Language remains a central determinant of epistemic access in South African Foundation Phase classrooms, where many learners begin schooling in a language other than their home language. The 2025 shift in language-in-education approaches renewed attention to mother tongue-based bilingual education (MTBBE) as a pathway to equity and meaningful curriculum access, particularly within Science, Technology, Robotics, Engineering, Arts and Mathematics (STREAM) education.
Aim
This review explored how MTBBE supported conceptual understanding, learner agency and educational inclusion in the Foundation Phase through sustained home-language instruction and additive English use.
Setting
A bibliometric analysis across global and South African contexts was conducted.
Methods
We conducted a bibliometric review, complemented by qualitative thematic analysis. Guided by the Preferred Reporting Items for Systematic reviews and Meta- Analyses (PRISMA) selection process, we sourced and screened Scopus-indexed literature published between 2015 and 2025 using predefined criteria and analysed the selected studies within a transformative research paradigm.
Results
Three themes emerged: (1) multilingualism and MTBBE as pathways to epistemic access and social justice; (2) additive bilingual pedagogies in early schooling; and (3) positioning African languages in global and local scholarship. Although global scholarship increasingly recognises multilingual learning and translanguaging, it seldom frames these within additive bilingual education as mechanisms of epistemic access.
Conclusion
By advancing both pedagogical and social justice aims, MTBBE deepened learning and reduced early cognitive barriers linked to language exclusion.
Contribution
This study advances bilingual education scholarship by synthesising evidence-based research on MTBBE in the Foundation Phase, subsequently informing teacher education and policy implementation debates central to education.
additive language teaching approachconceptual fluencyFoundation Phasemother tongue-based bilingual educationsocial justice in educationFunding information The authors received no financial support for the research, authorship, and/or publication of this article.IntroductionBackground
Within Foundation Phase education, the following quote from Ludwig Wittgenstein foregrounds the relationship between language, meaning-making and epistemic participation:
‘The limits of my language mean the limits of my world’. (Wittgenstein 1922, Proposition 5.6)
Young learners develop concepts, reasoning, cultural understanding and social identity through the linguistic and sociocultural resources available to them. When the mother tongue is prematurely displaced without adequate sociocultural mediation and additive bilingual support, learners may experience barriers to conceptual access, participation and meaning-making within the classroom (Cummins 2007; Sibanda & Tshehla 2025). The dynamic of learners’ worlds, especially for learning, is layered and socially produced. Learners grow their thinking to understand the world through a sociocultural approach to learning (Vygotsky 1978). South Africa has a diverse and multilingual learning environment (Department of Basic Education [DBE] 2024; Sibanda 2019) in which learners from various language backgrounds can thrive if given enough opportunities to engage with each other and learn from one another – regardless of their language differences.
Classrooms should not be spaces where learning in more than one language is discouraged (Cekiso 2026; García-Mateus & Palmer 2017), especially in the Foundation Phase – and more so in Science, Technology, Robotics, Engineering, Arts and Mathematics (STREAM) subject domains that are conceptually demanding (Schoonen & Motaung 2025). Linguistic repertoires are important as classrooms are diverse; a learner’s mother tongue might be Setswana; however, that learner might be in an English home language monolingual classroom, which can be challenging as their semantic and linguistic foundation is rooted in their mother tongue. With such a challenge, learners might find it difficult to participate in the classroom as they must cognitively process their thoughts before engaging in the discourse (Cekiso 2026; Jorolan-Quintero 2018). In the South African Foundation Phase curriculum, STREAM subjects are currently taught as Mathematics and the rest are encapsulated in the Life Skills Beginning Knowledge strand (DBE 2011a, 2011b, 2021). These subjects are conceptually demanding and rely on strong language-mediated meaning-making. The knowledge development, interactions, culturally situated practices, critical thinking, discourse and active engagement these subjects require necessitate sociocultural grounding (Essien & Moleko 2025; Parker & Gastrow 2026; Schoonen & Botes 2025).
From the above perspective, additive bilingualism does not simply add vocabulary to lessons or enable translation. Additive bilingualism enables learners to extend conceptual understanding across languages. Subtractive language models, or monolingual education, can be cognitively and ethically problematic in multilingual environments and in a global learning space, while additive language models, such as MTBBE (DBE 2024), are emancipatory and promote linguistic justice.
Multilingualism and additive language learning
Multilingualism in educational research is increasingly conceptualised as a dynamic linguistic repertoire rather than balanced native-like proficiency in multiple languages. Early definitions often framed multilingualism as the possession of equal mastery across languages, but contemporary scholarship rejects this symmetry model as both unrealistic and theoretically limiting (Cenoz 2013; Li 2018). Instead, multilingual competence is understood as context dependent and functional: Individuals mobilise different languages to varying degrees depending on social, cognitive and educational demands. From this perspective, multilingualism is not a static attribute, but a socially situated practice shaped by identity, interaction and institutional conditions.
This shift is significant for education because it reframes linguistic diversity as a resource rather than a deficiency, especially in mathematics (Essien & Moleko 2025) and early Science (Behrmann 2018). Li (2018) argues that multilingual learners draw on an integrated repertoire in which boundaries between languages are fluid, enabling flexible meaning-making. Such a view aligns with sociocultural approaches to learning, which position language as the medium through which knowledge is negotiated and internalised (Vygotsky 1978). In multilingual classrooms, learners’ ability to access concepts is therefore inseparable from their ability to engage with language systems that are meaningful to them. Multilingualism becomes a condition of epistemic participation: it determines how learners enter academic discourse and construct understanding within formal schooling.
Additive bilingualism extends the concept of multilingualism into a pedagogical framework. The distinction between additive and subtractive bilingualism, first articulated by Lambert (1975), remains foundational in bilingual education research. Additive bilingualism occurs when a learner’s first language is maintained and continues to develop while additional languages are acquired. Subtractive models, by contrast, replace the home language and often lead to linguistic erosion. Cummins (1996, 2007) argues that additive bilingual education supports cognitive development because conceptual knowledge transfers across languages. When learners build academic understanding in their strongest language, that understanding becomes available in subsequent languages, strengthening rather than fragmenting learning.
Modern dual language research reinforces this position by framing additive bilingualism as an educational orientation that treats multilingualism as an asset (Coetzee-Van Rooy 2022; De Jong, Yilmaz & Marichal 2019; García-Mateus & Palmer 2017). In additive models, languages are not compartmentalised but interact to support conceptual continuity. This orientation aligns with García and Li’s (2014) language-as-resource framework, which positions bilingual education as a means of expanding learners’ communicative and epistemic capacities. Additive bilingualism does not aim to produce perfectly balanced bilingual speakers (Li 2018). Its educational value lies in sustaining learners’ conceptual worlds while extending their engagement with additional languages. In the Foundation Phase, where conceptual foundations are formed, subtractive language practices risk undermining the sociocultural resources essential for meaning-making (Behrmann 2018; Coetzee-Van Rooy 2022; Parker & Gastrow 2026). Multilingualism is therefore conceptualised as a resource for epistemic access, informing the theoretical frameworks that follow.
Sociocultural and additive language theoretical underpinning
These observations are grounded in a sociocultural theory of learning (Vygotsky 1978) that positions language as the primary mediational tool through which children construct meaning and develop higher-order thinking. From the Vygotskian perspective, learning is not an individual cognitive event, but a socially situated process (Vygotsky [1941] 1997) in which language enables learners to internalise concepts through interactions with others.
As discussed earlier, additive bilingualism supports conceptual transfer across languages (Cummins 1996, 2007), strengthening rather than fragmenting academic development. In this sense, bilingualism is not a barrier to learning, but an expansion of learners’ cognitive and epistemic resources. García and Li’s (2014) language-as-a-resource framework similarly rejects deficit views of multilingualism and positions learners’ full linguistic repertoires as an asset for meaning-making (Li 2018). This positions language not only as a mediational tool but as central to questions of epistemic access (August-Mowers 2025). Framed within a transformative paradigm, this study treats language not as a neutral medium but as a site where power, identity and access to knowledge are negotiated.
Policies on multilingual education implementation
In the South African context, these theoretical concerns are inseparable from ongoing policy debates about language and education. Despite a constitutional commitment to multilingualism, the early transition to English as the dominant language of learning and teaching continues to reproduce inequalities in access to curriculum knowledge (Ngubane & Ngwenya 2025). The Language in Education Policy (LiEP) and renewed policy attention towards mother tongue-based bilingual education (MTBBE) as a national priority in 2024 (DBE 2024) signal a recognition that sustained home language instruction in the Foundation Phase is essential for equitable learning. Yet, implementation remains uneven, and subtractive language practices persist in many classrooms. For learners who are required to engage with conceptually demanding content through a language they do not yet understand, language becomes a structural barrier rather than a bridge for learning. Owen-Smith (2010) reiterated that receiving instruction in the language of teaching and learning that differs from learners’ home language undermines the development of a strong foundation of cognitive and academic development in the home language, as echoed by Olawale, Hendricks and Rusi (2024) and Wright and Baker (2025) in more recent years. Mother tongue-based bilingual education offers a policy-aligned framework that affirms African languages as legitimate languages of knowledge, while introducing English additively, positioning bilingualism as a pathway to both local identity and global participation (Ngubane & Ngwenya 2025). In this sense, language policy is not merely administrative; it is a justice-orientated intervention into how learning opportunities are distributed.
Despite growing recognition of multilingualism in both scholarship and policy, the field remains fragmented across cognitive, pedagogical and justice-oriented perspectives (August-Mowers 2025). There is limited systematic synthesis of how multilingual education is conceptualised as a mechanism of epistemic access in the Foundation Phase, and how MTBBE is positioned within this discourse. This article addresses this gap through a bibliometric analysis of global and South African scholarship, examining patterns of conceptualisation, thematic relationships and knowledge production. This approach provides a structured overview of the field and supports more coherent theoretical and practical engagement. The guiding question for the current bibliometric review is: How is multilingualism framed in the literature as a mechanism of epistemic access and justice in Foundation Phase education, and what role does mother tongue-based bilingual education play within this discourse?
Research methods and design
This study adopted the bibliometric research design to map and analyse scholarly trends related to multilingualism, MTBBE and its responsiveness to social justice in early education. Bibliometric analysis was selected to enable a systematic and transparent examination of how knowledge in this field has been conceptualised, clustered and prioritised across global and South African contexts over time. Beyond descriptive mapping, the analysis sought to interrogate how dominant discourses shape understandings of language learning and access within early schooling. The approach aligns with a transformative research paradigm, which views educational research as a means of exposing structural inequalities and inequities and advancing epistemic justice (Romm 2018). In this study, multilingualism is not treated as a neutral pedagogical variable, but a socially situated construct linked to power, identity and participation in knowledge making.
This stance is theoretically grounded in sociocultural and additive bilingual frameworks that position language as a mediational tool for learning and as a resource for equity. Framing the bibliometric analysis within a transformative paradigm therefore enables the study to move beyond cataloguing trends towards critically examining how our scholarly knowledge supports or constrains linguistic inclusion in the Foundation Phase. Peer-reviewed literature was sourced from Elsevier (2021), selected for its comprehensive coverage of education, social sciences and language-related disciplines, as well as its compatibility with bibliometric analysis software (Van Eck & Waltman 2010). Scopus was selected because of its comprehensive coverage of peer-reviewed literature, consistent metadata structure and compatibility with bibliometric analysis tools, while databases such as Google Scholar were not included because of limitations in data transparency and export reliability. A structured search string was developed and applied in Scopus to generate the initial results (Elsevier 2021). The following initial search string was used:
(ABS (“MTBBE” OR “mother tongue based bilingual education” OR “African language*” OR “indigenous language*” OR “linguistic repertoire*” OR “language background*” OR “native language*” OR “additive language*” OR “additional language*” OR “monolingual learning” OR “LOLT” OR “language of learning*” OR “language acquisition” OR “home language*” OR “mother tongue*” OR “dual language*” OR “first language*” OR “parallel medium” OR “additive bilingual*” OR “second language*” OR “English Language Learners” OR ESL OR ELL OR “bilingual*” OR “multilingual*” OR “translanguag*” OR “language of learning and teaching” OR “Language diverse”) OR KEY (“Linguistically responsive” OR “contextually relevant” OR “socio-cultural” OR “sociocultural” OR “conceptual fluency” OR “language comprehension”) OR ABS (ECD OR ECE OR “early educat*” OR “early learn*” OR “foundation phase” OR “early year*” OR “early childhood*” OR “early grade*” OR “primary education” OR “learning environment” OR “elementary school”) OR KEY (“Epistemic access” OR “equity” OR “inclusion” OR “social justice” OR “culturally responsive pedagogy” OR “transformative pedagogy”))
Searches were conducted within abstracts (ABS) and author keywords (KEY) to balance sensitivity and relevance. The initial retrieval prioritised breadth to ensure a rich and representative corpus, after which results were refined through journal selection, subject area filtering, relevance ranking and publication date. We made use of the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)-filtering approach (Page et al. 2021) in which raw datasets are filtered using parameters such as subject domains, topic-specific keywords, relevant journals, date ranges and document types to screen and select the focused documents from the broad search results. Table 1 explains the PRISMA-filtering process we followed.
Preferred Reporting Items for Systematic reviews and Meta-Analyses analysis of Scopus-filtering of available scholarship.
Phase
Description
Exclusion description
Documents (n)
1. Identification
1.1. Documents on Scopus database
No filter yet, search with source string
683 372
2. Screening & filtering
2.1. Excluded data using parameters set
Subject domain and keyword filter – domains and keywords unrelated to education and the topic were removed
−353 431
2.2. Documents after filtering
-
329 941
3. Eligibility
3.1. Documents excluded based on review
Removing non-subject-related journals (Source Title on Scopus)
−232 232
3.2. Documents after screening for eligibility
-
121 199
4. Included
4.1. Documents excluded after date range filter and document type selection
Date range filter to exclude publications prior to 2015 and document types – only academic publications
−74 340
4.2. Documents for bibliometric analysis of Global dataset
Final keyword and source type refinement – and removal of psychology subject domain)
46 859
4.3. Documents for bibliometric analysis of South African dataset
Filter for data for South African context (46 859–45 596)
1263
We extracted two subsets of data, one with the global publication results and one focusing on South African publications specifically to enable contextual comparison with global trends. From the 46 859 documents, a total of 1263 were publications focused on the South African context.
Bibliographic records were exported from Scopus and analysed using VOSviewer to generate keyword co-occurrence networks (Van Eck & Waltman 2010). Separate visualisations were produced for the global dataset and the South Africa–specific subset. Co-occurrence thresholds were applied to identify high-frequency and high-centrality terms, enabling the identification of dominant clusters and thematic relationships.
The visual maps were interpreted through thematic analysis (Cohen, Manion & Morrison 2018), guided by sociocultural learning theory, additive bilingualism and language-as-resource frameworks. This interpretive layer allowed the bibliometric patterns to be analysed in relation to issues of epistemic access, linguistic justice and early learning.
This study relied exclusively on publicly available, peer-reviewed secondary data and did not involve human participants, personal data or institutional records. As such, formal institutional ethics clearance was not required. Nevertheless, ethical principles of rigour, transparency and respect for scholarly representation were upheld throughout the research process.
We aimed to avoid misrepresentation of studies, to document inclusion and exclusion decisions transparently and to interpret bibliometric patterns reflexively, particularly given the historical and political sensitivities surrounding language, race and access in South Africa, through a transformative lens (Romm 2018). The analysis sought to highlight dominant trends as well as silences in the literature, in keeping with the study’s social justice orientation.
Ethical considerations
Ethical clearance to conduct this study was obtained from the University of the Witwatersrand Human Research Ethics Committee (non-medical) (No. H25/03/05).
Results
The bibliometric analysis, visualised through VOSviewer keyword co-occurrence maps, revealed three interrelated thematic patterns across the global dataset and the South Africa–specific subset. While multilingualism emerged as a shared organising construct across both datasets, the ways in which it was conceptualised, enacted and discussed differed significantly. The findings highlight how early schooling, language pedagogy and questions of justice intersect, particularly within the South African context. The figures below contain a comparison between the prominent themes in global and South African scholarship between 2015 and 2025 about multilingualism and additive language learning approaches in early learning.
Figure 1 visualises the global keyword co-occurrence network and highlights multilingualism as the dominant organising mode linking pedagogical, linguistic and cognitive clusters. The largest cluster centres around teaching, curricula and learning systems, suggesting that multilingualism is increasingly recognised as relevant to instructional design, rather than treated as a peripheral learner characteristic.
Scholarly writing trends about multilingual education in the global context.
A second cluster focused on language development, literacy and cognition reflects the continued influence of psycholinguistic frameworks that conceptualise language primarily as an individual skill. Although a social justice cluster that connects diversity and inclusion is present, it remains structurally less central, suggesting that global scholarship often separates multilingual pedagogy from broader equity concerns. Within this configuration, translanguaging and bilingualism operate as mediating concepts but are not embedded within sustained additive language planning.
From the perspective of MTBBE, the global landscape reveals a field that acknowledges multilingual learning but has not yet consistently framed it as a systematic mechanism of epistemic access. This contrast becomes particularly visible when situated alongside the South African dataset.
The South African map reveals a noticeably different configuration in which multilingualism is not only pedagogical but explicitly justice oriented, providing a context in which MTBBE emerges as a structural educational priority.
Figure 2 presents the South African-specific network and reveals a dense coupling of multilingualism with social justice, literacy and Foundation Phase education. Unlike the global configuration, multilingualism in this context is positioned as inseparable from questions of access, inequality and identity. The central visibility of South Africa alongside multilingualism and literacy highlights how language scholarship reflects national commitments to multilingualism in the early years of schooling.
Scholarly writing trends about multilingual education in the South African context.
A prominent cluster linking social justice, diversity and decolonisation indicates that language is understood as a structural determinant of educational participation. At the same time, a distinct cluster foregrounds in official and indigenous African languages, including isiZulu, isiXhosa, Setswana, Afrikaans and Northern Sotho, to name a few that are visible, and signals an explicit recognition of these languages as vehicles of knowledge-making rather than transitional supports. The prioritisation of multilingual engagement in the Foundation Phase reflects a positive policy and scholarly orientation towards additive bilingualism even as implementation remains complex in linguistically diverse classrooms. Within this context, MTBBE is not framed as an optional pedagogical strategy but as a necessary response to historical inequities and contemporary learning realities.
The contrast between the global and South African networks highlights how multilingualism shifts from a general educational concern to a justice-driven framework, a shift that underpins the thematic structure discussed in the following sections.
Theme 1: Multilingualism as epistemic access and social justice in the Foundation Phase
The first theme examines multilingualism not simply as a linguistic characteristic of classrooms, but as a mechanism of epistemic access in early schooling. It explores how the bibliometric patterns connect language to literacy, conceptual development and social justice and argues that the Foundation Phase emerges as a critical site where inequities in language policy translate directly into inequities in learning. This theme situates MTBBE within sociocultural theory and early education research to demonstrate that language choice is inseparable from questions of participation comprehension and educational legitimacy.
Language as mediation and conceptual access
Across both datasets, as seen in Figure 1 and Figure 2, multilingualism emerged as a central node linking literacy, learning and child development from a sociocultural perspective, thus clustering supports Vygotsky’s argument that language is not merely a communication tool, but the primary mediational means through which higher mental functions develop (Vygotsky [1941] 1997, 1978). The South African network map, however, extends this theoretical insight into explicitly political terrain. Multilingualism is clustered with inequality, access and social justice, suggesting that language functions not only as a cognitive mediator but also as a structural gatekeeper to knowledge. This reinforces Cummins’ contention that denying learners access to their home language and schooling constitutes a form of educational disempowerment (Cummins 1996, 2007). When learners are required to engage with foundational concepts through an unfamiliar language, epistemic participation is restricted before academic ability can be meaningfully assessed, as echoed by August-Mowers (2025).
Early schooling as a justice site
The prominence of the Foundation Phase and early literacy in the South African map is not incidental. It signals a recognition that linguistic inequity is reproduced most powerfully in the early years of schooling. International research demonstrates that sustained mother tongue instruction strengthens conceptual understanding across subjects (Behrmann 2018; Nikièma 2011). South African policy developments reflect a similar orientation (DBE 2024). The DBE’s incremental rollout of MTBBE positions early multilingual instruction as a systemic equity intervention rather than a remedial measure (Ngubane & Ngwenya 2025). Nevertheless, the bibliometric patterns also reveal tension between policy intent and classroom realities. While multilingualism is theoretically endorsed, subtractive practises persist, particularly in contexts where English remains symbolically tied to mobility (Sibanda & Tshehla 2025; Wright & Baker 2025). This tension becomes particularly visible when interpreted through the bibliometric patterns identified in this study. The bibliometric mapping revealed that global scholarship on mother-tongue-based bilingual education clusters strongly around themes of linguistic inclusion, translanguaging, language acquisition and language policy, with the responsibility for inclusion often positioned within classroom language practices (see Figure 1). In contrast, the South African publication patterns (see Figure 2) demonstrated stronger associations with broader educational transformation, including STREAM education, epistemic access, socially responsive pedagogies and curriculum inclusion. These findings suggest that within the South African context, mother-tongue-based bilingual education is conceptualised not only as a linguistic intervention but also as part of a wider transformative educational project aimed at addressing historical inequalities and enhancing meaningful participation in learning. The findings, therefore, suggest that additive language learning practices in South Africa operate simultaneously as an emancipatory policy vision while it remains contested pedagogical practices in STREAM disciplines. South African scholars consequently foreground the need for transformative pedagogical approaches that foster inclusion and contribute to MTBBE implementation (August-Mowers 2025; Cekiso 2026; Essien & Moleko 2025; Ngubane & Ngwenya 2025; Olawale et al. 2024; Parker & Gastrow 2026; Schoonen & Motaung 2025; Sibanda & Tshehla 2025) while simultaneously highlighting the persistent gap between policy aspirations and classroom enactment.
Theme 2: Pedagogical presentations of additive bilingualism in early schooling
While the previous theme establishes multilingualism as a condition for epistemic access, the second theme shifts attention to how this access is enacted pedagogically in cultural practices. If language mediates learning, the question becomes how teachers operationalise additive bilingualism in real instructional settings.
Theme 2, therefore, examines translanguaging, bilingual scaffolding and curriculum mediation as pedagogical responses to linguistic diversity. It critically interrogates the gap between theoretical endorsement of multilingualism and the uneven implementation of structured bilingual models, positioning MTBBE as a framework that seeks to stabilise practice with a coherent pedagogical architecture.
Translanguaging as pedagogical mediation
Translanguaging appears in both datasets as a bridging construct that links classroom interaction, assessment and language learning. García and Li’s (2014) language-as-a-resource framework conceptualises translanguaging as a deliberate pedagogical strategy that leverages learners’ full linguistic repertoires to support meaning making (García & Li 2014). In practice, however, the bibliometric network suggests uneven theoretical anchoring translanguaging is visible as a practice but not always embedded within additive bilingual planning. Research on dual language models demonstrates that when translanguaging is structurally supported, it contributes to positive learner identities and sustained academic growth (De Jong et al. 2019; García-Mateus & Palmer 2017). When it remains ad hoc, it risks becoming a compensatory tactic rather than a transformative pedagogy (Ramafikeng 2024).
Mother tongue-based bilingual education as structured pedagogical justice
The gap between practice and policy coherence highlights the role of MTBBE as a structured framework that integrates translanguaging within intentional language sequencing. International evidence from Ghana, the Philippines and West Africa demonstrates that systematic mother-tongue-based models yield measurable gains in comprehension and participation (Jorolan-Quintero 2018; Nikièma 2011; Rosekrans, Sherris & Chatry-Komarek 2012). South African research similarly shows that bilingual preparation in mathematics and science strengthens conceptual engagement in early classrooms (Olawale et al. 2024; Schoonen & Botes 2025). Mother tongue-based bilingual education therefore represents more than bilingual exposure; it constitutes a pedagogical architecture that redistributes epistemic authority by legitimising learners’ linguistic knowledge (August-Mowers 2025). Within a transformative paradigm, this redistribution is inherently political: It challenges monolingual norms that privilege dominant languages as the sole carriers of academic legitimacy.
Theme 3: Contextual and epistemic positioning of African languages within global and local scholarship
The discussion turns to the broader contextual politics of language, having explored multilingualism as epistemic access and pedagogical enactment. Theme 3 situates multilingual education within historical and social-political frameworks by analysing how African languages are positioned in the literature.
This theme interrogates the contrast between global abstraction and South African contextualisation, highlighting how language education operates with empowered structures shaped by colonial legacies, policy reform and identity politics. It argues that the visibility of African languages in the bibliometric network is not merely descriptive but indicative of an ongoing struggle for epistemic recognition with informal schooling.
Naming languages as epistemic resistance
A striking feature of the South African network map is the explicit naming of African languages. Unlike the global data set, which often abstracts multilingualism into generalised categories, the South African literature situates language within historical and cultural specificity. It aligns with decolonial critiques that argue that epistemic justice requires recognising marginalised languages as a legitimate knowledge system (Cummins 1996; García & Li 2014). By foregrounding indigenous and African languages as languages of teaching rather than conditional tools, the literature resists deficit framings that required English with intellectual authority.
Policy tensions and global–local contradictions
The contrast between global abstraction and local contextualisation reveals an ongoing tension in multilingual scholarship. Global frameworks frequently prioritise cognitive efficiency and standardised outcomes, whereas South African research foregrounds historic inequity and cultural identity. Policy development such as the DBE’s MTBBE rollout represents a positive commitment to multilingual justice, yet implementation remains uneven in linguistically diverse classrooms (DBE 2024; Sibanda 2019). The bibliometric evidence suggests that African languages occupy a paradoxical position, symbolically affirmed yet structurally vulnerable. Mother tongue-based bilingual education emerges here as both a pedagogical and political intervention, seeking to stabilise African languages within formal education while negotiating the persistent prestige of English. This negotiation underscores the transformative potential and the fragility of multilingual reform.
Bridging the themes together
The three themes reveal multilingual education in the Foundation Phase as a site where recognition, pedagogy and power converge. Theme 1 establishes language as the foundation of epistemic access, showing that early schooling is where linguistic inequities most sharply translate into learning inequities. Theme 2 demonstrates that pedagogical responses to multilingualism are uneven and that MTBBE represents an attempt to institutionalise additive bilingualism as a structured educational justice rather than a discretionary classroom practice. Theme 3 situates these dynamics within a broader socio-political landscape, where the recognition of African languages challenges inherited hierarchies of knowledge and legitimacy. Critically, the South African bibliometric map foregrounds strong links between multilingualism and conceptually demanding domains such as mathematics, science and engineering, which in the Foundation Phase are embedded within Mathematics and the Life Skills Beginning Knowledge strand. This alignment reflects a policy recognition that early engagement with STREAM-oriented thinking requires deep conceptual comprehension, not surface-level language substitution (Parker & Gastrow 2026; Schoonen & Motaung 2025), especially in early Science learning (Parker & Gastrow 2026) and mathematics (Essien & Moleko 2025). The DBE’s incremental rollout of MTBBE explicitly positions mother tongue instruction as foundational to equitable participation in these domains (DBE 2024). Within a sociocultural framework, language mediates access to abstract reasoning and scientific discourse (Vygotsky 1978), and Cummins’ interdependence hypothesis suggests that sustaining home language learning strengthens rather than constrains academic growth (Cummins 2007).
Across all three themes, multilingualism emerges not as a technical variable but also as a contested terrain through which participation in education is negotiated. Mother tongue-based bilingual education operates within this terrain as both a pedagogical framework and political project, linking social-cultural theory, policy reform and classroom practice to expand who is permitted to understand, contribute to and shape knowledge in early schooling.
DiscussionImplications and guidelines for additive language teaching in the Foundation Phase
The findings of this study suggest that multilingual education in the Foundation Phase requires coordinated action across curriculum design, pedagogy, teacher preparation and policy implementation. The following guidelines are derived from the bibliometric patterns and theoretical framing and are proposed as principles for strengthening MTBBE as a social justice intervention.
Curriculum guidelines: Language as infrastructure for early Science, Technology, Robotics, Engineering, Arts and Mathematics learning
Theme 1 demonstrated a strong coupling between multilingualism and conceptually demanding domains such as mathematics and early STREAM learning. This supports sociocultural arguments that language mediates access to abstract reasoning (Vygotsky [1941] 1997, 1978). International evidence further shows that sustained mother tongue instruction strengthens conceptual understanding across disciplines (Behrmann 2018; Nikièma 2011). The following guidelines may contribute to language as an infrastructure for early STREAM learning:
Foundation Phase curriculum design should integrate language objectives within Mathematics and Life Skills Beginning Knowledge rather than treating language as a separate strand (Cummins 2007).
Science, Technology, Robotics, Engineering, Arts and Mathematics-related concepts should be introduced through learners’ strongest language to support reasoning before translation into English terminology (Cummins 1996, 2007).
Curriculum materials should include bilingual scaffolds that preserve conceptual continuity across languages (De Jong et al. 2019).
African languages should be used for explanation, argumentation and problem-solving discourse, not merely classroom management (Sibanda 2019; Wright & Baker 2025).
Early STREAM success depends on linguistic access, and subtractive language practices undermine conceptual development at the very moment foundations are laid. These guidelines align with the DBE’s incremental MTBBE rollout, which recognises mother tongue instruction as foundational to equitable participation in early learning (DBE 2024).
Pedagogical guidelines: Structured translanguaging within mother tongue-based bilingual education
Theme 2 revealed that translanguaging is visible in classroom literature but often lacks structural anchoring. García and Li (2014) argue that translanguaging becomes transformative only when embedded within additive bilingual planning. Evidence from dual language models confirms that structured bilingual pedagogy strengthens learner identity and academic engagement (De Jong et al. 2019; García-Mateus & Palmer 2017). We suggest the following guidelines for structured translanguaging in MTBBE:
Teachers should plan bilingual lessons intentionally rather than rely on spontaneous translation (Cummins 2007).
Translanguaging should support conceptual reasoning, peer explanation and problem-solving (Ramafikeng 2024).
Classroom discourse should legitimise learners’ full linguistic repertoires as cognitive resources (García & Li 2014).
Formative assessment should allow multilingual expression during conceptual development (Olawale et al. 2024).
The aim is not language alternation for its own sake, but conceptual mediation across linguistic systems, consistent with additive bilingual principles.
The gap between policy vision and classroom enactment identified in Theme 2 points to teacher preparation as a critical leverage point. Research demonstrates that bilingual instructional competence must be explicitly taught rather than assumed (Purkarthofer & Mossakowski 2011; Schoonen & Motaung 2025). The following guidelines can support preparing teachers as bilingual pedagogical experts:
Teacher education programmes should include explicit training in bilingual curriculum design and language-integrated content teaching (Cummins 2007).
Pre-service teachers should develop strategies for scaffolding abstract concepts across languages (Schoonen & Botes 2025).
Professional development should frame multilingualism as pedagogical strength rather than classroom deficit (De Jong et al. 2019).
Reflective components should address language, identity and equity in schooling (Cummins 1996).
This positions teachers as agents of epistemic access, not mere policy implementers.
Theme 3 highlighted tensions between symbolic affirmation of African languages and uneven implementation. While the DBE’s MTBBE policy signals a positive commitment to multilingual justice (DBE 2024), research cautions that reform without sustained support risks superficial compliance (Rosekrans et al. 2012; Sibanda 2019). We believe that the following guidelines can support stabilising policy commitments to multilingual education:
Long-term investment in African language materials, particularly for early STREAM learning (Schoonen & Botes 2025).
Monitoring frameworks that evaluate conceptual learning rather than speed of English transition (Nikièma 2011).
Policy communication that frames multilingualism as academic enrichment rather than remedial intervention (Cummins 2007).
Institutional support for teachers implementing bilingual pedagogy (Sibanda & Tshehla 2025).
These measures are necessary to protect multilingual commitments from erosion under systemic pressure. The policies must protect multilingual learning from being reduced to symbolic compliance.
These guidelines emphasise that multilingual education in the Foundation Phase is not an optional enrichment strategy but a structural condition for equitable learning. When language is positioned as infrastructure for reasoning, teachers are equipped to mediate knowledge bilingually and policy stabilises additive instruction, MTBBE becomes a coherent system rather than an isolated intervention. The conclusion returns to this central claim.
Limitations of the study
This study is limited by the inherent constraints of bibliometric analysis. While keyword co-occurrence mapping provides valuable insight into dominant scholarly patterns, it cannot capture the full nuance of classroom practice or policy enactment. The analysis reflects published discourse rather than lived educational realities. Furthermore, reliance on Scopus-indexed literature may privilege Global North publication patterns and underrepresent locally produced scholarship that circulates outside major indexing systems. Although a South Africa-specific subset was extracted to mitigate this imbalance, bibliometric methods cannot fully account for informal knowledge networks, practitioner literature or indigenous epistemologies that exist beyond academic publishing.
The interpretive layer of the analysis is shaped by a transformative theoretical lens that foregrounds equity and justice. While this orientation is explicit and intentional, alternative paradigms might yield different emphases in interpreting the same dataset. The study therefore does not claim neutrality but argues for a justice-oriented reading of multilingual scholarship. Future research should complement bibliometric mapping with classroom-based, ethnographic and policy-implementation studies to examine how MTBBE functions in practice.
Conclusion
This study set out to examine how multilingualism is conceptualised within global and South African scholarship and to position MTBBE as a justice-oriented response in the Foundation Phase. The bibliometric analysis reveals a field that recognises multilingualism as central to learning; however, it remains divided between skill-based and equity-driven framings. While global scholarship frequently treats multilingualism as a pedagogical resource, South African research situates language within a broader landscape of inequality, access and identity. This distinction is not merely contextual; it reflects different assumptions about what language education is meant to achieve.
The findings demonstrate that multilingualism in early schooling cannot be reduced to vocabulary acquisition or classroom accommodation. In the South African context, language operates as a structural determinant of participation in conceptually demanding domains, particularly mathematics and early STREAM-related learning embedded in the Life Skills curriculum. The prioritisation of MTBBE within national policy signals a positive commitment to redistributing epistemic access, even as implementation remains complex in linguistically diverse classrooms (DBE 2024). From a sociocultural perspective, denying learners the mediational resource of their strongest language restricts conceptual development (Vygotsky 1978), while additive bilingual frameworks expand cognitive and academic capacity (Cummins 2007).
Mother tongue-based bilingual education therefore emerges not only as a pedagogical model but also as a political intervention into how knowledge is distributed. It challenges monolingual norms that privilege dominant languages as gatekeepers of academic legitimacy and affirms African languages as vehicles of reasoning, literacy and scientific thought. The bibliometric patterns suggest that the Foundation Phase is the critical site where this redistribution can occur most effectively. When multilingual instruction is sustained in the early years, learners are positioned to engage with abstract, technical and disciplinary knowledge without sacrificing linguistic identity.
At the same time, the article highlights ongoing tensions between policy objectives and classroom practices. Translanguaging practices are visible but often lack systemic support; multilingual commitments are articulated yet unevenly realised. Future research must therefore move beyond mapping discourse towards examining how MTBBE is implemented, supported and sustained in everyday teaching practices. Teacher education, curriculum design and resource development remain central to translating multilingual policy into equitable learning environments.
This article argues that multilingual education in the Foundation Phase is a question of justice, not convenience. Language determines who can access knowledge, who can participate in reasoning, and whose ways of knowing are legitimised within formal schooling. Mother tongue-based bilingual education offers a framework for expanding these boundaries. In linguistically diverse societies such as South Africa, the ethical imperative is clear: Early education must not narrow learners’ worlds by subtracting their languages but must expand them through additive bilingual engagement.
Acknowledgements
The authors wish to acknowledge and thank the Wits School of Education’s Research Committee for partially contributing to the article processing charges of this article.
Competing interests
The authors, Andri Schoonen and Itumeleng Lesabe, declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
CRediT authorship contribution
Andri Schoonen: Conceptualisation, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualisation, Writing – original draft, Writing – review & editing. Itumeleng Lesabe: Conceptualisation, Resources, Validation, Writing – review & editing. Both authors reviewed the article, contributed to the discussion of results, approved the final version for submission and publication and take responsibility for the integrity of its findings.
Data availability
The data that support the findings of this study are available from the corresponding author, Andri Schoonen, upon reasonable request.
Disclaimer
The views and opinions expressed in this article are those of the authors and are the product of professional research. It does not necessarily reflect the official policy or position of any affiliated institution, funder, agency, or that of the publisher. The authors are responsible for this article’s results, findings, and content.
ReferencesAugust-Mowers, C.M., 2025, ‘The slowly turning wheels on mother tongue-based bilingual education (MTBBE) finally stopped’, South African Journal of Higher Education39(3), 16–36. https://doi.org/10.20853/39-3-7348Behrmann, T., 2018, Evaluating the effects of mother tongue on math and science instruction, ISTES Organization, Monument, CO.Cekiso, M., 2026, ‘Mother tongue as a medium of instruction in the foundation phase: A comparison of classroom reality with the South African language-in-education policy’, South African Journal of Childhood Education16(1), a1777. https://doi.org/10.4102/sajce.v16i1.1777Cenoz, J., 2013, ‘Defining multilingualism’, Annual Review of Applied Linguistics33, 3–18. https://doi.org/10.1017/S026719051300007XCoetzee-Van Rooy, S., 2022, ‘Social cohesion and childhood multilingualism in South Africa’, in A.Stavans & U.Jessner (eds.), The Cambridge handbook of childhood multilingualism, pp. 24–41, Cambridge University Press, Cambridge.Cohen, L., Manion, L. & Morrison, K., 2018, Research methods in education, 8th edn., Routledge, London.Cummins, J., 1996, Negotiating identities: Education for empowerment in a diverse society, 2nd edn., California Association for Bilingual Education, Los Angeles, CA.Cummins, J., 2007, ‘Rethinking monolingual instructional strategies in multilingual classrooms’, Canadian Journal of Applied Linguistics10(2), 221–240.De Jong, E.J., Yilmaz, T. & Marichal, N., 2019, ‘A multilingualism-as-a-resource orientation in dual language education’, Theory into Practice58(2), 107–120. https://doi.org/10.1080/00405841.2019.1569375Department of Basic Education (DBE), 2011a, Curriculum and assessment policy statement (CAPS): Foundation Phase Life Skills Grades R–3, Government Printing Works, Pretoria.Department of Basic Education (DBE), 2011b, Curriculum and assessment policy statement (CAPS): Foundation Phase Mathematics Grades R–3, Government Printing Works, Pretoria.Department of Basic Education (DBE), 2021, Proposed amendments to the Curriculum and Assessment Policy Statement (CAPS) to make provision for Coding and Robotics Grades R–9: Draft Curriculum and Assessment Policy Statement Grades R–3, DBE, Pretoria.Department of Basic Education (DBE), 2024, Mother tongue-based bilingual education to be rolled out incrementally by 2025 report, DBE, Pretoria.Elsevier, 2021, Scopus, Elsevier, Amsterdam, viewed 31 January 2026, from https://www.scopus.com.Essien, A.A. & Moleko, M., 2025, ‘Research on language and multilingualism in mathematics education in sub-Saharan Africa’, Africa Education Review21(1), 25–49. https://doi.org/10.1080/18146627.2025.2455577García, O. & Li, W., 2014, Translanguaging: Language, bilingualism and education, Palgrave Macmillan, Basingstoke.García-Mateus, S. & Palmer, D., 2017, ‘Translanguaging pedagogies for positive identities in two-way dual language bilingual education’, Journal of Language & Identity & Education16(4), 245–255. https://doi.org/10.1080/15348458.2017.1329016Jorolan-Quintero, G., 2018, ‘Oral traditions: An aid to implementation of mother tongue-based multilingual education in the Philippines’ basic education programme’, International Review of Education64(6), 759–777. https://doi.org/10.1007/s11159-018-9743-9Lambert, W.E., 1975, ‘Culture and language as factors in learning and education’, in A.Wolfgang (ed.), Education of immigrant students, pp. 55–83, Ontario Institute for Studies in Education, Toronto, ON.Li, W., 2018, ‘Translanguaging as a practical theory of language’, Applied Linguistics39(1), 9–30. https://doi.org/10.1093/applin/amx039Ngubane, S.P. & Ngwenya, E.T., 2025, ‘Implementing mother tongue-based bilingual education through translanguaging in multilingual classrooms: A systematic review’, International Journal of Learning, Teaching and Educational Research24(6), 741–757. https://doi.org/10.26803/ijlter.24.6.34Nikièma, N., 2011, ‘A first-language-first multilingual model to meet the quality imperative in formal basic education in three “francophone” West African countries’, International Review of Education57(5–6), 599–616. https://doi.org/10.1007/s11159-011-9253-5Olawale, B.E., Hendricks, W. & Rusi, L., 2024, ‘Bilingual teaching in South Africa: A qualitative inquiry into the strategies used for the preparation of mathematics teachers in the foundation phase’, Cogent Education11(1), a2428884. https://doi.org/10.1080/2331186X.2024.2428884Owen-Smith, M., 2010, ‘The language challenge in the classroom: A serious shift in thinking and action is needed’, Focus56, 31–37, viewed 15 December 2025, from https://hsf.org.za/publications/focus/focus-56-february-2010-on-learning-and-teaching/the-language-challenge-in-the-classroom-a-serious-shift-in-thinking-and-action-is-needed/download.Page, M.J., Moher, D., Bossuyt, P.M., Boutron, I., Hoffmann, T.C., Mulrow, C.D. et al., 2021, ‘PRISMA 2020 explanation and elaboration: Updated guidance and exemplars for reporting systematic reviews’, BMJ Open372, n160. https://doi.org/10.1136/bmj.n160Parker, S. & Gastrow, M., 2026, ‘Language, learning and engagement: Science communication skills among scientists in South Africa’, South African Journal of Science122(3/4), 18084. https://doi.org/10.17159/sajs.2026/18084Purkarthofer, J. & Mossakowski, J., 2011, ‘Bilingual teaching for multilingual students? Innovative dual-medium models in Slovene–German schools in Austria’, International Review of Education57(5–6), 551–565. https://doi.org/10.1007/s11159-011-9255-3Ramafikeng, M.C., 2024, ‘Pedagogical translanguaging as a socially just strategy for multilingual students in occupational therapy’, Education Sciences14(5), 462. https://doi.org/10.3390/educsci14050462Romm, N.R.A., 2018, Responsible research practice: Revisiting transformative paradigm in social research, Springer, Cham.Rosekrans, K., Sherris, A. & Chatry-Komarek, M., 2012, ‘Education reform for the expansion of mother-tongue education in Ghana’, International Review of Education58(5), 593–618. https://doi.org/10.1007/s11159-012-9312-6Schoonen, A. & Botes, M., 2025, ‘Conceptualising an interdisciplinary approach to develop contextualised STEM resources for multilingual early childhood classrooms in South Africa: A socially constructivist approach to science literacy’, in INTED2025 Proceedings, Valencia, March 03–05, 2025, pp. 4228–4236.Schoonen, A. & Motaung, N., 2025, ‘Culturally responsive science pedagogy in multilingual early childhood classrooms: A bibliometric review of global trends and insights’, in EDULEARN25 proceedings, Palma, June 30 – July 02, 2025, pp. 7566–7573.Sibanda, R., 2019, ‘Mother-tongue education in a multilingual township: Possibilities for recognising lok’shin lingua in South Africa’, Reading & Writing10(1), 1–10. https://doi.org/10.4102/rw.v10i1.225Sibanda, R. & Tshehla, L.P., 2025, ‘From mother tongue to English: A language policy shift at a multilingual township school in Gauteng’, South African Journal of Childhood Education15(1), a1598. https://doi.org/10.4102/sajce.v15i1.1598Van Eck, N.J. & Waltman, L., 2010, ‘Software survey: VOSviewer, a computer program for bibliometric mapping’, Scientometrics84(2), 523–538. https://doi.org/10.1007/s11192-009-0146-3Vygotsky, L.S., [1941] 1997, The collected works of L.S. Vygotsky. Vol. 4: The history of the development of higher mental functions, R.W.Rieber (ed.), transl. M.J.Hall, Plenum Press, New York, NY.Vygotsky, L.S., 1978, Mind in society: The development of higher psychological processes, M.Cole, V.John-Steiner, S.Scribner & E.Souberman (eds.), Harvard University Press, Cambridge, MA.Wittgenstein, L., 1922, Tractatus logico-philosophicus, transl. C.K.Ogden, Routledge & Kegan Paul, London.Wright, W.E. & Baker, C., 2025, Foundations of bilingual education and bilingualism, 8th edn., Multilingual Matters, Bristol.
How to cite this article: Schoonen, A. & Lesabe, I., 2026, ‘Beyond the limits of language: An additive bilingual approach to Foundation Phase education as a social justice imperative’, South African Journal of Childhood Education 16(1), a1910. https://doi.org/10.4102/sajce.v16i1.1910
Note: The manuscript is a contribution to the themed collection titled ‘Teaching and learning in the context of Mother Tongue-Based Bilingual Education’, under the expert guidance of guest editors Prof. Anthony Essien, Prof. Jabulani Sibanda and Dr Ingrid Sapire.