Despite the rich South African literature on the logistical problems associated with the Language in Education Policy (LiEP), this body of work typically tends to treat these problems merely as local implementation issues (Heugh 2013; Probyn 2024). This article builds upon previous research by going beyond mere critique and providing a deeper analysis of the ideological structure of MTBBE. In doing so, this research maps out five common cultural myths – including those of transition and monolingualism – that reveal the ways Eurocentric hegemony continues to be safeguarded by progressive language policies. Moreover, while many studies advocate small reforms in early-exit programmes, this article makes an important contribution by developing a coherent radical bilingualism framework.

Africa has the paradoxical and unenviable position of being the only region where colonial languages still dominate as languages of learning and teaching (LoLT) or languages of learning, teaching and assessment (LoLTA) long after political independence. Prah (2009) confirms this enduring anomaly by noting that, for most African countries, indigenous African languages serve the education system as LoLTs in the initial 3 years or 4 years of formal schooling, before a colonial language takes over. South Africa is not an exception. In South Africa, only English and Afrikaans mother tongue speakers have the unique privilege of enjoying Grade R to university mother tongue education. This disparity is associated with persistent inequalities in literacy outcomes and educational attainment. For Makalela (2018), this linguistic inequality emanates from apartheid South Africa’s long history of linguistic separation where ‘indigenous African languages in South Africa have always been taught in isolation from one another’ (p. 227). This evidently reflects in educational outcomes, with learners who are prematurely compelled to transition to English medium instruction experiencing compromised literacy development and low academic achievement (Koch et al. 2009; Sailors et al. 2010).

The MTBBE in South Africa is a cyclical resurfacing of an old language-in-education policy paradigm that has been traditionally associated with failed promises. Although sold in contemporary discourse as a revolutionary step forward, MTBBE risks resurfacing a tradition that goes back to colonial times.

The issue of language authorisation in the educational context in South Africa has historically been politically framed. For instance, the 19th and early 20th-century missionary schools adopted mother-tongue teaching of largely literacy and religious curricula. Teaching in indigenous African languages was meant to achieve full understanding of the religious content. With the colonial and later apartheid periods, language policy shifted from native languages to the dominant use of European languages for education. The change was occasioned by the sinister motive of undermining any indigenous knowledge formation (Makalela 2018).

Further to that, the apartheid dispensation made mother tongue teaching compulsory until Grade 6 within the framework of the segregated system of Bantu Education. Given the malicious motive of depriving Black African students of cognitive growth and academic and socio-economic opportunities, the language education policy was oppressive, rather than empowering. However, what is ironic is that the segregated Bantu Education system and the repugnant regimes of colonialism and apartheid had language education models that extended the use of the mother tongue longer than the democratic government’s language-in-education models.

The democratic dispensation post 1994 necessitated legislative intervention to address deeply ingrained language hierarchies. For instance, Section 29(2) of the South African Constitution entitled learners to education in their preferred official language(s) (Republic of South Africa 1996) subject to ‘reasonable practicability’, an escape clause that provided a convenient loophole for denying these rights to most students. The constitutional requirement for language equality was meant to be implemented through the revolutionary Language-in-Education Policy (LiEP) of 1997, which promoted the use of additive bilingualism (Department of Education 1997). Subsequent complementary policies included the White Paper 6 of 2001, Ministerial Report on Indigenous Languages of 2003 and Draft Policy on the Incremental Introduction of African Languages (IIAL) of 2013.

Unfortunately, an extremely contradictory trend emerges in the South African educational context, where the same languages which have been declared essential in conceptual development are abruptly discontinued after the initial levels based on the wrong assumption that conceptual development is only required at the early years of education.

In South Africa, MTBBE is not a novel idea, but rather a resurgence of an old language in education model characterised by both promise and neglect. While it has often been marketed and touted as a progressive innovation, MTBBE has manifested a pattern of revival rather than evolution, a revival that is almost cyclical, dating back to colonial, apartheid and post-apartheid eras.

The contemporary rise of MTBBE has been built on the foundation of localised pilot programmes and ongoing structural crises. From 2015 to 2017, pilot MTBBE programmes were run in the Eastern Cape and Limpopo provinces, extending mother tongue instruction past the usual Foundation Phase (Grades R-3) boundaries. Ngubane and Ngwenya (2025) systematically reviewed the pilot implementations and the dynamic translanguaging process in multilingual classrooms. It emerged that state-run programmes produced results that were overly inconsistent and too uneven to be instructive. The programmes were not sustained long enough to allow for longitudinal assessment of the impact of mother tongue instruction beyond Grade 3. This contrasts with Funda Wande’s research-based, early literacy education curriculum in isiXhosa, Sepedi and Afrikaans implemented from the Foundation Phase to Grade 3, whose success was proven through randomised controlled trials performed in the Eastern Cape, Western Cape and Limpopo, yielding positive effects on children’s reading development (Mason et al. 2024). The Department of Basic Education (DBE) took a decisive step between 2020 and 2023 to introduce home language teaching to Grade 6, culminating in the 2024 national announcement of the DBE rollout of the late-exit approach from Grades 4 to 7 through the Basic Education Laws Amendment (BELA) Act. This arguably represents the boldest step yet to legitimise MTBBE as a compulsory educational model at national level. Through that legislative requirement, the introduction of the late-exit model will be phased from 2025 to reach Grade 7 by 2028. In the absence of a conceptual shift, the introduction of this revived model will likely fall victim to the same structural constraints as previous models.

This study employed a qualitative conceptual analysis approach informed by a critical interpretivist paradigm. Going beyond descriptive approaches that seek to reveal empirical data, the conceptual analysis approach sought to critically map, analyse, explore, and deconstruct the conceptual parameters, definitions and assumptions associated with a particular concept (Wiesner 2022), the MTBBE in this case. The focus shifted away from mere descriptions to a critique of how concepts work to construct meanings, rather than just establishing their presence or absence (Wiesner 2022). In this study, the critical perspective was applied in critically analysing the underlying presuppositions, power dynamics and knowledge structures embedded within dominant discourses to show how well-intentioned language policies end up reproducing linguistic hegemony associated with Eurocentrism.

The analysis was guided by three interrelated theories that offered critical lenses from which the analysis could be undertaken. Postcolonial language theory (Ngũgĩ wa Thiong’o 1986) was employed to examine the privileging of English and the marginalisation of African languages within educational contexts. It guided the critical examination of the assumption that mother tongues should be considered subsidiary languages. Decolonial theorisations about multilingualism and epistemic justice (Fricker 2007; Makalela 2019) offered an analytical framework for questioning the relationship between language policies and epistemic inequalities. The decolonial lens provided guidance to evaluate the extent to which MTBBE approaches are based on a concern for epistemic justice and language access, or if they go beyond linguistic concerns. The translanguaging theory (García & Wei 2014) offered a useful framework for the evaluation of language practices beyond additive bilingualism. It guided the critical evaluation of existing MTBBE models based on the extent to which these models acknowledge learners’ full linguistic resources. Figure 1 presents the theoretical positions and their contribution to the study.

The analysis process involved three steps. Initially, literature around the themes of MTBBE, multilingual education, language policy, translanguaging, and African language education was reviewed to unearth common arguments, assumptions and controversies characterising current debates about mother tongue education. The way in which MTBBE was presented in relation to literacy education, bilingualism, language transitions, educational reform, and social justice was especially significant.

Secondly, these common assumptions were examined using the analytical perspectives provided above. Through a process of iterative reading, comparison and synthesis of common themes, a series of dominant myths arose in connection with MTBBE. Finally, conclusions from the analysis of these myths were synthesised into a transformative framework for MTBBE. The framework thus, stands as an empirically underpinned theoretical framework for future application rather than a tested intervention.

This article followed all ethical standards for research without direct contact with human or animal subjects.

National assessment reports document that a sudden and early shift from teaching learners in their home languages to English-medium schooling leads to a significant drop in literacy and numeracy skills among such learners – a pattern of achievement that has been confirmed severally through Progress in International Reading Literacy Study (PIRLS) progress reports (Spaull 2019). While the learner’s home language is considered the best medium for gaining conceptual knowledge, it is discarded once the learner passes the initial stages of early childhood education. They become victims of language policies which demand a rapid transition from learning in their home languages to English after the initial levels, particularly in Grade 4. This highly publicised policy-practice gap demonstrates that in spite of all the progressive rhetoric in legislation, the actual advantages of mother tongue education apply only to the Foundation Phase (Heugh 2013; Stoop 2017). According to Heugh (2009), even though multilingual education policies exist, ‘their practice is driven towards expensive monolingual practices which invoke contemporary, external-to-Africa discourses about education’ (p. 185).

Several deficit narratives are frequently used to justify early withdrawal from the use of African languages. The first one is a Eurocentric vocabulary-based argument that African indigenous languages lack the capacity for complex technical vocabulary necessary to engage in advanced levels of academic research or science. In pragmatic terms, such an argument implies that early exposure to English has a protective element that would prevent students’ future academic difficulties when the colonial language takes absolute dominance in schooling. Other equally unsubstantiated reasons given are the alleged impracticality of using mother tongues due to linguistic diversity in Africa as well as the possibility of national disintegration. There is also the assumption that proficiency in English is the only way into globalised knowledge production.

Critical scholars dismiss such views as persistent relics of colonial language ideologies. Probyn (2024) observes that coloniality-informed language ideologies dominate education through English, while suppressing home languages of both teachers and learners as legitimate learning resources. Applied linguists argue that African languages have the capability to express sophisticated scientific ideas when they are exposed to systematic corpus construction and additive approaches to teaching and learning (Heugh 2009; Makalela 2018, 2024).

According to the postcolonial analysis, such widespread deficit theories are directly related to the phenomenon referred to by Ngũgĩ wa Thiong’o (1986) as ‘the linguistic colonisation of the mind’. Prah (2009) criticises the post-apartheid African elite for propagating colonial myths even as they pay lip service to the inherent value of mother tongues. For Van Pinxteren (2022), learners are forced to switch to English even before they achieve stable home-language literacy and second-language fluency. In turn, Sibanda (2017) dismisses two problematic assumptions associated with the transitional approach. Firstly, it assumes that students acquire full mother tongue literacy by Grade 3. Secondly, that home language literacy transfers to second language literacy to the point that by Grade 4, learners have sufficient proficiency in English to use it as a LoLT. Because of the falsity of these assumptions, the transition simply adds up to cognitive overload, linguistic attrition, and what scholars have called ‘strategic silences learners adopt in linguistically strange surroundings’ (Sibanda 2017).

Despite the official endorsement of additive bilingualism in policies, teachers often find themselves ‘smuggling the vernacular’ into the process of teaching through ad hoc code-switching and translanguaging to ensure understanding, although all instructional and testing materials are strictly monolingual (Cekiso 2025).

Against the backdrop of this dynamic, a dynamic that mirrors many other global trends in postcolonial educational exclusion, this article fundamentally redefines the very foundations of MTBBE as practised in South Africa. It presents MTBBE as a project of transformation, not merely a transient scaffold towards the acquisition of English language competency or a logistical expansion of mother tongue teaching. This article is neither a critique of any empirical instances of MTBBE implementation, nor a comparative study of diverse bilingual education schemes. The aim is to challenge existing conceptual misunderstandings and develop a transformative theory of MTBBE.

Persistent policy amnesia is evident in South Africa in the emergence of MTBBE being characterised by a pattern of ‘change without memory’ (Young 2004). Rather than learn from the problems attending decades of MTBBE implementation, policymakers continue to roll out MTBBE as a pedagogic innovation. More disturbing is how the political discourse around MTBBE selectively highlights the advantages of mother tongue deployment in the curriculum while conveniently ignoring and downplaying the logistical, material and structural factors (Bacchi 2009), and inconsistencies between policy formulation and practice (Wildsmith-Cromarty & Balfour 2019) for the failure of previous language-in-education policies. As a result of policy amnesia, the emerging discourse on MTBBE is rendered a cosmetic or an administrative affair (Wiesner 2022). The challenges of dealing with English exam washback, inadequate teacher education and the absence of disciplinary materials in African languages trap the curriculum in an endless cycle where progressive language rights are re-legislated on paper but destroyed in their practical implementation.

The renewed interest in MTBBE in South Africa can more aptly be referred to as a policy revival than an innovation. This is instructive to an appreciation of why earlier progressive language education policies have failed to deliver consistent results (Heugh 2013) or provided inconsistent and disjointed results (Taylor & Coetzee 2013).

Moreover, the sudden shift from the native language of instruction to English at Grade 4 constitutes a key systemic bottleneck for effective teaching (McKay & Romm 2015). The problem is not about lacking progressive policies, but about failing to provide consistent support for these policies. There is much reproduction of multilingual rhetoric (Kaschula & Maseko 2014) rather than meaningful contribution to educational transformation. In this regard, MTBBE is important, not only to enhance educational practice, but also as a significant avenue for promoting linguistic equality and epistemological inclusion.

The official resurgence of MTBBE is animated by, not only literacy concerns, but also the disruption of entrenched linguistic hierarchies. For Heugh (2013), there is a significant disconnect between history and theory in South African education policy, with the state continually adopting progressive and additive rhetoric of multilingualism in the introductory stage of curricula, but continuing to adopt models where learners have to undergo an early structural shift to English-medium instruction. Through framing indigenous languages as a ‘temporary stepping stone’ in the early stages of schooling, the curricula effectively construct African languages as structurally and/or cognitively inadequate for disciplinary thought.

On the contrary, the decolonial dimension calls for an entirely different epistemological paradigm, one which actively challenges these hierarchies imposed by the West in a monolingual context. In the decolonial view, language cannot be understood as static, separate and compartmentalised boxes of cognition that Makalela (2016) describes as a now-defunct Euro-Colonial categorisation created specifically for the purpose of separating languages from each other. Decolonial practice goes hand in hand with the philosophical framework of Ubuntu translanguaging, according to which African multilingualism is fluid, porous and interconnected. Makalela (2016) asserts that any transformational pedagogy should rely on the complete multilingual repertoire of the learner to achieve conceptual understanding and develop literacy. Thus, in the light of the current MTBBE paradigm, it becomes clear what choice should be made between mere symbolic representation of diversity, and long-lasting fluid decolonial multilingualism, recognising African languages as continuous sources of knowledge production.

During International Mother Language Day or Heritage Day, there is a flagging of, and emphasis on, the cultural and identity dimensions of mother tongue instruction to attain linguistic equity and epistemic decolonisation. If care is not taken, such commemorative events can position, in the minds of people, solely the symbolic function of the mother tongue in Education that amounts to token and performative multilingualism. For the MTBBE to constitute real transformation beyond superficial revival, rhetoric and advocacy need to be complemented by relevant and adequate resources, as well as sound pedagogical praxis.

A careful assessment of the existing Language in Education Policy (LiEP) frameworks shows that the problem of MTBBE stagnation does not just result from poor implementation, but is a reflection of fundamental pathologies in the existing concept of MTBBE (Heugh 2006). To create a transforming MTBBE practice, it is important to interrogate the emergent misconceptions proposed from the literature and policy discourse associated with MTBBE in South Africa’s education system.

Having debunked the hegemonic notions about mother tongue education as a temporary and precarious scaffolding device, this section of the article re-theorises MTBBE, not simply as a technocratic adaptation, but a complex ideology of instruction. There is a need for a shift from language being a tool for gaining content knowledge, to utilising it as the knowledge and context itself. For me, genuine MTBBE involves five mutually reinforcing concepts supported by various theoretical perspectives. These are:

Seeing that the MTBBE model is a resurgence on account of the previous models not having succeeded, there is a need for a radical bilingual framework that eschews the limitations of the previous models and translates the reimagination into practice, through a phased model of MTBBE implementation. I propose an MTBBE framework that draws on the three pillars as support for a gradual, phased MTBBE implementation. I propose a phasing in of the MTBBE through the three phases of Grade R-3, 4–7 and 8+.

The practical implementation of the framework is in the:

The linguistic realities of South African classrooms are so complex that deploying a single model, however good, cannot be enough. The MTBBE model I have envisioned can realise its full epistemic potential if there is a complementary and synergistic deployment of other related models alongside it. I discuss how the translanguaging pedagogy, dual-medium instruction, multimodal literacy and the Language Across the Curriculum (LAC) models can enrich the MTBBE model.

Translanguaging pedagogy acknowledges learners’ full linguistic repertoires as holistic and valid resources for communication. In the context of MTBBE, translanguaging legitimises the fluidity of language interaction and integration and dismantles the arbitrary boundaries between the mediums of instruction. Learners draw on their full linguistic and semiotic resources to negotiate meaning, enhance agency and foster cognitive flexibility (García & Wei 2014). Applied together, translanguaging animates MTBBE’s structural bilingualism in a way that helps learners to connect advanced or complex concepts employing familiar semiotic resources at their disposal.

With dual-medium instruction, there is an alternation or combination of two languages in a lesson within a subject area. For me, by two languages, I mean the mother tongue and the supposed global language, even when there is more than one mother tongue in the context. These allow for parallel bilingualism. While the early-exit model privileges the global language, dual-medium instruction fosters a sense of equilibrium between the mother tongue and the international language across the schooling phase. According to a study by Heugh (2013), dual-medium instruction enhances the acquisition of both the content and the language.

Multimodal pedagogy deploys a range of visual, digital and kinaesthetic modes alongside linguistic expression. The deployment of the full range of semiotic resources is particularly important in multilingual contexts, as that provides learners access to meaning even when linguistic mediation falters (Sibanda & Chikiwa 2024). Multimodality then comes in to complement MTBBE by extending literacy beyond words in ways that affirm diverse cognitive pathways that we use to convey meaning in everyday communication.

Language across the curriculum proceeds on the premise that every teacher is essentially a language teacher tasked with providing language-supportive strategies across all the subjects of the curriculum to enhance academic literacy in both languages operative within the specific context. The LAC model complements MTBBE by extending bilingual learning to the entire curriculum. The LAC model is pedagogically transversal.

The deployment of these models in conjunction with the MTBBE model will provide a holistic, multidimensional language-responsive system that enhances participation and access.

The viability and structural weaknesses of the proposed model find their parallels in mother tongue-based multilingual education (MTB-MLE) and intercultural bilingual education (IBE) around the world. Analysis of these examples helps to pinpoint the exact prerequisites necessary for the success of structural translation from policies into practice. Table 2 summarises important insights from comparable global contexts instructive to the South African context.

A comparative synthesis of the literature reveals that an effective MTBBE approach relies heavily upon the development of a unified and state-supported infrastructure. Genuine change will be brought about by coordinated changes in attitudes towards language, policy reform within institutions and significant investment in personnel and educational materials. Moreover, domestic implementations of international ‘best practice’ curricula face major obstacles in implementation due to lack of locally relevant and linguistically grounded planning (Mason et al. 2024).