach to LoLT policy.
At the point of LoLT transition from L1 to L2 in Grade 4, learners experience a significant barrier to learning where the required cognitive leap for learners to perform adequately from ‘learning to read’ the L2 to ‘reading (the L2) to learn’ is too great for the majority (Pretorius, 2002; Pretorius & Ribbens, 2005). A learner cannot be expected to learn a L2 as well as perform adequately using it as a LoLT at the same time (Heugh, 2006, p. 64). Should this be expected of learners, they are likely not to learn the L2 adequately, as well as to fail to understand explanations of concepts and skills that teachers attempt to teach them using it, thereby paralysing their learning. This will then result in such a learner declining in his or her performance and falling behind the performance of learners of his or her age who are taught in their L1.
Research conducted throughout Africa, such as that by Halaoui (2003) in Niger and Sampa (2003) in Zambia, has examined the effect of early-exit LoLT models, where learners move from L1 to L2 LoLT in Grade 4, on learner performance. This research found that learner performance declined within one or two years after learners had begun learning in a L2, and further continued to decline in subsequent grades (Heugh, 2006, p. 71).
Illustrating further findings comparable to those found in SA, Figure 2.1 below, reproduced from work by Thomas & Collier (1997, p. 53), shows the learner performance in reading scores of 43 217 L2 learners who followed alternate LoLT models between 1982 and 1996 in the USA. Amongst these learners, those in early-exit models (represented by Line 4 on Figure 2.1) showed a steady decline in their performance from the point of LoLT transition at the SA equivalent of Grade 4 when they began learning in a L2 LoLT, while those learners who continued to learn in L1 LoLT for a longer period (represented by Lines 1-3 on Figure 2.1) performed better over time.
Figure 2.1: Learner performance per LoLT model (Thomas & Collier, 1997, p. 53)
A reason for the decline in learner performance from the point of L1-L2 LoLT transition in early-exit models is that learners are required to know the L2 LoLT impossibly well in order to access and meaningfully engage with the primary school curriculum. While a minimum active vocabulary of between approximately 5 000 and 7 000 words is required for understanding the curriculum in Grade 4 (Thomas & Collier, 1997), by the end of Grade 3 SA learners have been found to have been able to amass a L2 vocabulary of only about 500 words (Heugh, 2006, p. 65). Such learners are also unable to engage fully with, or express, their prior knowledge in the L2 during learning processes.
Learners in early-exit systems are often not allowed the time needed to learn the required vocabulary to access the curriculum in a L2. While most learners in disadvantaged countries must learn in a L2 after having had little or no exposure to it at school, a learner needs to have spent between six and eight years learning the L2 before they are able to understand curricular content in that language adequately (DoE, 2007, p. 2).
Not one example in Africa of an early-exit model has been found to have succeeded (Heugh, 2006, p. 66). For disadvantaged primary schools that follow this LoLT policy, the repercussions for education include poor performance (Heugh, 2006, p. 71), low literacy attainment, the underdevelopment of both the L1 and L2 (Heugh, 2006, p. 67), and escalations in repetition and attrition rates from the point where learning begins in the L2 (Heugh, 2006, p. 65), as well as the use of ineffective and inefficient teaching methods (Alidou & Brock-Utne, 2006, p. 85) that produce education of a low quality.
The implementation of L2 LoLT policy within contexts where very few people can speak the LoLT is problematic as the language causes a barrier to teaching and learning when teachers and learners cannot use it to communicate and learn (Alidou & Brock-Utne, 2006, p. 85).
Within such contexts where teachers and learners cannot adequately use the L2 LoLT, classroom observations show that teachers tend to attempt overcoming the language barrier by providing learners with lessons, activities and assessments which are characterised by their tendency to facilitate cognitive processes that are limited to a lower order, and result in a poor quality of education (Alidou & Brock-Utne, 2006, p. 87). Lower-order cognitive processes are educationally ineffective because they do not facilitate the higher-level cognition required for learners’ effective learning. This is because teachers can engage learners’ higher-order thinking only by using questions of a high quality, which is generally possible only when facilitated in the learners’ L1 as a result of teachers’ and learners’ proficiency in it (Alidou & Brock-Utne, 2006, p. 87).
Low-quality teaching practices commonly found to be employed by teachers in contexts where the LoLT is neither the teachers’ nor the learners’ L1 include those such as safe talk, repetition, memorisation, and chorus teaching, as well as code switching and mixing (Alidou & Brock-Utne, 2006, p. 85; Brock-Utne & Alidou, 2006, p. 102). In such contexts, learners tend to hold back from actively participating and are typically passive and quiet during lessons taught in the L2 (Alidou & Brock-Utne, 2006, p. 85).
In contrast to patterns of low learner participation, teacher-centred and educationally-ineffective teaching styles and the poor quality of education typically found in L2 LoLT classrooms, experimental research from comparable contexts shows that in L1 LoLT classrooms:
· Teachers employ teaching methods that are more engaging and more educationally effective (Bender et al., 2005, p. 2);
· Learner participation is higher (Alidou & Maman, 2003; Dzinyela, 2001);
· Greater active teacher-learner interaction is fostered, which further allows learners to develop higher-order thinking skills (Chekaraou, 2004);
· The stronger development of problem-posing and problem-solving skills in learners is allowed (Fredua-Kwarteng & Ahia, 2005 [online]); and
· greater cognitive development in learners is evident (WCED, 2007, p. 2)
L1 LoLT use is also likely, for pedagogical efficiency reasons, to reduce inequality, since it minimises classroom miscommunication and benefits learners’ desire to learn while bolstering their self-esteem as able learners (Smits, Huisman, & Kruijff, 2008, p. 10).
Code switching and code mixing (CSM) are communication strategies frequently found in both L1 and L2 LoLT contexts employed by both teachers and learners in disadvantaged SA classrooms. Switching between L1 and L2 sentences during speaking is referred to as code switching, while the use of both the L1 and L2 within the same sentence is referred to as code mixing (Heugh, 2002, p. 190). The National Centre for Curriculum and Research Development in SA (DoE, 2000, p. 68) observes code switching in SA to be a “main linguistic feature in classrooms where the teacher and the learners share a common language, but ha[ve] to use an additional language for learning … the learners’ language is used as a form of scaffolding”.
CSM are variously employed by teachers, according to their views on the strategies and their own language skills, to help learners understand what is being taught and to increase learner participation during lessons (Probyn et al., 2002, p. 35), and involve teachers’ and/or learners’ mixed use of the LoLT and teachers’/learners’ L1 during lessons. These strategies are also commonly used by teachers to mitigate their own limited proficiency in using the LoLT (Banda, Mostert & Wikan, 2012, p. 24).
Generally, however, teachers employ CSM while they teach as they feel they have no choice but to do so within their circumstances, as the following quote from research conducted by Mwinsheikhe (2003, p. 16) illustrates: “I personally was compelled to switch to Kiswahili [learners’ L1] by a sense of helplessness born of the inability to make students understand the subject matter by using English”.
As such, the use of learners’ L1 for explanatory purposes is the main reason for code switching by teachers in lessons in L1 and L2 LoLT contexts (Probyn et al., 2002, p. 35). Teachers also make use of learners’ L1 via code switching for various additional functions that fall under the categories ‘classroom management discourse’ and ‘interpersonal relations’ (Ferguson, 2003, 16, p. 2). Within L1 LoLT contexts, teachers and learners have also been found tocode switch to English when the L1 lexicon is inadequate, inefficient, or does not exist. Code switching for this reason occurs most within subjects such as Mathematics and Science.
Learners in L2 LoLT classrooms have been found to code switch predominantly in order to respond to questions, during individual or group interactions, as well as during whole-class interactions when engaging in discussion (Setati et al., 2002, p. 81).
Advantages of code switching/mixing in classrooms
The primary advantages of code switching within L2 LoLT classrooms are also the main reasons why code switching is employed by teachers and learners: to be understood, to understand, and to communicate. As such, within multilingual classrooms, code switching can be considered a social and linguistic skill (Mesthrie, 2008, p. 67) used to facilitate communication between people of different cultures (Wolff, 2011, p. 66) as well as to facilitate learning by improving the communication of curriculum between teachers and learners (Brock-Utne, 2007, p. 521). As a pedagogical strategy, code switching is useful as it allows what research refers to as exploratory talk within classroom conversations, where learners’ L1 is used to explore concepts and ideas to a degree greater than would have been possible if they were to have been explored in the L2 LoLT alone (Setati et al., 2002, p. 78).
Research on L2 Mathematics teaching in SA has found that code switching, usually to give learners an explanation of concepts and processes in their home languages, is educationally beneficial and necessary for adequate learning to occur (Setati, Chitera, & Essien, 2009, p. 72).
In light of the educational advantages of code switching and mixing in classrooms, very little training is provided for teachers in SA on how to use code switching strategies as a linguistic resource in a manner that is planned and strategic (Probyn, 2008, p. 220).
Disadvantages of code switching/mixing in classrooms
Despite code switching as a practice in education being viewed positively as an educational resource (Setati et al., 2009, p. 66), its use remains dilemma-filled and an issue of considerable debate as a result of problems surrounding code switching’s impact on education. While most disadvantaged SA schools report an English L2 LoLT, this is largely inaccurate. Within such schools learners are expected to write and read in English while being taught using L1/L2 CSM (Heugh, 2013, p. 221). This difference between official LoLT and unofficial classroom practice has negative consequences for learner progress (Probyn et al., 2002, p. 42).
Despite having been often taught in L1 LoLT, when assessed, SA learners are required to answer questions in the L2 LoLT and will receive zero marks for any answer provided in their L1 even if they are correct (Brock-Utne & Alidou, 2006, p. 116). By their use of code switching during lessons, teachers set learners up for failure during assessment in this way. Studies conducted throughout Africa paint a related picture (Kalole, 2004), finding learners’ incompetence in the L2 LoLT to be the predominant factor negatively influencing learner performance during assessment.
Code mixing is even more prevalent in SA than code switching as a practice used by teachers during teaching (Heugh, 2002, p. 191). As such, the model of language provided in such schools is often a code-mixed model, the closest that teachers who are teaching in a second language can get to providing L2 LoLT learners with education in English. This is problematic within contexts where the goal is to learn a language, as very clear linguistic boundaries between the L1 and L2 need to exist in order for learners to learn either of the two languages well (Murray, 2004, p. 440).
This is because the process of developing a learner’s proficiency in a language is one facilitated by expanding the learner’s capacity to link an idea associated with a spoken word to text as a form of that word made visible in writing. For this to occur, the learner must be repeatedly provided with a correctly pronounced spoken word. This becomes impossible if a teacher is unable to provide learners with the correct pronunciation of a spoken word because of the teacher’s lack of proficiency in that language, or if the teacher frequently switches between words that are pronounced differently in the L1 versus the L2.
Developmental psycholinguistic research also indicates that learners in the foundation years of schooling are disadvantaged by CSM practices in this way because of the significant confusion that these practices cause for them while attempting learn a L2 while also navigating and comprehending new conceptual understandings (Henning, 2012, p. 69). The reason attributed to this is that such learners have not yet fully developed the socio-pragmatic skills (Myers-Scotton, 2002) that teachers use when they code switch or mix and are therefore psychologically unprepared (Cantone, 2007) to cope with the demands of learning from teachers who code switch or mix.
Research highlights code switching/mixing in early education as problematic because these practices: 1) create barriers to learning, and 2) undermine the cognitive development of learners because of the intersection that exists where language and conceptual development must meet in order for cognitive growth to occur – but do not meet (Henning, 2012, p. 75). It is also argued that code switching in classrooms undermines learners’ development in learning a language as they are not challenged by the inputs that they receive during teaching and learning to learn that language further (Gough, 1993, p. 2).
The experience of schooling for a learner who is taught in his or her L2 is more uncomfortable than that for a learner taught in his or her L1, for various reasons. Learners who are taught in a language that is foreign to them find school intimidating as they are not able to understand what his or her teachers are saying to them. Within such situations, learners become confused, bored and withdraw (Jhingran, 2005).
Further exacerbating the discomfort and frustration experienced by learners in L2 LoLT classrooms is the tendency in such circumstances for teachers to employ coercive measures intended to force learners to use the L2 LoLT. Such coercive measures range from learners who do not comply with the LoLT being humiliated to their being beaten in public. The effects of such measures include further reduced learner participation during lessons, increased learner anxiety associated with school, greater educational disaffection among learners, and the increased risk of learner dropout (Smith, 2003).
Teachers who must teach in a L2 also experience discomfort using the LoLT when they lack the proficiency in the language to do so effectively. In a survey of teachers in townships around Cape Town, it was found that many Xhosa teachers felt that they were not adequately trained to do so. This is reflected in a comment by one such teacher made during this survey: “Some of us are afraid of talking English. We feel uncomfortable because we don’t want to be laughed at, as well as making mistakes” (Hibbert, 1995, p. 5).
Research conducted in SA has led to concerns around the negative effects of subtractive bilingualism on cultural and linguistic heritage, these being the loss of L1 language proficiency and associated cultural identity (Murray, 2004, p. 438).
Relating to the linguistic threat of L1 loss that African learners are faced with when learning in a L2 LoLT is the challenge that L2 LoLT use poses to the African learners’ cultural identities when “the social conditions of learning devalue the child’s first language and associated culture” which may also “impede (their) cognitive and social development” (Luckett, 1992, pp. 46–47). Learners who are submerged at school in a L2 show losses of self-confidence and self-esteem (WCED, 2007, p. 2).
Contrary to L2 LoLT effects, L1 LoLT use has been found by research to benefit learners psychosocially and culturally. L1 instruction has been shown to hold significant benefits for learners, such as developmental self-confidence and a strong self-concept, as well as holding advantages for the preservation of minority languages and cultures (Smits, Huisman, & Kruijff, 2008, p. 8-9).
Various factors are typically found to exist within learners’ home environments, communities and schools that have an effect on and/or are affected by LoLT policy implementation.
While more privileged learners are supported in their learning at home by parents who are literate in the LoLT of their children and can provide learning conducive to the optimal development of literacy, for most L2 LoLT learners this is not the case. Their progress is hampered by factors that make achieving literacy more difficult. Such factors include lack of learning resources, inadequate parental support of learning, and the lack of parental involvement in learners’ learning, as parents have often missed out on education themselves and/or cannot speak, read, write or understand their child’s LoLT (UNESCO, 2008).
Compared to L2 LoLT use, which is shown to impact negatively on the degree to which parents support, and are involved in, their children’s learning, the use of learners’ L1 as LoLT in schooling is shown to lead to greater involvement of parents in their children’s education as well as having the effect of promoting more favourable parent and learner attitudes towards school and education (Brock-Utne & Alidou, 2006, p. 112).
Many parents within disadvantaged contexts consider the need for their children to learn English as vital for their success in life (Pather, 1994) and because of this most such parents choose to have their children taught in English from as early in the child’s life as possible, at the expense of the L1 (Wolff, 2006, p. 42). This sentiment is illustrated by the following quote:
Many people have come to accept that ‘real’ education can only be obtained in a world language such as English. Even the idea that a child will benefit if his or her initial education is given in the first language is disputed by many so-called educated parents (Bamgbose, 2000, p. 88).
This sentiment is one shared by African children and has contributed to the rise of what is known colloquially as the coconut or Oreo phenomenon. It is characterised by young African people shunning their cultures and languages in favour of English and Western ways. The result of this is the diminished use of African languages and loss of African culture, further worsened by a shortage of African writers as well the demise of African oral tradition. To use English as a LoLT aggravates this pattern of loss, as this quote reflects:
There is little doubt that the systematic but frequently ignored differences between the language and culture of the school and the language and culture of the learner’s community have often resulted in educational programmes with only marginal success at teaching anything except self-depreciation (Okonkwo, 1983, p. 377).
The survival of African languages is also threatened by inherent linguistic difficulties within certain contexts on usage. For example, African teachers and learners tend to use English terminology when engaging with communication around mathematical and scientific concepts. The reasons for not using the African terminology is that if it does exist, it is often difficult to use and cumbersome in its usage, or is otherwise unavailable or unknown (Probyn, 2006, p. 396; Setati & Barwell, 2008, p. 3; Vorster et al., n.d., p. 151; Wolff, 2006, p. 43).
SA learners from townships are further disadvantaged by the lack of exposure to any pure form of language, whether it be their own language or English. This results in confusion as learners are exposed to mixtures of various different languages in an unstructured manner that undermines their progress in learning their language as well as English, resulting in learners not knowing enough of either language in order to engage optimally with lessons using them (Probyn, 2006, p. 406).
Within disadvantaged SA contexts, lexical borrowing, defined by Haugen, (1972 [1950]) as “the adoption into one language of items, patterns and meanings from another”, in addition to code switching and mixing, has been seen to give rise to varieties of unofficial languages that differ from area to area as local hybrids that have resulted from the melting of different linguistic codes in one pot. These varieties of language have developed as different languages have been used together in a mixed manner over time to the extent that a degree of convergence occurs between the languages being mixed (McCormick, 2004, p. 219), forming new ‘languages’ that are typically used informally. Slang or gangster language is also evident within such contexts.
In addition to the linguistic factors mentioned above that contribute to a learner’s linguistic confusion while trying to learn several languages simultaneously, within the languages themselves, particularly in the case for isiXhosa in the SA context, great linguistic variation is found to exist from region to region in the form of differing dialects of the same language (Branford & Claughton, 2004, p. 199).
The ability of LoLT policies to facilitate learner performance is considered to be hampered by two sets of factors (Wolff, 2006, p. 50). The first set comprises those related to negative attitudes held by all stakeholders toward the use of African languages as LoLT. The second set of factors contributing to the failure of LoLT policy to produce good educational outcomes are those to do with the improper planning and implementation of LoLT policy.
Teacher training is also viewed as impairing LoLT implementation. Quality levels of pre-service and in-service teacher training in SA are low, and support to teachers in classrooms is inadequate (Modisaotsile, 2012, p. 2). The quantity of available in-service teacher development is minimal. Teacher training institutions in SA do not fully equip African home language-speaking teachers to teach English, or to teach in English – leaving newly qualified teachers unconfident in doing so, as well as likely to transfer their language errors to learners once in the classroom (Nel & Muller, 2010, p. 646).
The nature of teacher training available in SA is noted as largely inappropriate for the linguistic context of the country in that teacher training curriculum gaps around language education in multilingual classrooms (teaching methodologies, didactic tools and theoretical knowledge) exist (Alidou & Brock-Utne, 2006, p. 87; Calteaux, 1996, p. 158; Probyn, 2008, p. 220).
Learners’ educational progress is also further hampered within disadvantaged SA schools by a lack of basic material resources for use in teaching and learning, such as classrooms and furniture, reading books, textbooks, dictionaries, and computers, as well as stationery supplies (Prinsloo & Janks, 2002, p. 37; Probyn, 2006, p. 406). A lack of printed material resources in African languages is particularly apparent within such schools, and when such resources are available they are often of poor quality (Banda, Mostert, & Wikan, 2012, p. 33).
Having established the policy context within which the research was conducted, the next chapter will detail how the research was conducted. In this chapter, all of the elements that made up the research design employed will be specified, including research methods, research tools, sources of data and, procedures of data collection and analysis. Potential limitations and ethical considerations will also be addressed.
[1] Taylor & Coetzee (2013) provide a good summary of this research.