Review Article

Research on mother tongue-based bilingual education in mathematics, science and technology education across Africa

Shemunyenge Taleiko Hamukwaya, Anthony A. Essien

Received: 16 Feb. 2026; Accepted: 11 June 2026; Published: 20 Aug. 2026

Copyright: © 2026. The Authors. Licensee: AOSIS.
This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license (https://creativecommons.org/licenses/by/4.0/).

Abstract

Background: With the increased awareness of the role of language in enabling epistemic access in mathematics, science and technology education comes the importance of multilingual models that value the indigenous languages as important resources for meaning-making and knowledge construction. It is within this context that mother tongue-based bilingual education (MTBBE) was introduced in South Africa.

Aim: This scoping review synthesises research outputs published between 2014 and 2025 on MTBBE in Africa at the primary level in mathematics, science and technology education.

Setting: The surveyed research outputs were sourced from publications on MTBBE on the African continent from the years 2014 to 2025.

Methods: We first reviewed domain-specific journals and conducted additional searches in general education journals, including book chapters published by key academic publishers, using Google Scholar. A total of 45 research outputs met the inclusion criteria.

Results: Research outputs indicate a substantial increase in MTBBE research in mathematics, science and technology (MST) primary education across Africa in the last 6 years, with a strong geographic concentration in South Africa. Translanguaging practices and teacher preparation emerged as dominant themes. No research output on MTBBE in technology education was found in the review period.

Conclusion: The review indicates that the effectiveness of MTBBE is contingent upon the strengthening of teacher preparation and the establishment of greater coherence between language-in-education policy and classroom-level implementation practices.

Contribution: This review article contributes to the literature by providing the first comprehensive synthesis of research outputs on MTBBE specifically within MST education across African contexts.

Keywords: mother tongue-based bilingual education; multilingualism in Africa; teacher education; translanguaging; mathematics, science and technology education; primary education level.

Introduction

Learners’ achievement in mathematics, science and technology (MST) across many African primary classrooms is influenced by issues related to the use of a language of instruction which is often not the learners’ first language. In science, technology, engineering, and mathematics (STEM) subjects, learners are expected to interpret vocabulary and symbolic representations to develop their conceptual understanding. Where instruction is delivered in a language unfamiliar to learners, evidence from the literature reveals limited comprehension and participation, which pose challenges to MST subjects in terms of learning and understanding (see Banda 2017; Benson 2022; Planas et al. 2025; Probyn 2019). To address this challenge, UNESCO (2025) emphasises the importance of multilingual education models that recognise and value indigenous or local languages as a medium of instruction. In response, many African countries have initiated or strengthened the shift to Mother Tongue-based Bilingual Education (MTBBE) as an educational reform. For example, in South Africa, the Department of Basic Education (DBE) has prioritised MST subjects for the first phase of MTBBE rollout (DBE 2025). This reflects a growing need for subject-specific evidence to support MTBBE implementation (see Ngubane & Ngwenya 2025). Similarly, several other African countries, including Zambia, Malawi and Kenya, have adopted or piloted similar bilingual education shifts to enhance learning outcomes in MST disciplines (see e.g. Mose 2019; Nkonde et al. 2018).

Mother tongue-based bilingual education is a pedagogical model in which learners’ first language is employed as the principal medium of instruction during the initial phases of formal schooling, alongside the structured and progressive introduction of a second or additional language as a complementary medium of instruction. By its very design, MTBBE presupposes the sustained interaction of two or more languages within the instructional space. Although early formulations of MTBBE often conceptualised languages as functionally compartmentalised (where the learners’ first language serves as the primary medium of instruction while the additional language is introduced in a gradual and structured manner), the practical realities of multilingual classrooms rarely conform to such rigid separation. Instead, classroom discourse frequently reflects dynamic and fluid movement across linguistic codes as teachers and learners draw on their full linguistic repertoires to negotiate meaning, clarify concepts and mediate disciplinary knowledge. In this regard, MTBBE inherently entails forms of language mixing that extend beyond simplistic code-switching, particularly in contexts where learners’ linguistic resources are deeply intertwined. This pedagogical reality has led to the increasing association of MTBBE with translanguaging pedagogy whereby teachers and learners strategically draw on their full linguistic repertoires to facilitate meaning-making, enhance learning and promote active participation. Translanguaging reconceptualises bilingualism (when two languages are involved as media of instruction) not as the alternation between autonomous linguistic systems, but as the strategic and purposeful deployment of an integrated linguistic repertoire to enhance comprehension and participation. Within MTBBE settings, translanguaging practices enable learners to mobilise their first language to access complex academic content while simultaneously developing proficiency in the additional language.

The purpose of this scoping review is to provide an overview of research that has been conducted on MTBBE in the teaching and learning of mathematics, science and technology education on the African continent at the primary education level in the last approximately one decade (2014–2025). Several compelling scholarly and contextual considerations justify a review study of this nature: first, the past decade has been marked by intensified policy attention to foundational literacy and numeracy, as reflected in continental and global reform agendas in many African countries. A case in point is the introduction of MTBBE in South Africa. Within this reform climate, the language of learning, teaching and assessment (LoLTA) has re-emerged as a central determinant of epistemic access in MST subjects. A review of this kind is warranted to consolidate and critically interrogate the growing body of empirical work examining how MTBBE shapes conceptual understanding, classroom discourse and learner achievement in these domains; secondly, multilingualism remains a structural feature of most African education systems, where the language of instruction frequently diverges from learners’ home languages. Over the last decade, there has been a proliferation of interventions, pilot programmes and quasi-experimental studies examining the cognitive, linguistic and pedagogical implications of MTBBE. However, this research is often fragmented across countries, disciplines (mathematics education, science education, language education) and publication outlets. A scoping review would map and synthesise the existing MTBBE research by identifying patterns, methodological trends, theoretical frameworks and reported outcomes, thereby reducing knowledge silos; third, a review such as this would identify research gaps – geographical (under-represented regions), methodological (limited longitudinal or large-scale studies) and thematic (e.g. technology education, teacher professional development and assessment practices). By mapping the contours of existing scholarship, the study would provide a foundation for future empirical investigations and more contextually responsive language-in-education policies.

Finally, the decision to conduct this review study was further motivated by limitations identified in previous reviews of multilingual education in Africa, which have either focused on early literacy interventions (Piper, Zuilkowski & Ong’ele 2016), on policy frameworks (Heugh 2011), on multilingualism generally (Essien & Adler 2025; Essien & Moleko 2025) or on translanguaging practices in general education contexts (Ngubane & Ngwenya 2025; Probyn 2015). The present scoping review addresses these gaps by providing the first comprehensive synthesis examining MTBBE implementation specifically within MST education across multiple African countries between 2014 and 2025. In so doing, this review provides a more holistic and STEM-specific understanding of MTBBE implementation than previous reviews. Given the conceptual and linguistic complexity of MST learning at the primary level, a dedicated review of existing research focusing specifically on MTBBE in MST education is necessary to consolidate existing knowledge, identify gaps and inform future research and policy development. To this end, the current review study is guided by the following questions:

• What is the current state of research on MTBBE in the teaching and learning of mathematics, science and technology at the primary education level in Africa?
• How is translanguaging positioned in literature as a pedagogical strategy for facilitating meaning-making in MST classrooms within MTBBE?
• How do recent studies describe teacher preparedness and the role of in-service training in enabling MTBBE implementation in MST?
• What practical challenges are reported in African classrooms, and how can these challenges inform policy, curriculum design and teacher education to strengthen MTBBE across the continent?

To address these questions, we conducted a review of research articles published in mathematics, science and technology education journals and broader education journals within the African continent as well as in leading international journals. The specific inclusion and selection criteria guiding this review are elaborated in the methodology section.

Teaching and learning of mathematics, science and technology in multilingual classrooms

The UNESCO (2025:66) Report argues that ‘to improve education quality, the challenge is to promote multilingual education models that value indigenous and local languages as curricula languages’. Our conceptualisation of multilingual classrooms resonates with that of Barwell (2009), who uses the term to refer to classrooms in which more than one language is present and where these languages constitute potential resources for teaching and learning. From this perspective, multilingualism encompasses a range of linguistic configurations, including bilingualism. Accordingly, we regard bilingualism as a specific manifestation of multilingualism, characterised by the presence and use of two languages within educational contexts. This understanding acknowledges the linguistic diversity of classrooms and recognises the pedagogical potential of learners’ multiple language resources in supporting meaning-making and knowledge construction.

Research has shown that when home languages are used in learning mathematics and science, this facilitates both procedural and conceptual understanding (see Ndhlovana & Charamba 2023; Planas 2018; Robertson & Graven 2019). This is especially true when introducing new concepts. A growing body of literature in African classrooms shows that teachers and learners in multilingual classrooms consistently translanguage to communicate and co-construct knowledge. South African studies document that such pedagogical practice in mathematics and science enables learners to reason, clarify terminology and connect procedural steps to conceptual ideas (Maluleke 2019; Ndhlovana & Charamba 2023; Robertson & Graven 2020). Research from Ghana, Nigeria and Kenya finds comparable benefits in relation to this practice. An experimental and quasi-experimental study in Ghana reported improved arithmetic performance when Twi (home language) was used alongside English to teach mathematics as opposed to classrooms in their study where only one language (either English or Twi) was used as LoLTA (Abdul-Ganiyu et al. 2024). Furthermore, in Nigeria and Kenya, recent empirical work documents the role of translanguaging in increasing engagement and understanding in primary classrooms, while also noting that this teacher practice is often ad hoc rather than policy-driven (see Mose 2019; Nkonde et al. 2018). Across these diverse contexts, translanguaging emerges as an effective classroom-level pedagogical practice in MST education when it is implemented intentionally and aligned with clearly articulated learning objectives. The literature consistently indicates that its efficacy is contingent upon several interrelated factors, including teachers’ pedagogical and linguistic competence, the degree of curricular flexibility afforded within the instructional framework, and assessment regimes that do not constrain or penalise learners’ strategic use of their home languages as resources for meaning-making.

With regard to MTBBE, research from South Africa (e.g. Ngubane & Ngwenya 2025; Olawale, Hendricks & Rusi 2024) has emphasised that effective implementation of MTBBE requires systematic reforms in teacher education and ongoing professional development focusing on: (1) subject-specific bilingual pedagogies (how to teach mathematics and science bilingually); (2) development of subject terminology in local languages; and (3) strategies for planning and assessment in bilingual classrooms. This follows from earlier research from Tanzania and Zambia, which indicates that teachers often lack structured training in translanguaging pedagogy and report inadequate knowledge of bilingual MST education during professional development (see e.g. Mutagwaba & Jessner 2024; Nkonde et al. 2018). Drawing insights from these studies, it can be argued that to strengthen MTBBE in teacher education programmes, there is a need to prioritise and strengthen multilingual pedagogies in teacher development programmes so that these programmes equip teachers to purposefully use more than one language to teach for meaning-making.

Practical constraints towards MTBBE implementation feature in the literature. For example, shortages of teaching and learning materials in local languages, poorly developed subject terminology, monolingual national assessments, language hierarchies that privilege English and weak policy-to-practice coherence (Chimbutane 2015; Sanfo & Ogawa 2025; Sibanda & Tshehla 2025). Empirical evidence from the Eastern Cape and other regions of South Africa highlights the scarcity of teaching and learning materials, alongside persistent assessment misalignment, as significant structural impediments to the effective delivery of MST education through learners’ home languages (Sibanda & Tshehla 2025). Similar barriers are also evident in studies emanating from Ghana, Nigeria and Kenya where institutional and legacy assessment frameworks discourage sustained mother-tongue use in MST classrooms (see Mose 2019; Opoku-Amankwa, Edu-Buandoh & Brew-Hammond 2015). These point to the need for systemic supports in terms of teaching and learning materials and assessments that are adapted to multilingual pedagogies.

Methods

This study adopted a scoping review to map existing research on MTBBE in MST education across Africa in order to achieve its purpose.

Inclusion and exclusion criteria

We first reviewed domain-specific journals focusing on the top 20 journals in mathematics education (Williams & Leatham 2017), the top 17 journals in science education (https://guides.library.stanford.edu/science_ed/journals) and the top 15 technology education journals (https://guides.library.stanford.edu/tech_and_ed/journals) provided by Stanford University libraries. To ensure broader coverage, we also conducted additional searches in general education journals using Google Scholar. In addition, in our search, we also included book chapters published by key academic publishers such as Bloomsbury, Springer and SENSE Publishers. Studies were eligible for inclusion if they simultaneously met the following criteria:

• peer-reviewed research published from 2014 to 2025
• research conducted within African primary education contexts, regardless of the authors’ geographical location, and addressing issues related to teaching and learning within the three domains: mathematics, science and technology
• research explicitly focused on the implementation of MTBBE

Studies were excluded if they: (1) focused exclusively on literacy or language education without reference to MST education; (2) were conducted outside Africa; (3) focused on secondary or higher education contexts; (4) dealt solely with mother tongue instruction without a bilingual or multilingual dimension; or (5) were theses, dissertations, conference proceedings or papers published in predatory journals.

Searches were conducted between 28 January 2026 and 31 January 2026. The following combinations of search terms were employed: ‘“mother tongue-based bilingual education”, “mathematics”’; ‘“bilingual education”, “mathematics”’; ‘“multilingual education”, “mathematics”’; ‘“translanguaging”, “mathematics”’; and ‘“code-switching”, “mathematics”’. We repeated these combinations of search terms with the other two discipline domains – that is, with science and technology after each term. In addition, we undertook a supplementary search procedure (spin-off search) comprising: (1) a systematic examination of the reference lists of the initially selected articles to identify further relevant studies; and (2) a targeted search of publications authored by leading scholars on the African continent within the field. In total, the whole search yielded a total of 4668 records. After removing duplicate entries (n = 157), 4511 records remained for further screening. The screening process was conducted in two stages. First, titles and abstracts were screened against the inclusion criteria, which specified that studies must: (1) be empirical studies, conceptual papers or reviews; (2) address the implementation, teaching practice or policy related to MTBBE in mathematics, science and technology education; (3) be conducted within an African context; and (4) be published between 2014 and 2025. Based on these criteria, 3675 records were excluded. Following this initial screening, 836 full-text articles were assessed for eligibility. After applying the inclusion and exclusion criteria at the full-text level, 45 studies met the final inclusion criteria and were included in the review study. Figure 1, the Preferred Reporting Items for Systematic reviews (PRISMA) flow, summarises the process of identification and screening of the studies reviewed in this study (2014–2025).

FIGURE 1: Preferred Reporting Items for Systematic reviews (PRISMA) showing the article search and screening process.

Quality assessment

To ensure the methodological quality of the included research outputs, for each, key characteristics were recorded, including year of publication, title, author(s), country of study, participant population, research approach (qualitative, quantitative, or mixed methods), research focus and disciplinary focus (mathematics, science, technology, or cross-disciplinary STEM). The extracted data were analysed using a thematic analysis approach informed by the procedures outlined by Braun and Clarke (2019). Initially, each author independently coded all included studies to identify salient patterns and areas of emphasis. The authors subsequently met to compare and discuss their coding, resolve discrepancies and reach consensus on the interpretation of the data. Through an iterative process that combined both deductive coding, informed by the review objectives, and inductive coding, grounded in the data themselves, the coding framework was progressively refined. Ongoing critical discussions among he authors facilitated the consolidation and refinement of categories and themes. This process resulted in the identification of five thematic categories, as presented in Table 4. These thematic categories enabled the identification of dominant research trends. We individually coded all the articles into these themes and then met to compare our coding. We then refined our coding of each paper after reaching a consensus. In Figure 1, we present details of the article search and screening process.

Research trends in mother tongue-based bilingual education in mathematics, science and technology in Africa

Publication by year

Table 1 indicates that in our review period, research on MTBBE in primary MST education within African contexts began to emerge from 2017 onwards, with a notably more rapid growth from 2019, peaking in 2025 (n = 12). Many of the reviewed studies were published in African journals (n = 32). This distribution highlights the central role of African-based publication platforms in advancing contextually grounded research on MTBBE.

Table 1 further indicates a limited representation of publications originating from edited book chapters. The majority of the identified studies focus broadly on mother tongue education rather than explicitly on MTBBE. However, given the growing scholarly and policy attention directed towards MTBBE, it is reasonable to anticipate a shift in this distribution in the coming years. In the next section, we show the geographical distribution of these research outputs. It must be noted that some outputs are studies that map MTBBE across the continent. Given their transnational scope, these studies were not attributed to any specific country in the analysis.

Geographic distribution across Africa

While MTBBE research at the primary level is geographically diverse, it remains limited, with significant concentrations in Anglophone contexts and with limited representation from Francophone and other regions, as shown in Figure 2. Specifically, there is a strong concentration in South Africa (n = 26), reflecting the country’s sustained policy engagement and research focus on multilingual education. Importantly, sixteen of these studies were published before the formal introduction of MTBBE in 2025. Additional representation of two studies is also observed in Tanzania, Zambia and Nigeria and single studies from Malawi, Namibia, Burkina Faso, Rwanda, Senegal and Kenya. The geographical distribution of studies reveals a substantial imbalance. Hence, the findings of this scoping review should be interpreted with caution, as they largely reflect evidence from South African rather than continental patterns. The limited representation of other African regions highlights an important gap in the literature and suggests the need for more geographically diverse research.

FIGURE 2: Geographic distribution of included studies.

As indicated previously, seven studies (not represented in Figure 2) focus on cross-country or continental perspectives, examining MTBBE across multiple African contexts – mostly review papers.

Publication by knowledge domains

Table 2 presents the total number of surveyed research outputs by knowledge domains. It highlights mathematics education (n = 25) as the primary focus for MTBBE research in African contexts while also indicating emerging attention to interdisciplinary MST integration; for example, the integration of mathematics and science education (n = 12). A smaller number of studies focus exclusively on science education (n = 6), while two studies found an integrated domain of mathematics, science and technology education.

Notably, our review did not find studies that focused exclusively on MTBBE in technology education.

Publication by methodological approaches

Our findings indicate a strong predominance of qualitative research approaches (n = 38), compared to few quantitative (n = 2) and mixed-methods studies (n = 4). As summarised in Table 3, the dominance of qualitative and empirical approaches (n = 22) reflects the context-sensitive nature of MTBBE research at the primary level, which often seeks to understand classroom practices, teacher experiences and policy implementation within multilingual educational settings.

Fundamentally, there is a pressing need for rigorous quantitative studies that generate robust empirical and statistical evidence regarding the effectiveness of MTBBE. Such studies would need to employ methodologically sound research designs, clearly defined outcome measures and sufficiently large and representative samples to permit valid inferences regarding the impact of MTBBE on learners’ academic achievement and related educational outcomes.

Review findings

As indicated previously, two levels of iterative processes (inductive and deductive approaches) were used in our attempt to organise the reviewed articles into themes. This procedure culminated in the identification of five overarching themes, which are detailed in Table 4.

As shown in Table 4, translanguaging practices emerged as the most common research focus (n = 22) between 2014 and 2025. A substantial body of qualitative research demonstrates how translanguaging enables learners to access disciplinary concepts, engage in disciplinary reasoning and participate meaningfully in classroom discourse (Charamba 2022; Kazima et al. 2023; Ndhlovana & Charamba 2023; Ramothwala et al. 2025). These studies demonstrate that learners’ familiar linguistic repertoires serve as communicative resources that support conceptual mediation and engagement. In what follows, we engage further with these themes.

Translanguaging practices in mathematics, science and technology education

Translanguaging emerges consistently (22 research outputs) as a central pedagogical approach through which learners construct meaning and navigate multilingual MST learning environments. Across diverse African contexts, classroom-based studies demonstrate that when teachers and learners fluidly use multiple linguistic resources, learners are better able to access abstract concepts, articulate reasoning and engage in collaborative problem-solving (Charamba 2022; Kazima et al. 2023; Makonye 2019; Ramothwala et al. 2025). For example, Kazima et al. (2023) show how the use of local languages enabled Malawian learners to develop conceptual understanding of zero, a mathematically abstract concept, by connecting symbolic representations to familiar linguistic meanings. Similarly, Makonye (2019) found that translanguaging practices supported deeper understanding of geometric concepts by enabling learners to articulate reasoning processes using familiar linguistic structures. Translanguaging also plays a critical role in mediating epistemic access by enabling learners to bridge everyday and formal language (Essien 2025; Essien & Adler 2025; Maphefo & Awelani 2025). While these studies report enhanced conceptual understanding through strategic use of local languages, these findings are largely derived from small-scale classroom-based investigations involving limited numbers of learners and teachers. Although such studies provide rich insights into learner meaning-making, they offer limited evidence regarding the broader impact of translanguaging on learning across diverse African educational contexts. Moreover, the predominance of qualitative case-study designs within this body of literature also raises important questions regarding the dearth of research on translanguaging whose aim is the generalisability of its findings.

Furthermore, the literature reveals considerable conceptual differences in how translanguaging is understood and operationalised to support disciplinary learning. In science classrooms, for example, Charamba (2022) demonstrated that translanguaging facilitated conceptual understanding by enabling learners to engage meaningfully with scientific discourse. His use of the term ‘translanguaging’ to denote the mixing of ‘English language and African languages [ ] in roughly equal proportions with the objective of fostering full bilingualism and biliteracy with a focus on science education’ (p. 31) is particularly different from other studies (e.g. Makgabo & Modise 2020; Maluleke 2019) which use the term interchangeably with code-switching or language alternation with no allocated language proportionality. This conceptual inconsistency makes it difficult to determine whether reported translanguaging practices are attributed to pedagogically translanguaging practices or to spontaneous forms of language switching occurring in multilingual classrooms. As we have argued elsewhere (Essien, Hamukwaya & Takane in press), while at the academic or intellectual level, it may be important to distinguish between language mixing practices like code-switching and translanguaging, at the practical level, what is key is how language mixing – whether we call it code-switching or translanguaging – is used to enable epistemic access.

Several studies caution that the success of translanguaging practices depends on teachers’ pedagogical competence and intentional use of linguistic resources (Malindi et al. 2023; Maluleke 2019; Robertson & Graven 2019). In contexts where teachers lack training or confidence for multilingual teaching, translanguaging practices often appear to be ad hoc rather than strategic. Consequently, translanguaging does not systematically support conceptual learning. While translanguaging may facilitate learners’ participation and comprehension during instruction, institutional constraints such as assessment policies privileging English were reported to limit the effectiveness of multilingual classroom practices (Makgabo & Modise 2020). This may create a contradiction in which learners are encouraged to utilise multiple linguistic resources during learning but are often expected to demonstrate knowledge through a single dominant language in formal assessment contexts. Such tensions suggest that the benefits of translanguaging may be constrained by broader policy and assessment that remain largely monolingual in orientation.

Importantly, although translanguaging is associated with improved participation and conceptual access, less evidence exists in this review regarding its long-term effects. Only a small proportion of studies provide quantitative evidence linking multilingual pedagogies to measurable learning gains (Nchindila 2017; Sanfo & Ogawa 2025), while longitudinal investigations remain absent. As a result, claims in this review regarding translanguaging practices are supported primarily by evidence of improved classroom interactions and learner engagement. The reviewed studies suggest that translanguaging holds considerable promise as a pedagogical resource for supporting epistemic access in multilingual MST classrooms. Nevertheless, the evidence base remains dominated by small-scale qualitative studies concentrated primarily in South African contexts.

Teacher preparation and pedagogical strategies

Teacher capacity emerged as a pivotal determinant in shaping the quality, coherence and overall effectiveness of MTBBE implementation. Across the reviewed literature, many studies report that teachers often lack formal training to teach through bilingual or multilingual pedagogies and experience uncertainty regarding how to integrate multilingual strategies into mathematics and science instruction (Cekiso et al. 2019; Malindi et al. 2023; Shuukwanyama et al. 2022). This finding was evident across diverse national contexts and educational settings, suggesting that teacher preparedness constitutes a systemic rather than a localised challenge. A notable pattern across the literature is the emphasis on teacher education as a means of strengthening teachers’ capacity to use multilingual pedagogies. Emerging evidence suggests that targeted professional development interventions can significantly improve instructional practice. For example, Olawale et al. (2024) demonstrated that teacher training programmes focusing on bilingual pedagogical strategies enhanced teachers’ ability to support conceptual learning. Similarly, Ramadiro (2022) highlights the importance of integrating multilingual pedagogies into teacher education programmes to prepare teachers for linguistically diverse classrooms. The reported benefits are, however, often inferred from changes in teacher perceptions or short-term interventions rather than from sustained evidence. This raises questions regarding the long-term effectiveness of existing teacher professional development initiatives. Although multilingual pedagogies are promoted within policy documents and reform initiatives, teachers often work within educational systems characterised by English-dominant assessment, limited instructional resources and insufficient institutional support (Malindi et al. 2023; Shuukwanyama et al. 2022). Consequently, teachers are expected to implement MTBBE while simultaneously navigating expectations regarding assessment requirements and language use. This contradiction suggests that improving teacher preparation alone may be insufficient.

Other studies emphasise the importance of locally developed instructional materials and pedagogical tools, such as bilingual glossaries, translated instructional materials and culturally relevant teaching strategies, as some mechanisms for strengthening MTBBE implementation (Benson 2022; Booi et al. 2024). While these resources are regarded as beneficial, only a few studies critically examine how teachers use such materials in classroom settings or whether their use translates into improved learner understanding. The literature therefore tends to assume a positive relationship between resource provision and instructional quality, despite limited empirical evidence demonstrating this relationship. While many studies identify inadequate teacher preparation as a significant barrier to MTBBE implementation, the evidence base reveals important methodological limitations. Much of the literature relies on teachers’ self-reported perceptions (such as interviews) and small-scale case studies, with few studies directly observing classroom practice or evaluating the impact of teacher preparation programmes. Although there is evidence that teachers require additional support, the empirical basis for determining which forms of professional development are most effective to support the implementation of MTBBE remains limited. Importantly, the reviewed studies differ in how they conceptualise teacher competence. Some studies emphasise linguistic proficiency in both the home language and the language of instruction, while others consider pedagogical knowledge, translanguaging strategies or the ability to develop subject-specific terminology in African languages. This variation suggests that teacher preparedness for MTBBE is multidimensional and remains insufficiently theorised within the existing literature. Also, there remains insufficient evidence to determine which MTBBE implementation models of teacher education are most effective across diverse African settings.

Policy and implementation gaps

Despite policy support for MTBBE across African countries, implementation remains often constrained by systemic and institutional barriers. Several studies document significant gaps between policy intentions and classroom realities (Mini & Botha 2020; Ozoemena et al. 2021; Sibanda & Tshehla 2025). Across diverse contexts, policy support for MTBBE does not automatically translate into effective classroom practice. Rather, the reviewed literature suggests that implementation is shaped by multiple factors, including teacher preparedness, resource availability, language ideologies and the broader commitments to multilingual education. This suggests that MTBBE implementation failures should not necessarily be interpreted as evidence of ineffective policy but rather as indicators of weak alignment between policy intentions and institutional support structures.

The literature also reveals tensions regarding language transition policies. While studies support the use of learners’ home language during the early years of schooling, the timing of language transition policies also poses significant challenges to additional languages. Nkonde et al. (2018) found that premature transition from mother tongue to English undermined learners’ conceptual development. Similarly, Sanfo and Ogawa (2025) demonstrated that sustained exposure to a bilingual medium of instruction was associated with improved learning outcomes. However, these conclusions are drawn from studies employing different methodological approaches and from different educational contexts with different outcome measures. While the evidence supports bilingual instruction, the literature provides limited guidance regarding the specific conditions under which language transitions are most effective.

Resource constraints also emerge as another significant challenge that limits the effective implementation of MTBBE. Studies consistently report shortages of instructional materials in African languages, which undermine the institutionalisation of MTBBE (Ball et al. 2024; Mini & Botha 2020). Yet, the literature assumes a direct relationship between the availability of resources and successful implementation without adequately examining how such resources are used in practice. Few studies investigate whether the provision of translated materials leads to improved teaching and learning or how teachers adapt available resources to local classroom realities. This represents an important gap in the literature, particularly given the importance often associated with the development of multilingual educational materials. A further contradiction concerns the role of language ideologies within education systems. While policy frameworks promote multilingual and linguistic inclusion, some studies suggest that dominant societal and institutional preferences for English or French continue to shape classroom practices and educational decision-making (Ozoemena et al. 2021; Sibanda & Tshehla 2025). In this regard, MTBBE implementation challenges are not only technical or resource-related but are also ideological. This suggests that the continued use of colonial languages may undermine teachers’, learners’ and parents’ confidence in MTBBE, even where supportive policies exist. Importantly, much of the evidence-based claim in this report is derived from qualitative case studies. Although these studies provide valuable insights into contextual realities, they offer limited evidence regarding the extent to which identified challenges are representative of broader national or continental patterns. As a result, many conclusions regarding policy and implementation remain context-specific, highlighting the dangers of a one-size-fits-all approach.

Learner outcomes and assessment evidence

Although a growing body of research suggests that MTBBE supports learning, empirical evidence on learner outcomes remains limited. While some studies report a positive relationship between MTBBE and improved academic performance (Nkonde et al. 2018; Sanfo 2022; Sanfo & Ogawa 2025), the overall evidence remains limited in both scope and methodological diversity. These studies provide important evidence that exposure to a bilingual medium of instruction can enhance academic achievement. However, most studies employed small-scale or qualitative designs, limiting generalisability. Recent reviews highlight the need for more longitudinal research to examine long-term outcomes (e.g. Ngubane & Ngwenya 2025). Without such evidence, it remains difficult to assess the systemic impact of MTBBE.

The literature also reveals methodological differences across studies in terms of learners’ outcomes and assessment. For example, some studies examine learner outcomes using diverse outcome measures, ranging from classroom participation and conceptual understanding to standardised test performance and achievement scores. As a result, the term improved learning outcomes is often used to describe different forms of evidence. While some studies report statistically significant gains in learner performance (Sanfo 2022; Sanfo & Ogawa 2025), others infer positive outcomes from qualitative indicators such as increased participation or classroom interaction (Charamba 2022; Robertson & Graven 2019). Another tension concerning assessment practices from the reviewed literature is the argument from some studies that MTBBE supports conceptual understanding by enabling learners to access disciplinary knowledge through familiar linguistic resources, particularly English and French (Essien 2018; Kazima et al. 2023; Sanfo & Ogawa 2025). This creates inconsistency in which learning may occur through multilingual practices while achievements are measured through monolingual assessment frameworks. Therefore, the benefits associated with MTBBE may be underestimated when assessment instruments fail to capture learners’ understanding in the language through which learning occurred. Despite the importance of this issue, only a few studies investigate the relationship between language of instruction, language of assessment and learner performance (Makgabo & Modise 2020; Mtsatse & Combrinck 2018; Sibanda & Tshehla 2025).

The geographical concentration of studies also needs consideration. Much of this available evidence originates from a small number of countries, particularly South Africa, and a limited number of West African contexts. Thus, the reported outcomes may reflect context-specific policy environments rather than broader continental patterns. This raises questions regarding the transferability of findings to under-represented regions, including Francophone, Central and North African contexts.

Decolonial and indigenous knowledge approaches

Some reviewed literature positions MTBBE within broader decolonial debates concerning indigenous languages, knowledge systems in MST education and epistemic justice in African education. These studies argue that African languages provide epistemic resources that support meaningful learning (Babaci-Wilhite 2017; Essien & Moleko 2025; Mhlolo 2024). From this perspective, the use of African languages in MST education challenges the historical privileging of colonial languages and supports culturally responsive approaches to teaching and learning. An argument across the literature under this theme is that indigenous languages may provide conceptual understanding that facilitates learners’ understanding of disciplinary concepts. Larkin et al. (2025), for example, demonstrate how indigenous languages provide conceptual transparency that supports mathematical understanding of the place value system. These findings align with decolonial perspectives that emphasise the importance of linguistic and epistemic diversity in education. MTBBE thus serves not only as a pedagogical strategy but also as a means of promoting epistemic justice and educational equity. These studies suggest that integrating indigenous languages and knowledge systems into MST education can enhance learner conceptual understanding while challenging dominant monolingual approaches to teaching and learning, thereby strengthening the implementation and the potential of MTBBE in multilingual African contexts.

It is important to state that much of the decolonial and indigenous knowledge literature related to MTBBE relies on conceptual analyses, theoretical arguments and policy critiques. Although such literature contributes theoretical insights, it offers limited empirical evidence regarding implementation processes and educational outcomes. Therefore, future research should move beyond theory and examine how decoloniality and indigenous knowledge systems within MTBBE can lend epistemic access in MST classrooms.

Implications and recommendations

Despite growing evidence supporting the pedagogical value of MTBBE, the overall evidence base remains a challenge. Most studies employed qualitative or small-scale designs, limiting the ability to draw conclusions about system-wide impact. As noted in recent systematic reviews (see Essien & Adler, 2025), significant gaps remain, including limited research in Central and North Africa, a lack of longitudinal studies and limited quantitative evidence on learner outcomes. At the same time, a growing body of research adopts decolonial and indigenous knowledge perspectives, emphasising the epistemic legitimacy of African languages in STEM education (Babaci-Wilhite 2017; Essien & Moleko 2025; Mhlolo 2024; Larkin et al. 2025). These studies challenge deficit views of African languages and argue that MTBBE supports epistemic access by enabling learners to engage with disciplinary knowledge through culturally and linguistically meaningful frameworks.

Significantly, this review provides rich insight into classroom-level practice, but less evidence regarding systemic long-term learning outcomes. This imbalance reflects broader methodological trends in the field. Importantly, conducting this review, we found that the research on MTBBE in Africa has grown significantly, providing strong evidence of its pedagogical value. However, systemic barriers and methodological limitations constrain implementation and evaluation. Therefore, future research should prioritise examining the systemic impact of MTBBE beyond classroom practice, including its influence on educational equity, language policy implementation, teacher capacity, assessment systems, and sustained learner achievement. Such research should move beyond short-term intervention studies to include longitudinal, comparative, and large-scale evaluations capable of informing evidence-based educational reform across the continent.

Conclusion

In summary, this review study makes a significant contribution to the literature by providing the first comprehensive synthesis of MTBBE implementation specifically within MST education across African contexts. By extending beyond literacy-focused reviews and policy-focused analyses, the review provides new insights into how MTBBE supports conceptual learning, epistemic access and pedagogical transformation in mathematics and science education. In addition, by integrating decolonial and indigenous knowledge perspectives, this review highlights the broader epistemic and sociocultural significance of MTBBE for advancing educational equity and transformation in Africa. But despite the strength of the study, this review was limited to peer-reviewed publications and selected academic sources published between 2014 and 2025, which may have excluded relevant studies in grey literature.

Acknowledgements

The authors reported that they received funding from the National Research Foundation South African Numeracy Chair at the University of the Witwatersrand (Wits), which may be affected by the research reported in the enclosed publication. The authors have disclosed those interests fully and have implemented an approved plan for managing any potential conflicts arising from their involvement. The terms of these funding arrangements have been reviewed and approved by the affiliated University in accordance with its policy on objectivity in research. The author, Anthony A. Essien, serves as an editorial board member of this journal. The peer review process for this submission was handled independently, and the author had no involvement in the editorial decision-making process for this article. The authors have no other competing interests to declare.

Shemunyenge Taleiko Hamukwaya: Conceptualisation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing. Anthony A. Essien: Conceptualisation, Formal analysis, Methodology, 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.

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

This work is based on research supported by the National Research Foundation South African Numeracy Chair at Wits (grant number 74703). Any opinion, finding, conclusion or recommendation expressed in this material is that of the authors, and the National Research Foundation does not accept any liability in this regard.

The authors declare that all data that support this research article and findings are available in the article and its references.

The views and opinions expressed in this article are those of the authors and are the product of professional research. They do 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.

References