This chapter presents the results of the field-work in two parts; A and B.
Part A utilises the data from WCED Systemic Testing to demonstrate the relationship between choice of LoLT and performance in Mathematics and language. That is the different outcomes for FP L1 (early-exit transitional) and FP L2 (straight-for-English) LoLT models.
Part B utilises the interview and questionnaire data to illustrate how LoLT(s) are implemented in reality.
Part A compared learner performance from schools designated as following:
· An isiXhosa LoLT in the Foundation Phase (FP L1 LoLT); then
· Early-exit LoLT transitioning to English from Grade 4, to schools designated as following:
· An English LoLT in the Foundation Phase (FP L2 LoLT) and throughout subsequent grades (a form of straight-for-English submersion model), at particular points in time as well as over time.
The findings produced from this are presented quantitatively using descriptive statistics and then discussed. When interpreting learner performance test scores, it must be remembered that testing was conducted in the language of the LoLT. For example, Grade 3 FP L1 LoLT testing was conducted in isiXhosa, and Grade 3 FP L2 LoLT testing was conducted in English. Furthermore, it must be noted that because of the LoLT transition from isiXhosa to English in Grade 4 that occurred at all FP L1 LoLT schools included in the research, the testing of Grade 6 learners was conducted in English – as it was for Grade 6 learner testing in FP L2 LoLT schools.
FP L1 LoLT and L2 LoLT schools were compared by their Grade 3 learners’ performance in WCED Systemic Testing of Language and Mathematics in 2012 and 2015. The two groups of schools were also compared by their Grade 6 learners’ performance writing the same tests in 2015. The findings from these comparisons are reflected in Table 4.1, below.
a) Cross-sectional comparisons of Grade 3 performance in 2012 and 2015
As is evident in Table 4.1, below, it was found that Grade 3 learners in FP L1 LoLT schools performed better on average in both WCED Systemic Testing of Language and Mathematics than learners in FP L2 LoLT schools did in both 2012 and 2015:
· FP L1 LoLT learners out performed FP L2 LoLT learners during language testing by 5,3% and 9,5%, in 2012 and 2015, respectively. While this difference in performance was not found to be statistically significant in 2012, it was in 2015; and
· FP L1 LoLT learners achieved better results than FP L2 LoLT learners did during Mathematics testing to a very statistically significant extent in both 2012 and 2015, by 9,7% and 12,1%, respectively.
It is to be expected that Xhosa learners who wrote tests of language in their L1 would perform better than those who wrote it in a L2 because they had had more exposure at home to, and had mastered to a greater extent, the language assessed than was the case for learners assessed in English. Compared to FP L2 LoLT learners, FP L1 LoLT learners would, therefore, be better able both to understand the language used in the assessments as well as respond to the assessment in the language required. This advantage held by FP L1 LoLT learners over FP L2 LoLT learners pertains to understanding why the former group of learners performed better in Mathematics, as well.
When understanding the superior performance of FP L1 LoLT learners, it should be remembered that they were likely to have been able to understand explanations of, and learn, language and mathematical concepts during lessons to a greater extent than FP L2 LoLT learners were, because FP L1 LoLT learners were taught in a language in which they were proficent – unlike the case for Xhosa learners taught in English.
The reasoning provided above indicates that FP L2 LoLT learners performed in Language and Mathematics to a lesser extent than FP L1 LoLT learners because of the L2 LoLT used; suggesting that English was an additional barrier to both teaching and assessing FP L2 LoLT learners.
Furthermore, it is reasonable to speculate that the predominantly Xhosa teachers of FP L1 LoLT learners were able to provide teaching of a higher quality because they were teaching in their L1, whereas this was not the case for FP L2 LoLT learners.
b) Cross-sectional comparisons of Grade 6 performance in 2015
As reflected in Table 4.2, below, it was found that on average Grade 6 learners in FP L1 LoLT schools performed worse during 2015 WCED Systemic Testing of both Language and Mathematics than those in FP L2 LoLT schools did:
· FP L1 LoLT Grade 6 learners’ performance in Language was 3,7% lower than that of FP L2 LoLT learners, a statistically significant difference; and
· FP L1 LoLT Grade 6 learners’ performance in Mathematics was 2% lower than that of FP L2 LoLT learners.
When understanding these findings, it must be remembered that by the time testing occurred in English FP L2 LoLT, Grade 6 learners had been schooled using English designated as LoLT for a greater period of time (+/– 6 years) than was the case for FP L1 LoLT Grade 6 learners (+/– 2 years’ English LoLT exposure) and that this was likely to have advantaged FP L2 LoLT Grade 6 learners in terms of being able to understand the language of assessment, respond to it using the required language, and perform well.
Related to this, because of FP L2 LoLT Grade 6 learners’ greater exposure to English as designated LoLT from Grade 1, these learners were likely to have been advantaged in their proficiency understanding explanations of, and learning, concepts during lessons delivered in English, in comparison to FP L1 LoLT Grade 6 learners, who had begun being taught in English as designated LoLT only from Grade 4. If this was the case, which was probable, FP L2 LoLT Grade 6 learners would have been better prepared to write assessments because more learning had occurred subsequently than was the case for FP L1 LoLT Grade 6 learners.
From the above, the findings indicate that English as LoLT (and the language of assessment) was possibly an additional barrier to learning, and demonstrating learning, for FP L1 LoLT Grade 6 learners, when compared to FP L2 LoLT Grade 6 learners, who were better prepared to deal with overcoming it.
That learners following a straight-for-English LoLT model performed better than those in an early-exit transitional model context flies in the face of prevailing research on the matter. As noted at the beginning of this chapter, the LoLT model designation FP L2 LoLT was not found to be accurate in terms of the model’s implementation. This will be discussed later.
In order to secure an indication of how learners in the alternate LoLT contexts performed over time, FP L1 LoLT and FP L2 LoLT schools were compared by the difference in their learner performance in WCED Systemic Testing of Language and Mathematics between 2012, when learners were in Grade 3, and 2015, when learners were in Grade 6.
As reflected in Table 4.3, below, the comparison of learner performance difference in WCED Systemic Testing of Language and Mathematics between 2012 (when learners were in Grade 3) and 2015 (when learners were in Grade 6) showed that:
· FP L1 LoLT schools performed on average 6,9% lower than FP L2 LoLT schools did during the assessment of Language – which was found to be a statistically significant difference; and
· FP L1 LoLT schools performed on average 11,2% worse than FP L2 LoLT schools did during the assessment of Mathematics – which was found to be a very statistically significant difference.
As was the case for the previous comparison, when understanding the higher performance of FP L2 LoLT Grade 6 learners over FP L1 LoLT Grade 6 learners, it should be noted that the former were likely to have been comparatively advantaged when assessed by their greater proficiency in the LoLT in terms of better-developed abilities to understand, learn and perform during testing, using the LoLT gained as a result of larger exposure to English as the designated LoLT. Similarly, the L2 LoLT was again indicated as being a barrier to teaching, and the assessment of learning occurred for FP L1 LoLT Grade 6 learners because of their more limited exposure to English as designated LoLT.
As was the case for the previous Grade 6 learner performance comparison, discussed later, the interpretation of these results must be tempered by what the research found to be an inaccurate designation of the LoLT model, particularly FP L2 LoLT.
From Part A, the following conclusions were evident:
1. Learner performance was optimal when learners were taught and assessed in their L1. This was apparent from Grade 3 learner performance cross-sectional comparisons and was in line similar research conducted, such as that by Broom (2004, p. 506), who found the same trend in her comparative study of Grade 3 learner performance in disadvantaged SA schools.
2. The performance of learners within early-exit LoLT transitional models typically declined after the point of transition. Again this confirmed previous research (see for example Thomas & Collier, 1997).
3. The last conclusion was more nuanced than the first two. The comparison of FP L1 LoLT vs FP L2 LoLT learner performance over time between Grade 3 and Grade 6 indicated that FP L2 LoLT learners were likely to perform better in the long run, signalling perhaps that straight-for-English submersion LoLT model may be more advantageous to learners. On the face of it, this finding goes against the prevailing research (see, for example, Taylor & Coetzee (2013, p. 7)).
In order to explain this anomalous and perhaps far-reaching finding it was necessary to visit schools to see how the LoLT models described above were enacted in practice.
While Part A aimed to provide an indication of learner performance typically associated with FP L1 (early-exit transitional) and FP L2 (straight-for-English submersion) LoLT models, Part B aimed to gather an indication of how the LoLT models were implemented and what effect implementation had on teaching and learning, and ultimately learner performance.
Part B’s discussion begins by providing a snap shot of the manner in which LoLT policy was indicated as having been formulated within the schools researched, and how respondents felt about the LoLT policy at their schools. The nature of LoLT policy implementation (or non-implementation) forms the bulk of the discussion. An overview of the reported effects that LoLT policy had on learners brings this section to a near end. It concludes with a summary of findings.
Findings that contextualise the LoLT models researched in terms of factors contributing to how they were formulated are presented below, briefly. These include decision-making, consultation of legislation and theory, and how teachers felt about the LoLT in their schools.
a) School LoLT policy decisions
Regarding the manner in which LoLT decisions had been made in schools visited:
· Parents had rarely been consulted;
· Parents’ preferences carried little influence;
· SGBs had usually been involved; and
· The DBE’s preferred LoLT model was generally the deciding factor.
The findings above illustrated that the prevailing legislation, such as the Constitution of the Republic of South Africa (RSA, 1996a) and the South African Schools Act (RSA, 1996b), legislating the manner in which LoLT policy was formulated was not fully adhered to. This was further indicated, below.
b) Legislation and theory consulted
Regarding whether relevant legislation and language theory had been consulted during LoLT policy formulation, of 14 interviewees:
· Roughly 50% did not know about LoLT legislation.
· 29% said legislation was consulted.
· 86% said that they did not know about language theory or that it had not been consulted.
· 14% said language theory had been considered.
From the above, a distinct lack of legislation and theory consultation was evident during LoLT formulation processes.
c) Sentiment towards LoLT policy
Interviewees were more negative (77%) than positive (23%) about the LoLT policy in their schools.
Figure 4.1: General LoLT policy sentiment
Reasons for differences in sentiment formed a pattern of why respondents were either pro- or anti-LoLT policy.
Negativity towards isiXhosa LoLT use up until Grade 4 was predominantly a result of the problematic LoLT transition at the start of that grade. Further negativity toward isiXhosa as LoLT was found to stem from isiXhosa-specific linguistic challenges that will be detailed later. Negativity towards isiXhosa as LoLT was also found to spring from the belief that the language did not serve learners’ future best interests. Interviewees were positive about isiXhosa LoLT use during the early years of schooling because of the benefits of mother tongue education for learners.
Negativity towards English as LoLT was found because of the fear that learners lose their language and culture by not learning in their L1. Positivity toward English as LoLT stemmed from the belief that English is necessary and beneficial for learners’ future education and employment. English LoLT use was viewed positively because it was the most accessible LoLT choice within heterolinguistic contexts.
The findings above indicated that respondents were of the opinion that prevailing LoLT policy did not serve their purposes. In light of those, respondents were asked what their LoLT preference was, indicated below.
d) Prevailing LoLT policy preference
The prevalence of LoLT preference among 14 interviewees was:
1. 57% indicated a preference for the straight-for-English LoLT model.
2. 29% indicated a preference for an early-exit transitional LoLT model.
3. 14% indicated a preference for a late-exit transitional LoLT model.
Having presented an indication of how research into LoLT policy in schools came to exist, as well as how teachers felt about it, the discussion now turns to evaluating how LoLT policy was implemented.
This section begins with a discussion of the extent to which LoLT policy was found to be implemented, moving on to include a reflection upon how varying levels of teacher and learner proficiency in the LoLT affected implementation, and the teaching and learning practices associated with the implementation of LoLT, and concludes with an evaluation of factors beyond the classroom indicated as affecting implementation.
4.2.2.1 Extent of implementation
LoLT was found to be implemented to a greater extent within L1 contexts than in L2 contexts, see Figure 4.2, below. The extent to which LoLT implementation differed between L1 and L2 contexts was described by a principal using figures to illustrate his assertions:
In the Foundation Phase [where L1 LoLT is policy] up to 100%, but in the Intermediate and Senior Phase [where L2 LoLT is policy] I would say it is about 50% (Principal C, principal at FP L1 LoLT school).
Interviews confirmed this quote. They also found that the causes of non-implementation differed between L1 and L2 LoLT contexts, forming two patterns. Both patterns congregated around a major factor. This factor was a lack of LoLT accessibility of use – in terms of efficiency and effectiveness – by both teachers and learners as a result of smaller factors specific to either context, or to both, to be discussed later.
Figure 4.2: Extent of LoLT implementation
Generally, non-compliance was driven by necessity when teachers and/or learners needed to use a language other than the LoLT to communicate, employing mixed-language-use practices such as CSM to do so. These practices constituted the almost-exclusive source of LoLT non-implementation to the extent that the prevalence of CSM within LoLT contexts invalidated the designated LoLT descriptor, especially in L2 LoLT contexts. Therefore, conclusions relating to the LoLT model as it was specified could not be drawn. Instead, any conclusions made regarding the LoLT models researched needed to be tempered by a disclaimer that stated that LoLT designations were inaccurate. Additionally, the nature of the LoLT as it was implemented in practice within classrooms was required to be specified.
No evidence of intentional defiance by teachers and/or learners as a cause for LoLT non-compliance was found.
When reflecting during interviews on who was responsible for this, significant uncertainty was expressed by respondents. Various role-player groups fingered one another. A lack of accountability for LoLT implementation within schools was also evident.
Now that the extent to which LoLT policy was found to be implemented within the schools visited has been indicated, various factors found to affect LoLT implementation are explored.
4.2.2.2 Teacher and learner proficiency
How teacher and learner LoLT proficiency reportedly affected implementation of the LoLT (or not) during teaching and learning are discussed below. LoLT proficiency is explored as ability, comfort, and participation using the LoLT. Findings are presented in order of significance.
a) Teacher and learner ability
Teachers’ and learners’ ability to use the LoLT was found to have the greatest impact on the degree to which the LoLT was conducive to learning, and able to be implemented.
Teachers in both contexts during every interview specified that learners’ inability to understand the LoLT was the largest factor preventing education within their school. While “the kids don’t have [a] flipping clue of what English is about” (Teacher J, Grade 3 teacher and FP HOD at FP L1 LoLT school), learners were also reportedly challenged learning in isiXhosa.
What this reflected for education in disadvantaged schools of the Western Cape was summed up like this: “In actual fact we have two problems at the same time. In English and in isiXhosa” (Principal D, principal at FP L1 LoLT school).
Confirming interview findings, questionnaire responses indicated that learners understood teaching in L1 LoLT better than they did in a L2 LoLT: 100% of L1 LoLT teachers responded positively with ‘Almost always true’, or ‘Usually true’, or ‘Learners understand isiXhosa teaching’, while 75% of L2 LoLT teachers responded negatively with replies that ranged from ‘Often true’ or ‘Rarely true’ to ‘Learners understand English teaching’.
In light of the above, it is understandable how a L2 LoLT was considered by this principal to be a learning barrier:
Because English itself not as a subject is a barrier to any isiXhosa-speaking person. English is a barrier because it is a foreign language. As much as isiXhosa is a barrier to an English-speaking person (Principal C, principal at FP L1 LoLT school).
While language constituted a barrier to learning within both L1 and L2 classrooms, as reflected in the quote above, the extent to which English was a barrier to learning for Xhosa learners was reportedly particularly detrimental as these learners were not able to learn in English optimally.
These findings highlighted that assessment of learners in a language that they are not adequately able to understand or express themselves in is problematic, because the results produced were invalid as they did not accurately reflect learning held by learners.
Learners within both LoLT contexts tended to use their L1 when speaking because they were less able to use the L2 to do so. Of 11 interviewees who mentioned proficiency of speech, 73% noted this trend. Within FP L2 LoLT contexts, this was a significant source of non-implementation. In this regard, questionnaire responses (see Table 4.4, below) showed learners were better able to communicate in the L1 than in the L2.
The inability of learners to speak the LoLT was not something particular to L2 LoLT contexts, however, and was thus a cause of LoLT non-implementation in these contexts, too. 36% of interviewees in L1 LoLT contexts mentioned how learners encountered difficulties speaking the ‘original’ isiXhosa required in the classroom because they were used to speaking informal versions of language that were ‘improper’, ‘slang’, ‘township’, ‘impure’, ‘tsotsitaal’, ‘kweri-kweri’, ‘social’, ‘mixed’, ‘abbreviated’, of the incorrect dialect, or, as one teacher at a FP L1 LoLT school put it, “something gibberish, something localish”.
Table 4.4: Comparison of teacher and learner ability in the LoLT
Insufficient learner vocabulary in the LoLT was found to be a LoLT implementation challenge within both contexts. It was noted that in L1 LoLT contexts this challenge was particularly problematic when learners moved into Grade 4 where a L2 LoLT was used. Within L1 LoLT contexts, it was indicated that Xhosa learners often did not hold vocabulary in isiXhosa for mathematical, scientific and technological concepts such as, for example, the names of colours, concepts of time, or number names. The contrary proved to be true; Xhosa learners often knew such vocabulary in English. As such, for example, one teacher expressed how she had often encountered learners who were not able to use letters of the alphabet in isiXhosa, but knew the A-B-C in English. She added to this that even some Xhosa teachers did not know the alphabet in isiXhosa.
The ability of learners to read and write in the LoLT was flagged as problematic within both contexts. Interestingly, in light of the fact that isiXhosa was learners’ L1, there was 28% more feedback from interviewees regarding the inability of learners to read and write in the language than from interviewees regarding the same problem in English. This barrier to implementation was reflected in questionnaire responses (see Table 4.4, above).
Reasons for learners’ insufficient ability to read and write isiXhosa centred around the challenging nature of the language, which made it difficult for learners to use. This was further aggravated by a high prevalence of mixed-language learner vocabularies and a predominance of English within learners’ everyday environments.
While these findings regarding learners’ inability to read and write in isiXhosa were significant pertaining to the use of the language itself, they were also significant pertaining to learners’ ability to read and write in English. As one teacher pointed out, if Xhosa learners are not able to read and write in isiXhosa, learning to do so in English is made even more difficult. As such, learners’ ability to read and write in the LoLT was found to be a stumbling block to its implementation.
Of all the reported challenges faced by teachers in both contexts implementing the LoLT, the greatest was being able to be understood by, and understand, learners in order to teach. Both isiXhosa and English use as LoLT affected the degree to which teachers were able to teach as a result of various factors, some specific to each language. Interviewees indicated that Xhosa teachers were better able to implement LoLT teaching in isiXhosa than they were in English as learners were better able to understand isiXhosa than English.
It was noted that more teachers stated that they had problems teaching in isiXhosa because of the challenging linguistic nature of the language than whose who noted that they had problems teaching in English for the same reason.
Despite challenges to LoLT implementation posed by the difficulty of isiXhosa, questionnaire responses indicate that L1 LoLT teachers were more comfortable planning lessons than L2 LoLT teachers were. Teachers were reportedly also better able to assess in the L1 LoLT than they were in the L2 LoLT.
The delivery of curriculum was reportedly undermined where obstacles to teachers’ ability to teach existed. From this it was evident that obstacles to LoLT implementation resulted in inadequate curriculum delivery. Interviews and the survey indicated that insufficient curriculum coverage was double the problem for L2 LoLT teachers than it was for L1 LoLT teachers.
Unexpectedly, interviews found that inadequate teacher vocabulary was expressed exclusively as an implementation barrier within L1 LoLT contexts. The reasons provided for this fell into two categories, the first being a vocabulary deficiency on the part of the teachers themselves and the second being, to a lesser extent, a deficiency of vocabulary within the isiXhosa lexicon. Questionnaire responses reflected this issue (see Table 4.4, above).
Teachers’ ability to teach Mathematics and Science was found to be limited by the vocabulary available within the LoLT to do so. Within L1 LoLT contexts, Mathematics and Science teaching was impaired by a combination of the aforementioned teacher-and-learner vocabular inadequacy as well as the lexical deficiency of the isiXhosa language in Mathematics and Science. This was aggravated by the prevailing predominance of English usage when engaging with mathematical and scientific concepts.
Despite challenges L1 LoLT teachers faced in teaching Mathematics and Science in isiXhosa, the questionnaire reflected what interviews had found; that L1 LoLT teachers were reportedly better able to teach Mathematics and Science than L2 LoLT teachers.
Xhosa teachers within L1 LoLT contexts were also reportedly challenged in their understanding of ‘academic isiXhosa’, which impaired the extent to which they were able to implement the L1 LoLT while teaching. This was because of issues pertaining to the prevailing nature of isiXhosa usage.
Teachers’ ability to engage learners’ higher-order thinking by implementing the LoLT was of interest as doing so is required for of facilitation of quality education. L1 LoLT teachers indicated they engaged learners’ higher-order thinking to a greater extent than L2 LoLT teachers did.
b) Teacher and learner comfort
Teachers’ and learners’ comfort teaching and learning using the LoLT was of interest because teachers and learners teach and learn best when comfortable with the LoLT used. Furthermore, LoLT was less likely to be implemented if teachers and learners were uncomfortable using it. In this way, comfort levels provided an indication of the degree to which implementation was likely to occur.
The most prominent learner comfort factor influenced by LoLT was the extent to which learners felt comfortable (and able) to use the LoLT to express themselves. Within both contexts learners were reportedly more comfortable using isiXhosa to express themselves than English. Of the 10 interviewees who addressed learners’ freedom of expression using the LoLT, 9 mentioned that learners were more comfortable using their L1 than English. Accordingly, teachers indicated that learners enjoyed learning in L1 LoLT more than in L2 LoLT. These findings were reflected by the questionnaire (see Table 4.5, below).
The degree to which teachers were comfortable speaking the LoLT was of interest since teachers provide optimal education when comfortable using the LoLT. It was found that teachers were more likely to implement the LoLT when they were comfortable speaking it.
Table 4.5: Comparison of teacher and learner comfort between LoLT context
From 12 respondents who explored teachers’ comfort when speaking the LoLT, 67% expressed that teachers were more comfortable speaking isiXhosa during teaching while 33% indicated that teachers were more comfortable speaking English during teaching. Questionnaire responses confirm this finding (see Table 4.5, above). The reasons provided for discomfort speaking isiXhosa related to the difficulty using the language for teaching.
It was also apparent from questionnaire responses (see Table 4.5, above) that L1 LoLT teachers were more comfortable writing in isiXhosa than L2 LoLT teachers.
As further measures of comfort using the LoLT, the questionnaire (see Table 4.5, above) found L1 LoLT teachers reported smiling, laughing and joking more frequently while teaching than L2 LoLT teachers did. Noteworthy discomfort was expressed by L1 LoLT teachers regarding their ability to read official documents in isiXhosa.
Overall, it was found that teachers and learners are more comfortable with the L1 LoLT than with the L2 LoLT. Discomfort in using the LoLT was noted as a significant hurdle to implementation within L2 LoLT contexts.
Like comfort using the LoLT as an outcome of implementation, learner participation was of interest as it is required for optimal teaching and learning. The dynamic that was found to exist between LoLT and learner participation is explored in the next section.
c) Learner participation
It was indicated that the extent to which learners participated during lessons was contingent on learners’ ability to use the LoLT to express themselves. See Table 4.6, below. As such, respondents reported higher levels of learner participation in L1 LoLT contexts than in L2 LoLT contexts. It was evident that learners were inhibited in their participation as a result of their inability to respond using the LoLT in both L1 and L2 LoLT contexts, but mostly in the latter. As such, teachers indicated that learners participated during lessons to a greater extent within L1 LoLT contexts than in L2 LoLT contexts. L1 LoLT learners participated more frequently than L2 LoLT learners did, in various ways.
From this the research concluded that within L2 LoLT contexts, the occurrence of learner participation and LoLT implementation were to some extent mutually exclusive.
Table 4.6: Comparison of learner participation between LoLT contexts
As discussed during this section, the research found that although L1 LoLT implementation was not without its challenges, L2 LoLT implementation faced substantially greater challenges relating to the following factors existing within the classroom. Compared to L1 LoLT contexts, within L2 LoLT contexts teachers and learners were less able to use the LoLT for teaching and learning, teachers and learners were less comfortable doing so when they could, and learners participated to a lesser extent.
Together, these findings tentatively suggested that the quality of teaching and learning occurring within L2 LoLT classrooms is inferior to that within L1 classrooms. Exploring this possibility, the discussion will now reflect upon the nature of the teaching and learning practices that constituted the manner in which designated LoLT models were indicated as being implemented within classrooms.
4.2.2.3 Classroom practices
It was found that particular patterns of classroom practices particular to each LoLT context existed. The use of these practices evidently resulted from the extent to which teachers and learners were able to use, and were comfortable using, the LoLT in lessons. The nature of the practices employed were of interest because they influenced the quality of education that teachers provided as well as learners’ performance, directly. These practices were also indicative of the extent to which LoLT was implemented.
A discussion of CSM follows a brief overview of teaching practices indicated as commonly occurring within the schools visited. In this discussion, a large focus is placed on CSM practices within classrooms since of all the classroom practices mentioned by respondents, CSM practices were overwhelmingly prevalent.
a) Teaching practices
The following teaching practices that were indicated as in use are ordered by prevalence of respondents’ mention, from most to least.
Repetition was the most prevalent of the teaching practices reportedly used within classrooms. This methodology was reportedly used out of preference in both contexts, but more out of necessity in L2 LoLT contexts in order to ensure learner understanding, as illustrated by the following quote: “You really have to explain yourself. You really have to repeat yourself more than once and you have to ensure that they [the L2 LoLT learners] understand (Teacher H, Grade 3 teacher at FP L2 LoLT school).
Chorus teaching as well as chalk-and-talk-style teaching were found within both contexts but to double the extent within L2 LoLT classrooms.
Memorisation was also found to be equally as prevalent within both contexts. However, like repetition, the practice was reportedly used more out of necessity than preference within L2 LoLT classrooms. Learners were reported as using memorisation as a way of coping with learning in a language that was not fully understood by them in L2 LoLT classrooms.
All of the above-mentioned teaching practices were reportedly more prevalent within L2 LoLT classrooms than in L1 classrooms. As teaching practices that typically produce teaching and learning of a sub-standard quality, these findings indicated that the quality of education that existed within L2 classrooms was inferior to that which existed in L1 classrooms.
b) Code switching and code mixing
CSM are communication strategies involving the alternate and mixed use of different languages in close proximity to varying degrees. It was found that:
· CSM usage was significantly prevalent within both LoLT contexts;
· The degree to which these strategies was used was linked to the degree that the LoLT was usable by teachers and learners; and
· The use of these strategies during lessons affected learning performance positively or negatively, depending on how they were employed.
i) Prevalence in LoLT contexts.
During interviews, all interviewees in both LoLT contexts reported that teachers employed some form of CSM during teaching at the school; either by switching between isiXhosa and English between sentences or/and by using both languages within one sentence. It was indicated that learners typically employed CSM in a similar manner.
CSM were reported during interviews as prevalent to a greater extent (as much as double as prevalent in some instances) within L2 LoLT contexts than within L1 LoLT contexts. Teachers’ questionnaire responses reflected this trend.
The reportedly high prevalence of CSM practices within the classrooms of the schools researched was very significant in that this prevalence indicated that the official LoLT model designation of schools was invalid as the use of CSM practices constituted LoLT non-implementation.
Towards further understanding of the nature of CSM practices in relation to LoLT implementation and the resultant learner performance outcomes; how the practices were used within the respective LoLT contexts; the reasons why these practices were used, and which ones were used; teachers’ attitudes to the practices; and teachers’ responses to learners who employed the practices, as well as the reported consequences of the practices for teaching and learning, are reflected upon below.
ii) Pattern of usage between LoLT contexts
The manner in which CSM was reportedly employed differed from context to context:
· English words were typically used within isiXhosa sentences in L1 LoLT contexts;
· isiXhosa words were typically used within English sentences in L2 LoLT contexts; and
· Teachers in L2 LoLT contexts also tended to switch into isiXhosa when explaining concepts or words that learners had difficulty understanding.
CSM practices were found to be employed for different specific reasons within, and particular to, each LoLT context – but for the same general purpose: in order to facilitate communication during teaching and learning. The reported reasons for CSM are presented below in order of prevalence of indication during the research.
iii) Reasons for code switching and code mixing
All of the teachers interviewed expressed the necessity of CSM for learner understanding during teaching, within both contexts. However, CSM practices were reportedly more prevalent within L2 LoLT contexts than in L1 LoLT contexts.
The predominant reason for teachers’ CSM within L1 LoLT contexts was to facilitate the learner understanding of isiXhosa words that were difficult or unknown, typically those related to the realms of Mathematics, science and technology.
Other challenges to the use of isiXhosa LoLT that contributed to teachers and learners using CSM included:
1. A high prevalence of informal or non-academic isiXhosa use by Xhosa speakers, generally, and
2. A predominance of commonly known ‘easy’ words from English that substitute difficult isiXhosa words, mentioned previously.
Within L2 LoLT contexts, the reportedly most common reason for teachers’using CSM was to translate lesson content from English to isiXhosa and to explain new English words in the learners’ L1 so that learners were able to understand and learn them.
Teachers also code switched from English into isiXhosa out of necessity when, because of their inadequate ability to use English as a LoLT, the only way for teachers to communicate confidently was to use a language that they were proficient in.
CSM was typically employed by L2 LoLT learners in order to communicate with their teachers and one another, enabling their participation and learning during lessons. Within both L1 and L2 LoLT classrooms, learners tended to switch to the language they were able to express themselves in better. Learners were also inclined to respond to questions and participate in learning using their L1 since they were more comfortable speaking it than the L2 LoLT, and because they were aware they shared isiXhosa with their teachers as a mother tongue.
A further reason for CSM within L1 and L2 LoLT contexts was inadequate teacher and learner vocabulary. When teachers and learners needed to communicate a concept they did not hold the necessary words for, they switched using words from the language they held the needed vocabulary in.
Within L1 LoLT contexts specifically, teachers and learners switched to English when the isiXhosa language equivalent was laborious in its use, ineffective, did not exist or was unknown. Related to this, isiXhosa lexical inadequacy on the part of teachers and learners, in addition to as a language itself, gave rise to code mixing when teachers and learners appropriated non-isiXhosa words for use within isiXhosa sentences.
Within L1 LoLT classrooms, learners and teachers reportedly switched to English as a result of the difficult and inefficient nature of isiXhosa words and phonics. All of the L1 LoLT teachers interviewed expressed this sentiment. Learners and teachers within L1 LoLT contexts also reportedly switched from isiXhosa to English as a way of avoiding the use of isiXhosa because of what was described to be its laborious and time-consuming nature.
A combination of English predominance and impure isiXhosa usage of language within the environments of teachers and learners furthered the likelihood that they employed CSM practices within L1 LoLT contexts.
iv) Attitudes to code switching and code mixing
The sentiment expressed by management was that CSM were not encouraged as their use during lessons was considered to be deviating from the school’s LoLT policy, but that the practices were tolerated. The prevailing sentiment expressed by teachers toward the practices was the same; that they were not permitted. Additionally, interviewees generally indicated uncertainty of what the official DBE stance on the use of the practices during lessons was.
v) Responses by teachers to code switching and code mixing
The responses by teachers to learners who CSM during lessons were explored because such responses were either conducive to quality teaching and learning or not.
The most common response by teachers within both contexts to CSM by learners was reportedly to correct the language of the learner in the language in which they were having difficulty and to encourage the use of the LoLT. Teachers also often accepted answers from learners in a language other than the LoLT during lessons as well as during assessments.
Less common responses by teachers to learners who CSM were to ignore, respond in a manner that resulted in learner humiliation, or learner punishment – or some combination of these.
Only teachers within L2 LoLT contexts indicated that they reacted to learner code switching in this manner. To ‘I punish learners who do not use isiXhosa in class’, zero L1 LoLT teachers indicated they typically punished learners who did not speak the L1 LoLT, compared to half of L2 LoLT teachers whose responses to ‘I punish learners who do not use English in class’ indicated that they typically punished learners who did not speak the L2 LoLT in class.
vi) Consequences of code switching and mixing for teaching and learning
As dominant features within the classrooms of schools researched, the consequences of CSM practices during teaching and learning must be understood and considered when deciding upon appropriate policy for contexts similar to that within which to research was conducted.
It was indicated that the use of CSM during lessons held two significant consequences for the quality of education provided within both L1 and L2 LoLT contexts.
57% of interviewees within both contexts mentioned the first repercussion of CSM in classrooms; the development of a ‘language crutch’ in learners. This repercussion was problematic during assessment, when learners’ language crutches resulted in the assessment of their learning being invalid to some extent.
According to participants, learners developed the ‘crutch’ because learners’ development in the LoLT was undermined as they were not challenged to master the LoLT because of the assistance provided by teachers using CSM, resulting in learners not being as prepared to use the LoLT during assessment as they would have been if they had been challenged. Additionally, the assistance provided to learners in the form of CSM was absent when it came to learners’assessment as assessments were generally written in one language only, the LoLT, and did not include translations in learners’ L1.
The second consequence of CSM for teaching and learning was reported by 43% of interviewees within both contexts as being learners’ ‘linguistic confusion’, occurring when learners became confused between the languages in terms of phonetics and pronunciation, spelling, and vocabulary as well as syntax, and that this confusion hampered learners’ ability to learn both the L1 and L2. The linguistic confusion of learners was reportedly intensified within classrooms where teachers were inadequately able to model correct L1 and/or L2 usage.
vii) Code switching and code mixing for learning
It is evident from the reasons for teachers and learners reportedly employing CSM practices within classrooms that the use of these practices with in L2 LoLT contexts is generally unavoidable. Furthermore, despite the negative consequences reported as resulting from CSM, the purpose served by these practices during lessons is valuable in that they are used to the facilitate the communication ultimately required for teaching and learning to occur, as well as for the learning of a L2 to take place. Policy recommendations with regard to the use of these practices are made in the next chapter.
4.2.2.4 Challenges to LoLT policy implementation beyond the classroom
Various factors reportedly existed outside of the classroom within the home, within society and within the school that affected LoLT implementation. These will now be discussed.
a) LoLT challenges within the home
Within the home environment of learners, the reportedly most significant factor that negatively affected L1 and L2 LoLT implementation for learning was the low level of literacy amongst parents, resulting in an inadequate level of support of learning available to learners outside of school. The inability of parents to speak the LoLT further undermined the extent to which parents were able to support to their children’s education at home. This was found to be particularly the case in L2 LoLT contexts. A general lack of parental participation in learners’ learning was reportedly the next biggest problematic factor in the home environment, that impaired learner performance.
b) LoLT sociolinguistic and cultural challenges
During the research, it was noted that specific challenges to learners’ mastery of the respective LoLTs existed within learners’ society and these were influenced by culture. Such challenges included the effects of globalisation on minority cultures and their associated languages. These are reflected upon below.
i) Globalisation versus mother-tongue education
A definite tension was noted as existing between the need to learn the mother tongue as well as preserve its culture, and the need to learn English for social mobility.
Participants typically expressed their desire for learners to learn their home language and retain its culture. Similarly, participants indicated the need for learners to learn their L1 in order to maintain their identities. The need to adhere to principles of L2 acquisition involving the L1 being firmly established before learning a L2, including the idea of linguistic transference, was also evident in findings. At the same time, however, participants strongly indicated the inevitable need for learners to learn English in order to ensure a bright future, a sentiment that reportedly gave rise to Xhosa parents typically preferring that their children should be taught in English.
ii) Challenges to isiXhosa as LoLT
A set of inter-related challenges specific to isiXhosa reportedly existed that prevented the optimal implementation of the language as a LoLT for quality education. These challenges are addressed separately, below, as they were found not to apply to English to the same extent.
The greatest challenge to isiXhosa as a LoLT was a reportedly high prevalence of variation within the language, hampering learners’ mastery of the language because of the confusion that it caused. This confusion was indicated to be aggravated by the following contributing factors:
1. Prevailing impure usage of isiXhosa inside and outside of school, where the language was mixed with other languages during use;
2. Learners’ inadequate exposure to ‘pure’ isiXhosa as a result of prevailing impure isiXhosa usage;
3. Regional dialect differences within the isiXhosa in general use to which Xhosa learners were exposed; and
4. The influence of a high predominance of slang as well as abbreviated language use.
The confusion referred to was reportedly worsened by a predominance of English usage for certain purposes within learners’ social environments that had the effect of undermining isiXhosa LoLT implementation when learners were unable to use isiXhosa for these purposes at school when they needed to because they had not learnt to do so outside of school.
Learners’ learning of isiXhosa, and therefore its use as a LoLT, was also reportedly undermined by the generally decreasing prevalence of isiXhosa language use within learners’ environments outside of school. The undesirability of the isiXhosa as language of communication amongst learners was indicated as contributing to this and further undermining its use as a LoLT.
An additional factor that reportedly undermined the extent to which isiXhosa was able to serve as a LoLT was the language’s ‘user-unfriendly nature’, in comparison to English. isiXhosa was reportedly tiresome to use because of its phonetics and words being difficult to use, and also for being difficult to teach and learn.
Related to this, interviewees expressed their opinion of isiXhosa as a language that was inefficient and ineffective as a LoLT. The use of isiXhosa as a LoLT was to a notable extent, according to them:
1. Inefficient because of what was described as being the time-consuming nature of using the language;
2. Ineffective because of what was described as a deficiency in the isiXhosa lexicon.
Significantly, the productive implementation of isiXhosa as a LoLT was reportedly impaired by several mismatches of language that existed inside and outside of the classroom:
1. The isiXhosa that learners used within their home environments was to some extent not the isiXhosa that was required to be used within the classroom.
2. The isiXhosa used within the official documents differed to some degree from the isiXhosa that Xhosa teachers used and understood.
3. The isiXhosa used within isiXhosa teaching and learning resources varied from the isiXhosa that learners and teachers used.
4. The isiXhosa that teachers and learners commonly used differed from the isiXhosa that was used during external assessments in the language.
The sociolinguistic and cultural challenges to LoLT implementation mentioned above were noted as weighing heavily against the likely success of isiXhosa as a LoLT, instead framing English as a more suitable LoLT.
c) LoLT challenges within the school
Various challenges were found within schools that impeded the successful implementation of LoLT, as indicated by participants. They are presented in order of prevalence of mention by interviewees.
i) Inadequate prior learning of the LoLT
The greatest challenge faced by teachers implementing the LoLT within both L1 and L2 contexts was the reportedly inadequate prior learning of the LoLT within previous grades or before learners had enrolled at their current school. It was commonly indicated that the extent to which learners had mastered either of the LoLTs by the time they reached Grade 4, when they were required to begin learning in a L2, was inadequate. Learners’ inadequate prior learning of the LoLT was identified by respondents as somewhat resulting from learners’ insufficient exposure to pure language.
ii) Insufficient exposure to pure language
As found, the effective implementation of both LoLTs was hindered by learners’ insufficient exposure to pure languages and learners’ detrimental exposure to impure languages within learners’ environments outside of school, and interviewees indicated that the same applied inside of school. This insufficiency was indicated as somewhat relating to teachers’ inability to speak pure forms of language.
iii) Human resource deficiency
A human resource deficiency reportedly existed within schools that undermined the implementation of the both LoLTs. The most prevalent of these was teachers’ lack of proficiency using the LoLT.
Of the interviewees that mentioned this problem, 80% did so in relation to isiXhosa, 13% mentioned it in relation to English, and 13% in relation to both. These findings were reflected in the following opinion given by a principal: “There are few isiXhosa resources and a few people who can teach isiXhosa. But there are many people who can teach English”. Referring to the indicated shortage of teachers who are able to teach in either isiXhosa or English, the same principal added: “With a lack of such human resources the quality of teaching will never be at the good standard that we want it to be” (Principal A, principal at FP L1 LoLT school).
Respondents were of the opinion that teachers were not able to use either or both LoLTs because they had not been trained properly to do so.
iv) Teacher training inadequacy
As such, an inadequacy of teacher training, resulting in a lack of proficiency in teachers using the LoLT, was indicated to be a problem that stood in the way of implementing both English and isiXhosa LoLTs.
v) Lack of material resources
A lack of material resources was indicated as a factor that undermined the implementation of both isiXhosa and English LoLTs, but this was particularly the case for isiXhosa. isiXhosa teaching and learning material resources were reportedly very limited in their availability and often of a poor quality.
Having discussed the nature of LoLT implementation, as well as the factors surrounding the likelihood of implementation being successful, the following section outlines the reported effects on learners of the manners in which LoLT was implemented within the schools visited.
While the research found both positive and negative reported implementation effects of both LoLTs on learners, these effects were found to be significantly more negative than positive.
4.2.3.1 Positive effects of LoLT on learners
LoLT implementation reportedly affected learners positively in in the following ways.
a) Learner performance
36% of interviewees indicated good learner performance as being a positive effect of L1 LoLT use. Only one teacher was of the opinion that English LoLT use was conducive to learner performance. Similarly, simply put:
I have been teaching for 10 years, so I have seen the difference. If the child is learning in their mother tongue the child does better (Teacher G, Grade 3 teacher at FP L2 LoLT school).
b) Inclusivity in heterogeneous sociolinguistic contexts
43% of interviewees indicated that the use of English as a LoLT was positive in its capacity to be educationally inclusive within heterogeneous sociolinguistic contexts; that non-Xhosa learners were better able to access teaching and learning in English than they were able to access it in isiXhosa
The opinion that English LoLT use was conducive to the social integration of SA learners from diverse backgrounds was evident to a greater extent than was the case for isiXhosa.
c) Access to the globalised world and opportunities for social mobility
Participants indicated that English LoLT use was unavoidable if learners were to exercise social mobility later by:
· Undertaking to further their education at tertiary institutions where English was dominant; and
· Gaining access to professional employment typically available only to speakers of a dominant language, such as English.
While it could be argued that the dominance of English was detrimental to the disadvantaged SA learner for various reasons, the research took the view that an English LoLT holding social mobility for learners who learn English as a positive effect.
d) Linguistic and cultural preservation
The use of L1 LoLT was indicated as beneficial in that its use proliferated the language and associated culture beyond the classroom. In other words, using a language as a LoLT was conducive to maintaining the general prevalence of that language as well as its associated cultural knowledge within the society surrounding the school.
4.2.3.2 Negative effects on learners
Having outlined the positive effects that LoLT reportedly had on learners, the negative reported effects are now presented in order of the prevalence that there were mentioned during the research.
a) Barrier to teaching and learning
The greatest negative effect of LoLT on teaching and learning was indicated as being the LoLT constituting a barrier to learning as a result of teachers’ and learners’ inability to use it for teaching and learning, this being particularly the case within L2 LoLT contexts. Of those indicating that the LoLT was a barrier to learning in their schools, all did so referring to English, while a third did so referring to both isiXhosa and English.
b) Disadvantaged learners’ chance at education
That the LoLT had the effect of disadvantaging learners’ chances of attaining education was referred to in relation to isiXhosa L1 LoLT and English L2 LoLT equally, but for different reasons, as already discussed. In brief, respondents were of the opinion that neither LoLT model, in the way in which each was implemented, satisfactorily advanced learners’ educational opportunities.
c) LoLT policy transition
The LoLT transition from isiXhosa L1 LoLT to English L2 LoLT at the beginning of Grade 4, as is typically the case for the vast majority of SA disadvantaged schools where the L1 is used as a LoLT only in Grades 1 to 3, was viewed as highly problematic by all of the interviewees within L1 LoLT contexts. This was because, from the point of LoLT transition, learners were reportedly unable to learn adequately because they were not able to understand and apply English as a LoLT well enough to do so. It was indicated that this situation was aggravated by the simultaneous additional burden on learners to cope with a larger workload – as the number of subjects they were taught expanded from 4 to 7. The effect on learners of this LoLT transition was indicated as a decline in performance evident throughout subsequent grades.
At the point of LoLT transition, learners’ inadequate ability to express themselves using English was also indicated as problematic for various reasons that negatively impacted on learning. These included learners’ inadequate ability to use the LoLT in order to participate during teaching and learning activities; demonstrate learning during assessments; as well as to draw upon and express prior knowledge.
d) Inadequate mastery of L1 and/or L2
71% of all interviewees indicated that learners’ inadequate mastery of isiXhosa, English, or both languages was an effect of LoLT policy implementation in their schools. While 20% of such interviewees were from L2 LoLT contexts, 80% were from L1 LoLT contexts. Of those from L1 LoLT contexts:
· 80% indicated the LoLT model at their school resulted in learners’ inadequate mastery of English;
· 60% indicated the LoLT model at their school resulted in learners’ inadequate mastery of isiXhosa; and
· 40% indicated the LoLT model at their school resulted in learners’ inadequate mastery of both languages.
Participants typically indicated that learners’ inadequate mastery of either English or isiXhosa, or both, was caused by the poor quality of the manner in which the languages were taught and that this resulted in learners being ill-equipped to learn using the languages, negatively affecting learner performance and drop-out rates.
e) Learner drop out
As a consequence of learners’ inadequate mastery of the LoLT and the poor learner performance that this results in, learner drop out was indicated as resulting from the manner in which the LoLT policy was typically implemented within disadvantaged schools.
f) Linguistic, cultural and identity loss
Xhosa learners’ linguistic, cultural and identity loss was indicated as being caused by L2 LoLT use in the following ways:
1. The exclusive use of English LoLT pushed the isiXhosa language towards extinction.
2. The implementation of English LoLT during the early years of schooling limited the extent to which Xhosa learners are able to master their mother tongue.
3. By decreasing the prevalence of learners’ usage of isiXhosa, English LoLT use eroded the extent to which learners held Xhosa heritage as well as undermined the proliferation of Xhosa culture.
4. English LoLT implementation weakened the cultural identity of Xhosa learners.
g) Socio-economic marginalisation
Interviewees indicated that the prevailing LoLT policy in SA schools had the effect of socio-economically marginalising the most disadvantaged of the country’s learners.
This sentiment was alluded to in the points below, taken from a L1 LoLT principal’s discussion of how he saw LoLT affecting learners’ futures. His view was that for Xhosa learners in schools like his:
· The chances that they would leave the schooling system having not mastered both English and isiXhosa were strong;
· Mastering or not mastering English (as influenced by prevailing LoLT policy) meant the difference between learners obtaining a language that they would be able to use to mobilise themselves socially via further education and employment or not; and therefore
· That the level of learners’ proficiency in English and isiXhosa, albeit to a greater extent for the former, would ultimately determine the extent to which they were socially and economically isolated or not.
h) Deprivation of choice
Interviewees indicated that LoLT policy in SA did not allow most of the country’s learners the constitutional choice of what language they used for education and that in this way these learners were forced to learn in a language that was not their L1. Interviewees also expressed that for most black SA teachers, teaching in a language that was not their L1 was not a matter of choice, either.
Part A provided an indication of the learner performance associated with each respective LoLT model tempered by the finding that LoLT designations should not be assumed to accurately represent the LoLT implementation that they designate. It was found that those who learnt in their L1 performed better than learners who learnt in a L2. Furthermore, findings illustrated that learner performance benefited from additional exposure to the language of teaching, learning and assessment. This confirms what Taylor & Coetzee (2013, p. 15) found with regard to English.
Part B provided an indication of the nature of LoLT formulation and implementation within the schools researched. It was found that LoLT policy formulation was irregular in terms of the extent to which legislation was adhered to and theory consulted during the process. It was indicated that LoLT policy was only somewhat implemented, particularly in L2 LoLT contexts, and that non-implementation predominantly took the form of mixed language use. It was also indicated that the quality of teaching and learning typically found within disadvantaged L2 LoLT classrooms was inferior to that typically found within similar L1 LoLT classrooms. Furthermore, various challenges to LoLT implementation that reportedly existed within society, within the home and within the school were surveyed. These challenges were particularly stacked against isiXhosa LoLT implementation, but negatively affected that of English LoLT to a substantial extent, too. Part B concluded by listing the reported effects of prevailing LoLT policy implementation on learners within the schools that were researched.
LoLT policy recommendations that are suitable to contexts similar to that within which the research was conducted will be made in the next chapter, based on the findings.