Introduction

South Africa conjures up the embodiment of a linguistically and culturally diverse and rich country. In fact, the South African Constitution recognizes 11 official languages and consequently nine major African languages are on an equal footing with English and Afrikaans (since 2023 sign language became the 12th language). Politicians hence moved from the view of language as a problem to the conceptualisation of “language as a right” (Heugh, 2016: 249), although the supremacy of English in all official domains puts into doubt the move to a rights approach.

In fact, authors such as de Swaan (2023) and Ndimande-Hlongwa and Ndebele (2017: 68) are critical about the actual impact of such “paper policy” and claim that “the official language policy of South Africa, as embodied in the Constitution” (de Swann, 2023: 4), could be deemed a failure. This belief is held on the grounds that, although the original objective of promoting all indigenous languages equally was worth praising and undoubtedly well intentioned, reality indicates that indigenous languages have more often than not been overlooked. De Swaan (2023: 5) puts it bluntly: “Where a multiplicity of languages in all spheres of social life was the ostensible objective, only English profited and is on its way now to English only, at least in the sphere of national politics, administration, big business, and possibly, higher education.” The aim of this paper is thus twofold. First, to analyse students’ views on whether multilingualism at tertiary level is upstaged by the hegemony of English and “on its way now to English only”, as manifested by de Swaan. And second, to focus on the challenges that such predominance entails with regards to the language practices (i.e. translanguaging and classroom interactions) found in lectures. This will pave the way to unearth students’ language ideologies and their views on multilingual literacy practices.

Multilingualism as language policy in South African Universities

For decades now multilingualism has become a mantra at tertiary level. In this multilingual friendly environment, South Africa is characterized by its inherent linguistic complexity which has led some authors to label it as a linguistic laboratory (Rudwick, 2021). In fact, studies have confirmed that multilingualism, defined by the 2020 Language Policy Framework for Higher Education Institutions as "The effective use of multiple languages either by an individual or by a community" (p. 9), is widespread among South African university students irrespective of their L1 (Berghoff, 2021).

However, throughout history the linguistic situation has varied considerably and this has exerted a profound influence on language policies over time. According to Du Plessis (2006: 96), “Bilingual higher education in South Africa has its roots in the establishment of the Union of South Africa in 1910”, when the objective was to produce English and Dutch/Afrikaans bilingual citizens, although teaching at university was still primarily in English. From 1918 bilingual universities “slowly evolved into monolingual Afrikaans-speaking universities” (p. 97) and the use of Afrikaans as medium of instruction unfolded after the mid-1930s alongside the strengthening of the Afrikaans movement. “By 1948 there were four Afrikaans-medium universities in South Africa and four English-speaking universities. UNISA (a non-residential institution) remained the only ‘technically’ bilingual university” (Du Plessis, 2006: 97). Until 1994 Afrikaans and English continued to function as the languages of higher education with no space for the different indigenous languages spoken by the majority of the population. This situation changed with the passing of the Language Policy in Higher Education (Ministry of Education, 2002), which urged universities to strive to develop and use the nine indigenous official languages as academic languages besides Afrikaans and English.

In 2020 the Language Policy Framework for Higher Education Institutions (Deparment of Higher Education & Training, 2020) was published as a review of the 2002 Language Policy for Higher Education. The revision was needed because the latter’s development for transformation and social inclusion had floundered due to lack of resources, monitoring and systematic implementation. As a result, the supremacy of English had been reinforced, whereas indigenous languages had been upstaged from South African universities’ language policies (Cele, 2021). One of the main objectives of the Framework is to help to develop and strengthen the use of indigenous languages in scholarship, teaching, learning and communication at university level, while “enhancing the status and roles of previously marginalised South African languages to foster institutional inclusivity and social cohesion” (p. 12). The controversy of the 2020 Language Policy Framework was brought about by the exclusion of Afrikaans, Khoi, San and Nama as indigenous languages, as only those languages that have their heritage roots in Africa and belong to the Southern Bantu language family were regarded as indigenous languages. Despite the recurrent use of the term multilingualism in university language policies, the current hegemonic role of English as a lingua franca challenges the very essence of initiatives aimed at bolstering the use of different languages in graduate and postgraduate courses. Although the main objective remains to grant students from marginalized groups (i.e. indigenous home language speakers) access to learning on an equal footing, epistemological access may be refrained by the preponderance of English (Antia & Dyers, 2016). Since the predominant language ideology may impinge on students’ acceptance of the use of English as medium of instruction, this article aims at giving them the opportunity to express their opinions through focus groups so that their (more socially oriented) language ideologies and (more individually oriented) attitudes are unveiled.

Previous research (Mgqwashu, 2014) reports that the use of African languages as means of instruction (MoI) has only been paid lip service, as such experiences are conspicuous by its absence. Mgqwashu observed that one of the few pilot studies that used Zulu as MoI in a Bachelor of Education module allowed students to engage with the academic registers needed to succeed in formal higher education. Mgqwashu concludes that the use of African languages as MoI increases not only the opportunities for epistemological access, but it also brings indigenous ways of being into the university system while fostering social justice. Importantly, studies (Lasagabaster, 2004; Lasagabaster and van der Walt, 2024; Heugh, 2016; Rudwick, 2021) have also shown that students’ L1 and the degree they are enrolled in exert a significant influence on their attitudes towards languages and particular language policies, which is why we will pay special heed to these two variables (L1 and degree) in our study. In this vein, it has been found that Education undergraduates are more likely to support multilingualism than students enrolled in other degrees (Lasagabaster and van der Walt, 2024).

Whereas previous research (Banda & Peck, 2016; Bangeni & Kapp, 2007; Berghoff, 2021) has attested the uncontested acceptance of English as the language of academia, this article aims to go a step further by examining how the use of English as main MoI affects the development of multilingualism, the use of other languages as MoI, the implementation of multilingual practices in university lectures, as well as students’ stance towards a particular language policy (explained in the next section).

The context of the study: Stellenbosch University

Initially established as Victoria College in 1865 in what is nowadays the Western Cape province, Stellenbosch University (SU) changed to its current name in 1918. According to its language policy (2021), SU is a public higher education institution that aims to foster equitable access for all students by promoting individual, institutional and societal multilingualism, while it opens up opportunities to develop multilingual academic literacies. With this multilingual mindset in mind, SU commits to using the three official languages of the Western Cape Province, namely Afrikaans, English and Xhosa as a way to empower diverse communities. As for demographics, the Western Cape province is home to the largest proportion of L1 English (22% of the province’s population) and L1 Afrikaans (41.2%) speakers in the country, whereas Xhosa is the L1 for 31.4% of the province’s population (Department Statistics South Africa, 2022).

In SU’s language policy each language is described in the following terms: Afrikaans is seen as an international language that opens doors for academic partnerships with Dutch and Belgian universities, and the institution shows strong commitment to keep it as medium of instruction; English is valued for being an academic, professional and international asset; and Xhosa is included as a language spoken by one of the largest communities in South Africa and its promotion as an additional academic language is recommended and sought (Stellenbosch University, 2021). The three languages are thus regarded as resources for successful construction of knowledge, but their use should be based on what is practicable in each degree, depending on a panoply of factors such as the number of students, the academic language proficiency of students, the availability of linguistically proficient staff members or other resources (e.g. translation or interpreting services).

SU’s language policy strives to promote inclusivity and the appreciation of diversity, which is why, although Afrikaans and English are the primary MoI, translanguaging in multiple languages is boosted to support learning. Similarly, the language policy establishes that all official institutional communication has to be conveyed in the three languages and that they will be used judiciously at official events such as graduations. The three are also languages of external communication, but the caveat is entered that, in the case of Xhosa, it will be used when reasonably practicable. It is important to note that SU also commits to the development of other South African official languages, apart from the aforementioned three.

Research questions

With SU’s language policy in mind, in this paper the following research questions were put forward:

  • RQ1. What is the students’ stance towards SU’s multilingual language policy?

  • RQ2. Are students in favour of using Xhosa as an additional academic language?

  • RQ3. What is students’ attitude towards translanguaging?

  • RQ4. Do students’ observe any problems in the implementation of EMI?

  • RQ5. In their view, does English proficiency affect classroom interaction in EMI courses?

The study

The participants

Eleven focus groups were carried out at SU with a total of 30 participants, who were randomly invited to participate in the study while they were walking on campus. The sample encompassed students from 11 different degrees and diverse L1s (Afrikaans, English, Xhosa, and other indigenous languages), the majority of them female (70%), and their mean age being 21.1. The detailed description of the participants is provided in Table 1.

Table 1 Focus group participants
Full size table

It is worth noting that those who had Afrikaans and English (or both languages) as L1 represented almost three quarters of the participants (73.3%), which is a reflection of the fact that these two languages are clearly predominant in the student body. Xhosa L1 speakers amounted to 13.3% of the sample, but this can be regarded as proportional, since the June 2023 data show that Xhosa home language speakers represent 6.4% of the student body at SU. Those whose mother tongue was a different language than the three just mentioned amounted to 13.3% of the participants, which evinces the linguistic and cultural diversity characteristic of SU.

The procedure

The study was approved by the Ethics Committee at SU (Project ID: 29122) and informed consent was obtained from all participants. All focus groups were carried out in English by the author, which is the main language of instruction of the university under scrutiny and in which all the participants were proficient. The questions raised by the interviewer to foster discussion and debate revolved around four main issues: the SU’s multilingual language policy, the use of the three official languages at university, the implementation of translanguaging practices, and the impact of English as the main medium of instruction. It is worth pointing out that in the focus groups the terms codeswitching and translanguaging were used interchangeably because some of the participants were not familiar with the term translanguaging, although the author is well aware of the differences between both concepts (Lasagabaster and García, 2014).

Overall, the focus groups had a duration of 432 min, and an average duration of 39 min. They were audio-recorded and then completely transcribed with the aid of the Sonix online software (https://sonix.ai/). The transcriptions were double checked by the author and a research assistant who is a postdoctoral researcher highly experienced in this kind of qualitative research study. All focus groups were then content analysed (Dörnyei, 2007) using NVivo 14 and the coding system was discussed by both researchers until an agreement was reached.

Results

RQ1. What is the students’ stance towards SU’s multilingual language policy?

When inquiring into students’ stance towards the university’s multilingual language policy, the following main themes emerged: (not enough) appreciation of language diversity on campus, (not enough) promotion of inclusivity of the three official languages, English as a universal language, wish to have a larger learning offer in English or Afrikaans or Xhosa, and (not) being in favour of adding other indigenous languages apart from the three official ones (see Table 2 for the number of instances under each theme).

Table 2 Total coding references for each theme (absolute numbers) (RQ1)
Full size table

It is interesting to notice how the situation at SU was described. As regards life on campus, interviewees tended to agree that language diversity is promoted (cf. FG7_1, quote below) and that SU is trying its best to promote the inclusion of English, Afrikaans and Xhosa. While they did recognise such effort of inclusion of the three languages, the majority affirmed that it has not been reached yet. Education students in particular showed a high degree of awareness of these themes, and recognised that dealing with language issues in highly diverse multilingual settings is not an easy task.

We won’t ever be able to fully understand one another’s languages as much as we understand our own language. And I think there’s a reason why everybody has pride in their own language, because it’s your language, and that is a good thing and a beautiful thing in and of itself. So language issues will always be there. However, we should always try our best to work towards including people understanding one another’s languages as best as we can and respecting it as best as we can, and trying to work together to get that one job done. [FG1_2]

While life on campus accommodates for the presence of different languages, education provision was regarded as “a complete separate section”, and the little inclusion of languages other than English was underscored.

At the end of the day, the education has to be in a certain way that is useful to everyone else. And more people speak English in, for a specific section or degree than in another language. Uhm … but in terms of, like, general life and culture around the university, it’s still like, uhm… people are promoted to speak in their first language, the languages that they’re comfortable in. But I think the education aspect is a complete like separate section. [FG7_1]

Frequently, students stated that the university tries hard to promote the inclusion of the three official languages. From what they said, however, it was understood that Afrikaans was the only language that was promoted and used other than English, while Xhosa, more often than not, was not even mentioned and inadvertently overlooked. Moreover, the use of Afrikaans seemed to be generally aimed at facilitating the transition from Afrikaans to English only, which appears as the predominant language.

Based on my personal experiences on campus, all of my modules have consistently been taught in English. But even as, as FG1_4 mentioned, even in an Afrikaans module, where one of South Africa’s official languages is the main focus, the lecturer does give the first additional language-speaking students who still have difficulty in understanding by providing translations vocally during the lessons, should it be necessary. In other words, it has already been implemented that our modules cater to our languages, our languages, that is, either English or Afrikaans. [FG1_1]

It is interesting to observe the differences that were found in opinions expressed by focus group participants based on their L1 (Table 3), and on the degree programme attended (Table 4). As for degrees, nine participants were enrolled in Education, while 21 were attending other programmes. This differentiation (Education vs. Other) is motivated by our interest in understanding what opinions were expressed by prospective teachers, and whether these were distinct from those of students from other degrees (as observed in Lasagabaster and van der Walt, 2024).

Table 3 What is the students’ stance towards SU’s multilingual language policy? (RQ1): Coding results based on L1
Full size table
Table 4 What is the students’ stance towards SU’s multilingual language policy? (RQ1): Coding results based on degree programme (Education vs. Other programmes)
Full size table

Table 3 reveals that the Afrikaans-L1 group showed the highest percentage of coding of the theme "appreciation of diversity" on campus (55.1%). On the flip side, the English-L1 speakers considered that appreciation of language diversity is currently not enough (39.85%). Interestingly, the latter group is the one that most often stated that the SU policy promotes—or tries its best to promote—the inclusiveness of the three official languages (56.84%), but, at the same time, that the commitment is not yet sufficient (49.08%), since the third language, Xhosa, has little presence in courses.

With regard to the desire for a greater presence of the three official languages or other indigenous languages, the largest percentages were found, by far, in the Afrikaans-L1 group. Interestingly, all the instances of the desired greater presence of Xhosa were found in the Afrikaans-L1 group (100%), who expressed a favourable attitude towards this language, whereas in the Xhosa-L1 group no one spoke in favour of this possibility (aligning with the results of RQ2 below). A possible interpretation of this finding could be that Afrikaans L1 students are aware that they cannot request for their own language to be incorporated in the university by closing down the pathways for other languages, or Afrikaans L1 students are aware that their claim for Afrikaans to be used will be stronger if it is paired with the claims of other official languages.

When analysing the impact of the degree programme (Table 4), Education students were in favour of a greater presence of not only Afrikaans (86.09%) and Xhosa (89.08%) as languages of instruction, but also of other indigenous languages (98.94%). In contrast, those enrolled in other degree programmes emphasised above all the role of English as a universal language (88.2%) and expressed their desire to increase its presence at university (77.16%), an attitude vividly illustrated by the quote by FG7_1 (engineering student) above. Education students’ more positive stance towards indigenous languages could be due to the impact of the curriculum, in which they have the option to take two languages (chosen from Afrikaans, English and Xhosa) during their four years of study; in addition, a third language has to be chosen in the third year of study. Moreover, concepts such as multilingualism, translanguaging and interculturality pervade the curriculum (and this is not the case in many other degrees such as engineering, for instance), which may explain the differences found when education students were compared to students enrolled in other degrees.

RQ2. Are students in favour of using Xhosa as an additional academic language?

When analysing RQ2, the following main themes emerged: Xhosa as a difficult language, multiple meanings of its vocabulary, Xhosa as a problem, Xhosa as a compulsory subject, use of Xhosa just as a bridge to make the transition to English easier, and (not) being in favour of having Xhosa as an additional academic language. Table 5 compiles the number of instances coded under each theme.

Table 5 Total coding references for each theme (absolute numbers) (RQ2)
Full size table

Intriguing results were observed when the participants’ L1 (Table 6) and degree programme (Table 7) were looked into. Afrikaans-L1 and English-L1 participants harboured a much more favourable stance towards Xhosa (41.42% and 42.55%, respectively) than Xhosa-L1 participants themselves, none of whom stated they were in favour of having it as an additional academic language. The highest percentage of coding in the “not in favour” theme (48.44%) was registered among Xhosa speakers, and the main reason put forth was that Xhosa words tend to have multiple meanings, and such circumstance hinders understanding, especially when it comes to high-level academic subjects. They also underscored the difficulty of coining technical terms in Xhosa. The two exchanges below from FG3 and FG6 illustrate this point.

No, because Xhosa is so complicated… [FG3_1] I’m not in favour because I still stand on my point that it’s unfair for other students. Zulu and Xhosa is similar, but the words don’t mean the same thing. So imagine if there was a course, and I understand that word in the way that I was taught, and the question is not really asking that. And we have 12 languages in South Africa, out of 12 they cannot take one. It’s either they take all of them or they just leave it [FG3_2]. It’s so complicated, they’re deep, you know. One word could mean several things. It’s… it's not easy to translate them into science-based terms, for example, you know. So I wouldn’t be… I’m not in favour of having Xhosa as an official additional language of the university. English should be the only language that is used for the university. [FG3_3].

Not really. Like Math in Xhosa would be a different thing. I feel like it’s harder in Xhosa. Anything, everything would be harder in Xhosa, you know, because it’s… it’s a complex language [FG6_2]. It’s very, very hard [FG6_1]. And as I said, you can’t directly translate anything. So they’ll have to make up words now. Make up new words so that we can try to get an understanding [FG6_2]. But for example, at school, when you were kids, did you learn through Xhosa or English? [Interviewer]. Xhosa [both]. And was it difficult? [Interviewer]. No [both]. Not really. It was easy because we were only doing one plus one, two plus two. [FG6_2]

Table 6 Are students in favour of using Xhosa as an additional academic language? (RQ2): Coding results based on L1
Full size table
Table 7 Are students in favour of using Xhosa as an additional academic language? (RQ2): Coding results based on degree programme (Education vs. Other programmes)
Full size table

In contrast to the rather negative attitude towards Xhosa expressed by Xhosa-L1 participants, Afrikaans-L1 interviewees seemed to hold a much more favourable stance towards the language. Such positive stance is nicely illustrated in the following passages below, where a greater presence of Xhosa was supported (also based on historical reasons).

I feel, like my first year, uhm… I had Xhosa as a language, uhm… and I felt, like, it’s very important to have that implemented, especially because of our past, uhm… in South Africa, apartheid. And I do feel like to be inclusive of everyone, we should have at least a module in Xhosa. [FG2_2] I would like to include that. And also because I feel, because Xhosa is being overlooked, uhm… it would also be, I think, appreciative for other people speaking Xhosa that we also try to understand them in their language. So I would like to learn that as well. And I feel if more people could speak all three, uhm… then everyone would feel included. [FG2_1]

A favourable stance towards Xhosa was expressed by future teachers, that is, students enrolled in Education (Table 7). The following quote by an Afrikaans-L1 future teacher exhibits their positive attitude towards the language.

It’s… it’s a positive thing for us because we don’t know, like, what language we are going to teach in the future. So having an additional language will be a positive thing for us as future teachers [FG9_3].

In contrast, 83.33% of students attending Other programmes declared they did not support the idea of having Xhosa as an additional academic language. Figures for the theme “In favour (or Not against) of having Xhosa as an additional academic language”, which seem contradictory to what has just been stated, are due to the imbalance in the number of participants in the two groups: while there were 21 participants attending Other degree programmes (some of whom expressed a favourable stance), only 9 were those attending Education. This imbalance is the reason why the Education percentage (34.99%) is lower than that of the Other programmes (65.01%).

An interesting theme that emerged only among Other programmes students (100%) was their willingness to accept Xhosa, but not as a mainstream offer for everyone, but rather as an opportunity Xhosa-L1 speakers deserve in order to bridge their gaps and facilitate their transition to English. Noteworthy is also the parallelism established between what could be carried out with Xhosa and what is actually done with Afrikaans, as nicely manifested in the following quote.

Maybe just bring it, if, I think firstly it would… it would be helpful to determine how many students have the need for it. Uhm… because there could be students, like in our case, that would be very comfortable with maybe transitioning directly into English, which they now have to do, I think, if they, if their home language is Xhosa. But if the case is not so, I think it would be very helpful to just bring the… it on par with the transitioning of, the same level that Afrikaans is transitioned into English, maybe. Maybe bringing Xhosa just up to that level, if there is a need for it. [FG11_3, Agricultural economics]

RQ3. What is students’ attitude towards translanguaging?

In the case of RQ3, the following eleven themes emerged: translanguaging as helpful, being in favour of translanguaging, translanguaging as not helpful (or not being in favour), need for international students to be flexible and accept it, need for lecturers to stick to English when international students are present, translanguaging as a means to learn content, negative aspects of translanguaging, classroom materials, translanguaging on PowerPoint (PPT) slides, translanguaging in tasks or exams for evaluation, and finally translation better than translanguaging (Table 8 displays the number of instances under each theme).

Table 8 Total coding references for each theme (absolute numbers) (RQ3)
Full size table

Once again the participants’ L1 (Table 9) leads to a remarkable finding, as English-L1 students tended to express the most positive stance towards translanguaging overall. Not only were they most favourable (46.29%), but they also recognised its effectiveness in supporting students to learn content (37.78%) and in tasks and exams for evaluation (42.97%). There seemed to be a general appreciation of the use of translanguaging by English-L1 participants, as nicely illustrated in this quote by an Engineering student, who perceived it as a means also to make the lesson more “lively” and “interesting”.

It does make the lecture feel at least a bit more lively and more interesting. […] personally, I just I enjoy other languages, so it’s just nice to, and generally when, because it’s not like a requirement, switching isn’t a requirement. It’s lecturers who do it are generally just more, uhm… more like outgoing, more friendly. [FG7_2]

Table 9 What is students’ attitude towards codeswitching/translanguaging? (RQ3): Coding results based on L1
Full size table

Contrarily, Xhosa-L1 speakers (62.73%) and speakers of other indigenous languages (37.27%) were consistently less in favour of translanguaging. The latter were also those who most frequently mentioned negative aspects of translanguaging (43.06%) (e.g., feeling somehow uncomfortable or frustrated for not understanding, or wasting time that should be dedicated to content), and the only ones to speak in favour of translation (100%), as shown in the following quote:

I’d rather prefer that maybe to say, if, what FG5_4 pointed out, was during Covid that they had an Afrikaans translator and like an English option and all that, that would probably potentially be more effective because then at least content and like the course actually moves along quite fast. And there’s more time for the lecturer to actually explain things rather than shifting different languages and by shifting languages. [FG5_3]

The contrast between English L1 students and speakers of indigenous languages is well worth noting. Xhosa L1 speakers’ attitudes seem to be shaped by the fact that they have worked their entire lives to be proficient in English in order to accommodate others, and importantly, to the detriment of their own language as a side effect. When Xhosa L1 students have undergone their entire schooling transitioning to English as best they could, and arrive at university only to be told that their institution is now opening the door for multilingualism, they do not see the implementation of translanguaging as a progressive move, which seems to spark their frustration that the linguistic goalpost has shifted yet again. From a language policy perspective, this is an issue to bear in mind, as the disconnection between pre-university education and higher education needs to be addressed.

When focusing on the degree (Table 10), the vast majority of instances of the theme “translanguaging as a means to learn content” (54.12%) was detected among Education students. In response to the question of being in favour or not of translanguaging, an interesting reflection came from a prospective teacher. After expressing herself in favour of it, she provided a deep reflection on its use at SU, and observed that, in her view, it is not always favourably received. She argued that this lack of openness to inclusivity—promoted by translanguaging—stands at the roots of the problem of achieving multilingualism at SU.

I think that’s the root of the problem with multilingualism in Stellenbosch at the moment. Because everybody is kind of selfish about their language, selfish about their culture. So they want to be able to express themselves. But then by being selfish, they are excluding everyone else as well. So like that, like clashes in classes like, and that also, is also a negative thing because it makes people not pay attention, it makes people not come to class because they’re like, this is just "I don’t understand", "It’s not working for me" "I can’t concentrate". But I think the goal of multilingualism is there. Inclusivity is there, but there’s always going to be that outlier of people who just aren’t accommodated for. [FG1_3]

Table 10 What is students’ attitude towards codeswitching/translanguaging? (RQ3): Coding results based on degree programme (Education vs. Other programmes)
Full size table

RQ4. Do students’ observe any problems in the implementation of EMI?

As for the implementation of EMI at SU, the following five themes emerged: English language-related hindrances, being in favour of English language support and, more specifically, being in favour of English language courses, not being in favour of English language support, and possible solutions to solve English language-related problems. Table 11 shows the number of instances that were coded under each theme.

Table 11 Total coding references for each theme (absolute numbers) (RQ4)
Full size table

Looking at coding results based on participants’ L1 (Table 12), a rather clear pattern emerges. Afrikaans-L1 speakers were the most aware of the existence of problems with English (65.4%), and both English-L1 speakers and Afrikaans-L1speakers were in favour of language support both in the classroom, during lessons (as provided by lecturers themselves), and in dedicated support courses. Moreover, it was from Afrikaans-L1 students that most of the proposed solutions to English-related problems came from. A solution that was mentioned was the use of translanguaging by the lecturer, as delineated in the following quote (Table 13).

I feel, for example, if it is not your first language and you have difficulty understanding certain words, that’s a bit, big words, the lecturer can maybe say… Oh, well, this is a big word that maybe translate it to the other language and explain what it means. [FG1_2, Afrikaans L1]

Table 12 Do students observe any problem/hindrance in the implementation of EMI? (RQ4): Coding results based on L1
Full size table
Table 13 Do students observe any problem/hindrance in the implementation of EMI? (RQ4)—Coding results based on degree programme (Education vs. Other programmes)
Full size table

On the other hand, the participants who were the least favourable were Xhosa-L1 speakers, who happened to be less prone to this type of support (54.64%). In their view, translanguaging is generally performed with Afrikaans in EMI classes to broaden the understanding of Afrikaans-L1 students. This represents a “benefit” that is offered to them and which, on the other hand, disadvantages natives of other languages, such as Xhosa or Zulu speakers.

So address your lecturers and make sure that being an Afrikaans student isn’t an advantage, you know? Because it’s not like you choose it. Nobody chooses to be black and speak Xhosa, you know? You’re born like that. So speak to your lecturers and make sure that the whole advantage, the whole benefits that are given to Afrikaans speaking students is staffed entirely. [FG3_3]

According to Xhosa-L1 students, just as they have striven throughout their pre-university years to learn English well in preparation for university (even at the expense of relinquishing their L1 as an academic language), so should Afrikaans students. This is bluntly put by FG3_3 below, summarising and adding details to what her fellow participants (FG3_1 and FG3_2) mentioned before.

Black kids were taught to be fluent in English, to try and bridge the gap with white people. That’s the reason why, according to me, that’s the reason why everything we did back in school was in English. You were always pushed to be fluent, know English, know your vocabulary, know how to speak. When all this time we could have been communicating in our home language. Clearly, you find black people now that can’t pronounce their own names, because throughout their whole lives they have been pushed to speak in English, know how to speak, know how to speak. And then you come into this white place, full of white people that say, ‘Listen, all the work you’ve done your entire life is not worth it. Because guess what? I don’t even want to speak English. I want to speak Afrikaans. And because you don’t want to speak Afrikaans, I’m not going to talk to you!’ It’s unfair! […] But you look at how Afrikaans people... at home you speak Afrikaans, you go to school you speak Afrikaans, you go to high school you speak Afrikaans, your entire life you speak Afrikaans! You are at no point obliged to try and meet another person halfway. And then you come here and you’re still carrying that whole mentality, when opposite of you is a black person who has been pushed their entire life to try and meet you! [FG3_3]

Student FG3_3 complains that Afrikaans L1 students have had the opportunity to experience their language being used in the school and university as an academic language, whereas isiXhosa L1 students have not had the same degree of use in schooling.

As happened in RQ4, the Education figures suffer from the imbalance in the number of participants with respect to the Other programmes group. Nonetheless, future teachers fully acknowledged the existence of English language-related problems (41.03%), and seemed to be very aware that such hindrance negatively impacts content learning, as this quotes nicely illustrates.

So I think one should really consider both things [i.e., content and language]. But I don’t think it should be something we should also just throw away, uhm… the language issue, just because there’s so much content to do. Because if you do throw the language issue away, it affects the content learning as a whole. So what’s the use of learning content when you can’t learn it to the best of your ability and bring to the table the best of yourself because of the language learning issue that is already there? [FG1_2]

None of the Education students were against language support; 100% of the coding of this theme was found in the Other programmes participants, as exemplified in this quote by a science student:

Now if the teacher focuses on the language, the content now, you see, there’s… there’s no learning of the content. So it’s just a waste of time. I feel like if a student can’t understand, they should look up the words or ask for more further, more like explanation from the teacher aside, or do more research themselves, you know, because the university anyways, you have to. [FG6_1]

Prospective teachers, instead, welcomed the idea of underpinning language through possible English support courses, provided that such courses are hands-on, interactive and student-centred.

So that depends on how interactive and how successful these English language courses are going to be. Right? If we look at courses such as TEFL, which is Teaching English as a Foreign Language, uhm… those courses are predominantly focused on English. But it’s not as, they not, it’s not hands-on. It’s not! So it depends on how interactive and how hands-on you’re going to make these short courses or courses for the learners. [FG4_2]

RQ5. In their view, does English proficiency affect classroom interaction in EMI courses?

English proficiency emerged as an important topic in students’ discussion, the following main themes being highlighted: formation of cliques based on language, questions in front of the class, student–student and teacher-student interaction, and feeling uncomfortable when having to speak English (see Table 14).

Table 14 Total coding references for each theme (absolute numbers) (RQ5)
Full size table

When examining the coding results according to the participants’ L1 (Table 15), it was evident that Xhosa-L1 (40.53%) and other indigenous languages (31.33%) speakers were the ones that more often reported the formation of cliques based on the L1. In their opinion the formation of groups at university was based on knowledge of Afrikaans or lack thereof. That is, Afrikaans-L1 students tend to create a separate group, where Afrikaans is spoken, whereas English-L1 and other-L1 students form a different group, where English becomes the main language of interaction. Interviewees said that this happens not only in informal conversations but also in class, because groups are split based on language knowledge. As a result, English represents an inclusive language, since everybody can fall back on it, as everybody is expected to have some command. Instead, Afrikaans, whose presence according to Xhosa-L1 and other indigenous languages speakers is still very strong at SU, is basically just for Afrikaans students, and it is therefore exclusive.

I think that a big split in classes, most of our classes, where there is like a predominantly English-taught class or anything that will always be split, like in terms of students between Afrikaans students and English-speaking students, right? So English-speaking students are not even just white students. Just any student that doesn’t speak Afrikaans will then like fall into that group. [FG10_2]

Table 15 In their view, does English proficiency affect classroom interaction in EMI courses? (RQ5): Coding results based on L1
Full size table

In coherence with what was seen in RQ4, Xhosa-L1 students were also those who registered the highest percentage in student–student interaction, which they tended to find difficult due to their lower English proficiency. In line with this, Afrikaans-L1 students recognised that insufficient proficiency in English creates difficulties in teacher-student interaction (53.33%) and is responsible for a feeling of discomfort in class (37.68%). The latter theme (i.e., feeling uncomfortable) was also frequently mentioned by English-L1 students (56.18%), who seemed to be well aware of the difficulties of their non-English-L1 peers with the main language of instruction, difficulties that could impede asking questions comfortably in class or that could lead to dropping out of studies, as portrayed in the following quotes. Curiously enough, feeling uncomfortable was not mentioned by Xhosa speakers because they focused more on interaction issues.

And generally students will talk to one another, say… Hey, don’t you want to ask this for me? If they’re not comfortable [FG7_2 ]. That, you have seen that? Asking a classmate to ask the question on their behalf? [Interviewer] [Both FG7_1 and FG7_2 agree, laughing] Okay, interesting. So you are asked, as native speakers, to ask questions [Interviewer]. [Both FG7_1 and FG7_2 agree, laughing]

Uhm, one problem that might occur is obviously, as you get further into the degree, people drop out, people leave. And I think maybe a large percentage of that could be non-native speakers who, the reason they’re dropping out is not because they can’t do the content, it’s more so they can’t do the language. So as we progress through the degree you are around people like that a lot less. You don’t have a lot of experience with those sorts of people. So you sort of get into this bubble where everyone can speak English and no one has a problem with English. [FG7_1]

Education students also placed special emphasis on the impact of English proficiency on student–student (35.54%) and teacher-student interaction (40.02%) (see Table 16), as manifested in the following quote.

Sometimes, when you, when you’re in a classroom and you’re having a discussion or you’re having a debate, uhm… it gets to the point where you want to interact, you want to answer, you know the answer, but because your… your proficiency within the language is not as good or up to standard as it, for example, maybe should be, now you feel that if you now speak, uhm… the lecturer is, at the end of the day, is going to have to translate what you said, if you speak in your mother tongue… So I feel that sometimes that makes students not want to interact in classrooms, because now you feel that, uhm, since my language is not, or the use of my language is not that good, I would rather sit back and let someone else give the answers, maybe. [FG4_1]

Table 16 In their view, does English proficiency affect classroom interaction in EMI courses? (RQ5): Coding results based on degree programme (Education vs. Other programmes)
Full size table

Discussion and conclusions

Our findings reveal that it is complex to strike a balance between the three official languages at SU. Participants in this study agree on the impossibility of supporting and giving presence to all South African languages, but still largely consider that the three official languages at SU are part of their identity and they should all play a role on campus. In fact, students attach symbolic and cultural value to languages other than English (Banda & Peck, 2016).

At SU English is perceived as a universal cultural capital that is highly valued, but the institution also supports scholarly content developed in Afrikaans and its continued use as a scientific language. English is felt as a neutral and universal language that is not identified with any specific group and this makes its hegemonic role as MoI easy to accept, whereas Afrikaans is identified with a very concrete group of speakers which seems to make its knowledge less palatable for Xhosa speakers. Some of our participants even bear a grudge against Afrikaans and still seem to associate it with the language of the apartheid (Heugh, 2016), which even leads them to have a reluctant and apprehensive attitude towards translanguaging. Paradoxically, Xhosa speakers look down on their own language as MoI, a negative view when it comes to the academic domain that needs to be addressed.

SU language policy aims to develop Xhosa as an indigenous formal academic language and to increment its presence in various disciplinary domains when this is practicable and pedagogically sound, and especially when its use is important for career purposes, such as Education. In fact, Education students are the most vocal when defending the need to implement a really multilingual language policy. This is a key issue, since South African universities do not qualify enough teachers to teach content subjects in specific languages (Mpofu, 2019).

However, SU has covertly conferred the status of primary MoI on English and this decision looks natural to students, to the extent that its key role remains indisputable (Banda & Peck, 2016), while Afrikaans performs a supportive role and Xhosa plays second fiddle or none at all, depending on the degree. Thus, these findings match with those obtained by Antia and van der Merwe (2019) in the University of the Western Cape (UWC), where the Xhosa students whose position was supposed to be enforced by a multilingual language policy ended up being excluded, while English was “enthroned” and Afrikaans “dislodged” in the process (p. 426), although it has to be acknowledged that UWC’s policy does actually elevate English over the other languages. What is more, Xhosa speakers in our study disavowed their own language (Antia, 2015) and expressed strong opinions –without space for any shades– about English being the most appropriate language in academia. The lack of concerted efforts to increase the presence of Xhosa as MoI has infused them with the negative belief that their language cannot be effectively used in the academic world (Mgqwashu, 2014). In the students’ protests in South African campuses between 2015 and 2016 in defence of a free decolonized education, Antia and van der Merwe (2019: 408) found it striking that “the students were largely mute about the role of indigenous African languages in higher education.” Our results seem to indicate that this is largely still the case and that “the way to English only” (de Swaan, 2023) is just being strengthened, although some resistance is found on the part of Afrikaans-L1 speakers and Education students.

The existence of a language hierarchy (Weber & Horner, 2012) is thus neatly established, since English is clearly the predominant language, followed by Afrikaans and, in a decreasing order of importance, by Xhosa. In Xhosa home language speakers’ defence, it needs to be noted that their language ideology may have been shaped by the fact that, in many instances, they would have had their L1 used as MoI for just three years in primary education (Antia & Dyers, 2016; van der Walt & Pfeiffer, 2021). As a result of this language policy, English is only consolidated as a strong identity factor in our interviewees’ multilingual identities at university level.

Our findings indicate that a paper policy (Ndimande-Hlongwa & Ndebele, 2017: 68) aimed at fostering multilingualism is actually not equalising the status and actual functions and use of university languages, which is why some measures should be implemented to try to reverse the situation. First, university students at SU need to become aware of the importance of developing their multilingual skills at least in the three official languages. Short courses could enable teaching staff and students to develop basic communication skills and discipline terminology in Afrikaans and, above all, Xhosa.

Second, Xhosa speakers would benefit from multilingual language awareness courses that would made them aware of the importance of developing their language for academic purposes. A greater number of courses in which their language were the MoI would also contribute to this aim (Mgqwashu, 2014). Other international experiences have shown that minority languages can successfully be used as MoI at university level (Lasagabaster, 2023), while African indigenous languages do not need “to be ‘developed’ or ‘enriched’ with an endemic academic vocabulary, they should just adopt the bastard, pseudo-classical terminology that almost all modern languages use” (de Swaan, 2023: 5). However, the responsibility cannot be placed only on students. In fact, the onus should also fall on higher education departments to create language and curriculum policies that are aligned to one another and to SU’s language policy (2021).

Third, our participants highlight that languages other than English are not often resorted to in their lectures. However, studies prove that the mere use of Xhosa and Afrikaans to deliver content helps to change students’ cognition (Antia & Dyers, 2016), which is why what is stated about translanguaging in SU’s language policy should be fostered by the actual implementation of multilingual literacy practices in lectures. Whereas the responsibility of this type of linguistic initiatives are constantly put on students’ and teachers’ shoulders, the university should take a more active role and foster its use as a bridge to help some students overcome their linguistic hindrances with English (as pointed out in discussions in FG1 and FG3 above).

It has to be acknowledged that one of the limitations of the study is that teaching staff was not included in the sample. Since teachers have been theorised as language policy arbiters and exert an impact on shaping language attitudes, future research should also aim at this university body, as this would help to complement our findings.