Language of instruction policies define the language used to teach reading, mathematics, science, social studies, and other school subjects. They also decide when a second or international language enters the classroom, how long home-language teaching continues, which language appears in textbooks and exams, and how schools support learners who do not use the school language at home. In global education data, this is not a small administrative choice. UNESCO reports that about 40% of learners worldwide do not receive education in the language they speak and understand best, and in some low- and middle-income countries the share rises to 90%. More than a quarter of a billion learners are affected by this gap.[a][b]
A school system may choose a teaching language for learning reasons, legal reasons, administrative reasons, teacher-supply reasons, assessment reasons, or family-demand reasons. The best policy is rarely a single-language rule. It is usually a sequenced language model: a familiar language in the early years, planned oral exposure to additional languages, later subject teaching in one or more wider-use languages, and assessments that do not confuse subject knowledge with language barriers.
Data Lens: Language of instruction sits between curriculum, teacher deployment, literacy policy, assessment design, migration, minority-language rights, and global labour-market preparation. A child can attend school every day and still lose learning time if the teacher, textbook, and exam operate in a language the child has not yet mastered.
What Language of Instruction Means in School Systems
The language of instruction, often called medium of instruction, is the language used to deliver curriculum content. It is broader than a language subject. English, French, Arabic, Spanish, Mandarin, Portuguese, Hindi, Swahili, Turkish, German, Japanese, Korean, or any local language may be taught as a subject. The policy question is different: which language carries the lesson itself?
In early grades, the issue connects directly to foundational literacy and numeracy. Children must learn letter-sound relationships, vocabulary, number concepts, classroom routines, and subject words at the same time. If the teaching language is unfamiliar, the student must decode both the lesson and the language channel. That creates a double load. The effect appears most clearly in reading and mathematics because both depend on precise comprehension.
UNICEF’s Foundational Literacy and Numeracy Hub states that teaching children in languages they understand is central to improving foundational literacy and numeracy, and that language-of-instruction issues should be placed in education policy dialogue rather than treated as a side matter.[c]
Language Subject vs Teaching Language
| Policy Area | Main Question | School-Level Example | Measurement Risk |
|---|---|---|---|
| Language as a subject | Which languages should students learn? | English, French, Arabic, Spanish, Mandarin, or national languages taught in weekly lessons | Test scores measure language course outcomes only |
| Language of instruction | Which language delivers mathematics, science, reading, and social studies? | Mathematics taught in the home language in Grades 1–3, then gradually bilingual | Low scores may reflect language access rather than subject ability |
| Assessment language | Which language appears in exams? | Reading test in the national language, science test in English, oral assessment in a local language | Exam language can understate real knowledge |
| School support language | How do teachers help learners bridge languages? | Glossaries, bilingual reading books, oral scaffolding, language-rich math lessons | Weak support can widen early-grade gaps |
The Main Models Schools Use
Countries and school networks usually combine several models. The label may look simple on paper, but the classroom version depends on teacher skills, textbooks, exams, parent expectations, and the number of languages spoken in one class. A language policy works like a railway switch: a small decision early in schooling can send later learning onto a different track.
Home-Language or Mother-Tongue Instruction
Home-language instruction uses the learner’s strongest language for early teaching. The goal is not to isolate the child from national or international languages. The goal is to build reading, reasoning, and subject concepts first in a familiar language, then transfer those skills into additional languages through planned teaching.
The World Bank’s language-of-instruction policy approach recommends teaching children in their first language from early childhood through at least the first six years of primary schooling, using that language for academic subjects beyond reading and writing, introducing a second language first with oral language focus, and continuing first-language support after a second language becomes the main teaching language.[d]
National-Language Instruction
National-language instruction uses an official or widely shared national language as the main teaching language. It can support system coherence because textbooks, teacher training, exams, and administration operate in one shared language. The problem appears when many children enter school without oral fluency in that language. The policy may look efficient for the ministry but difficult for the first-grade classroom.
This model often appears in countries with many local languages. It can work better when schools provide bridging materials, oral-language preparation, bilingual teaching aides, community-language support, and extra time before high-stakes exams in the national language.
International-Language Instruction
International-language instruction uses a language such as English, French, Spanish, Portuguese, or Arabic as the medium for wider academic and economic access. Families often associate this model with higher education, mobility, and professional opportunity. That demand is real. Yet early full immersion in an unfamiliar language can reduce comprehension if teachers and students do not have enough proficiency.
The policy challenge is sequencing. Students can gain strong international-language proficiency without losing early comprehension if schools introduce the language as a subject first, build oral fluency, then use it for selected subjects after learners have enough academic vocabulary.
Bilingual and Multilingual Instruction
Bilingual instruction uses two languages for teaching. Multilingual instruction uses more than two languages across school stages, subjects, or regions. The strongest designs are not casual code-switching. They assign clear functions: one language for first literacy, another for wider communication, and sometimes a third for regional or international mobility.
UNESCO’s 2025 guidance on multilingual education describes migration, technology, and wider recognition of multilingualism as forces changing classroom language needs. It presents multilingual education as a practical policy route for systems where learners arrive with different language backgrounds.[e]
Global Data Behind the Policy Choice
The strongest data pattern is simple: language access and learning access are linked. Not every low score comes from language mismatch, and not every language policy fixes learning by itself. Still, large-scale sources show that unfamiliar instruction language can make early literacy, classroom participation, subject vocabulary, and assessment results harder to interpret.
| Data Point | Reported Figure | Why It Matters for Teaching Language | Source Note |
|---|---|---|---|
| Learners not taught in the language they understand best | About 40% globally; up to 90% in some low- and middle-income countries | Shows that language mismatch is a large school-access issue, not a rare classroom case | UNESCO language education data[b] |
| Learners affected | More than 250 million | Indicates scale for textbook production, teacher training, and assessment planning | UNESCO[b] |
| Countries with low first-language instruction coverage | In one fifth of countries with data, less than half of early-grade children are instructed in their first or home language | Shows a data gap and an implementation gap in early grades | UNESCO Institute for Statistics Data Browser[f] |
| Learning poverty in low- and middle-income countries | 70% of 10-year-olds estimated unable to read and understand a simple written text after the pandemic shock | Language policy is one part of the wider literacy recovery agenda | World Bank, UNESCO, UNICEF, FCDO, USAID, and Gates Foundation report release[g] |
| Immigrant students and home language in PISA 2022 | Across OECD countries, 62% of first-generation immigrant students and 44% of second-generation immigrant students mainly spoke another language at home | Shows why host-language support matters even in high-income systems | OECD PISA 2022[h] |
The World Bank defines learning poverty as being unable to read and understand a simple text by age 10. The measure combines schooling deprivation and reading proficiency. This matters because a language-of-instruction policy can affect both: children may stay enrolled yet fail to reach reading fluency if early teaching is not comprehensible.[i]
How Schools Decide the Teaching Language
Education ministries rarely choose a teaching language from one variable. The decision usually combines student language profiles, teacher availability, curriculum materials, exam design, legal status of languages, local administration, and the language used in later schooling or employment. The real question is not “Which language is best?” It is: Which language sequence gives students the most access to learning at each stage?
Student Language Profile
The first variable is the language students understand before formal schooling. In a linguistically mixed classroom, a single home language may not cover everyone. Schools then need mapping: which languages children speak, which languages they understand orally, which languages families use for reading, and which languages teachers can use accurately.
India’s Central Board of Secondary Education issued 2025 implementation guidelines requiring affiliated schools to adopt language education provisions from the 2025–26 academic session. The circular states that the first language of literacy should ideally be the student’s mother tongue or a familiar regional or state language, and that the medium of instruction should be the home language, mother tongue, or familiar regional language in the foundational stage.[j]
Teacher Supply and Language Proficiency
A language policy fails if teacher deployment does not match it. A system may promise mother-tongue instruction, but if trained teachers cannot teach mathematics, reading, and science in that language, schools may revert to a dominant language. Teacher education must cover both subject pedagogy and academic language use.
Teacher supply has several technical layers: oral fluency, reading fluency, writing conventions, subject vocabulary, classroom questioning, assessment rubrics, and the ability to bridge from one language to another. A bilingual teacher is not automatically prepared for bilingual pedagogy. Training must show how to develop vocabulary, compare concepts across languages, and prevent translation from replacing explanation.
Textbooks, Assessment, and Learning Materials
Instruction language depends on materials. A school cannot deliver stable early-grade reading in a local language without graded readers, alphabet charts, decodable texts, mathematics vocabulary, teacher guides, and assessment tools. Oral teaching can help, but reading policy needs print or digital material in the same language children use to build literacy.
Assessment is another pressure point. If students learn content in one language but sit exams in another, low marks may reflect assessment language rather than subject mastery. UNESCO and UNICEF’s guidance on classroom-based assessment of multilingual learners calls for assessment planning that recognizes oral language, literacies, and curriculum learning across subjects, not just one written test language.[k]
The Early-Grade Question: Why Timing Matters
Early-grade language policy carries high weight because children are building the base of later schooling. Reading, number sense, working memory, and classroom confidence develop together. If a student hears explanations in an unfamiliar language, every lesson becomes a language test before it becomes a subject lesson.
The World Bank’s language-of-instruction position is clear on timing: first-language teaching should begin in early childhood services and continue through at least the first six years of primary education. The same position says the first language should remain present even after another language becomes the main teaching language.[d]
Why Six Years Often Appears in Policy Guidance
Six years is not a magic number, but it reflects a practical learning sequence. Children need time to gain oral fluency, decode print, read with comprehension, write connected text, handle academic vocabulary, and use language for problem solving. A short mother-tongue period of one or two years may not be enough if the later switch language becomes the language of mathematics, science, and national exams.
UNESCO’s language education page notes that six years of mother-tongue instruction are needed to reduce learning gaps for minority-language speakers. It also notes that more than 7,000 languages are spoken around the world today, which shows why one national solution cannot fit every school context.[b]
Common Policy Patterns by School Stage
Teaching-language choices change by age. Early schooling needs comprehension and foundational literacy. Lower secondary schooling adds abstract vocabulary and subject specialization. Upper secondary schooling connects to national exams, vocational pathways, university entrance, and cross-border academic mobility.
| School Stage | Typical Language Aim | Policy Risk | Data Need |
|---|---|---|---|
| Pre-primary and Grade 1 | Oral comprehension, early literacy, classroom routines | Starting in an unfamiliar language can delay participation | Home-language mapping, teacher language profile |
| Grades 2–3 | Reading fluency, writing, early numeracy, vocabulary growth | Switching too early may weaken reading confidence | Reading fluency by language, attendance, repetition |
| Grades 4–6 | Subject vocabulary, bilingual bridging, academic writing | Content learning may fall when language demand rises | Subject scores by assessment language |
| Lower secondary | Abstract concepts, science and mathematics terminology | Hidden language barriers may look like weak subject ability | Diagnostic language data and subject vocabulary tests |
| Upper secondary | Exam preparation, higher education language readiness | One exam language may narrow pathways | Completion rates, exam performance by language background |
Country Examples Showing Different Routes
Country examples show that language-of-instruction policy is shaped by system design. Centralized systems may implement national language sequences faster. Federal systems may allow provinces, states, or territories to adapt. Multilingual countries may use regional models. Immigration-receiving countries often focus on support for learners who study in the national or majority language while using another language at home.
India: Home Language and R1 in Early Schooling
India’s 2025 CBSE language-instruction circular uses the terms R1 and R2. R1 is the first language of literacy in school, ideally the mother tongue or familiar regional or state language. The circular states that R1 should serve as the medium of instruction until foundational literacy in another language is achieved. It also allows practical adjustment where classroom diversity, limited resources, or oral traditions make one language choice difficult.[j]
South Africa: Extending Mother-Tongue-Based Bilingual Education
South Africa’s Department of Basic Education describes Mother Tongue-based Bilingual Education as a programme that uses learners’ home-language resources while they also learn other languages. The department states that the programme will extend mother-tongue instruction beyond Grade 3 into Grade 4 in 2025. It also notes that the aim is bilingualism, not one-language schooling.[l]
This example matters because many systems face a sharp switch around Grade 4. If reading, mathematics, and assessment move suddenly into a second language, learners may appear to decline just as subject content becomes more abstract. A gradual bilingual route reduces the shock and keeps subject learning visible.
Australia: English as an Additional Language or Dialect
Australia’s curriculum guidance defines EAL/D students as learners whose first language or dialect is not Standard Australian English and who need support to develop proficiency in it. The guidance includes First Nations learners, immigrants, refugee-background students, children of migrant heritage, international students, and children of Deaf adults who use Auslan as a first language.[m]
This route differs from a mother-tongue medium model. The main curriculum language remains English, but policy recognizes that students need explicit language support to access subject content. The teaching-language issue therefore appears through scaffolding, learning progressions, teacher planning, and classroom differentiation.
OECD Systems: Migration and Language Spoken at Home
OECD PISA 2022 data show why language policy is not limited to low-income or highly multilingual countries. Across OECD countries, 11% of all students spoke a language at home different from the PISA assessment language. Among immigrant students, the gap was much larger: 62% of first-generation immigrant students and 44% of second-generation immigrant students mainly spoke another language at home.[h]
For these systems, the policy question is often not the official teaching language itself. It is whether schools provide enough academic-language support for students to learn mathematics, science, reading, and writing while they are still acquiring the school language.
Technical Variables That Decide Whether the Policy Works
A language-of-instruction policy becomes measurable only when it links to implementation data. Ministries, districts, and schools need indicators that go beyond official language labels. The most useful data show the match between learner language, teacher language, material language, and exam language.
The Four-Language Match
A practical system audit can examine four language layers: home language, teacher language, textbook language, and assessment language. When all four align in early grades, students face fewer barriers. When they split, schools need bridging support.
| Layer | Measurement Question | Possible Indicator | Interpretation |
|---|---|---|---|
| Learner language | What language does the child understand best on school entry? | Share of Grade 1 learners by strongest oral language | Shows demand for teaching-language options |
| Teacher language | Can the teacher explain concepts in that language? | Teacher proficiency and certification by language | Shows staffing feasibility |
| Material language | Are textbooks and readers available? | Textbook-to-pupil ratio by language | Shows whether policy has classroom tools |
| Assessment language | Do tests match the taught language? | Assessment language by grade and subject | Shows whether results measure learning or language access |
| Transition language | When does another language become the main medium? | Grade of transition and years of overlap | Shows whether bridging is gradual or abrupt |
Transition Design
Many systems do not disagree on the value of familiar-language learning. The dispute lies in transition timing. A short transition moves learners quickly into the national or international language. A long transition keeps the home language in subject teaching while gradually increasing academic use of the second language.
Good transition design includes oral exposure before written load, shared vocabulary lists, bilingual subject glossaries, teacher modelling, parallel reading texts, and low-stakes assessments in both languages. Without these pieces, a policy may become a sudden switch rather than a learning bridge.
Assessment Problems Many Short Explanations Miss
Language policy shapes test interpretation. A mathematics item written in a student’s weaker language can test reading comprehension before it tests mathematical reasoning. A science question may contain unfamiliar academic verbs, passive structures, or culture-specific examples. When this happens, exam data can mislead school leaders.
For multilingual learners, assessment needs at least three checks: language load, subject validity, and response mode. Can the student explain the concept orally but not yet write it in the assessment language? Does the test use vocabulary taught in class? Does the marking scheme separate language errors from content understanding?
UNESCO and UNICEF’s 2024 assessment guidance is useful here because it treats multilingual learners as students with language assets, not as students with a single deficit. It covers oral language, reading, writing, and curricular content such as mathematics and sciences.[k]
Current Developments Connecting Language Policy to Wider Education Reform
Recent education reforms place language of instruction next to three wider trends: learning recovery after school closures, migration-driven classroom diversity, and digital language tools. These trends do not replace teacher-led language planning, but they change the data schools need.
Learning Recovery and Reading Data
The 2022 global learning poverty update estimated that 70% of 10-year-olds in low- and middle-income countries could not read and understand a simple written text after the pandemic shock, compared with a revised pre-pandemic estimate of 57%. The same release estimated a possible $21 trillion loss in lifetime earnings in present value for this generation of students.[g]
Language of instruction is not the only cause of low reading performance. Teacher preparation, attendance, poverty, school resources, health, and assessment systems all matter. Yet language is a high-leverage variable because reading recovery depends on whether students understand the words used to teach reading.
Migration and Host-Language Support
Migration changes classroom language profiles even in systems with one dominant school language. OECD PISA data show that many immigrant-background students speak a language at home different from the assessment language. Schools therefore need strong host-language teaching, but they also need to value students’ existing languages as learning resources rather than treating them as obstacles.[h]
Digital Tools and AI Translation
Digital tools can help produce bilingual glossaries, teacher notes, audio materials, and early reading resources in more languages. Still, technology cannot decide the teaching language by itself. A translated worksheet may carry vocabulary errors, unsuitable examples, or a reading level too high for the grade. Human review remains essential, especially for early literacy and assessment.
Country-by-country education pages can help readers compare how different school systems organize governance, curriculum, assessment, and language-related decisions across national contexts.[n]
How to Read a Language-of-Instruction Policy
A strong policy document should not only name the language. It should show the grade sequence, subject coverage, teacher plan, material plan, assessment plan, transition point, and support model for students who speak other languages. Without these details, the policy remains a statement rather than a school system design.
- Grade coverage: Which grades use the home language, national language, or international language?
- Subject coverage: Is the language used only for reading, or also for mathematics, science, and social studies?
- Teacher readiness: Are teachers trained to teach subjects bilingually or only to translate?
- Materials: Are textbooks, readers, teacher guides, and assessments available in the chosen language?
- Transition point: Does the policy shift language suddenly, or does it keep two languages active for several years?
- Assessment validity: Do exams measure subject knowledge separately from language proficiency?
Policy Trade-Offs Schools Must Balance
Language-of-instruction policy involves several trade-offs. A familiar language may improve early comprehension, but it requires trained teachers and materials. A national language may simplify administration, but it can create barriers for children who do not speak it at home. An international language may support later mobility, but early full-medium instruction can weaken learning if introduced before students are ready.
The strongest systems avoid a false choice between home-language learning and wider-language opportunity. They build both. Early teaching uses a language students understand. Additional languages enter through planned oral exposure, vocabulary growth, reading development, and later subject use. Assessment then follows the language sequence rather than surprising students with a language they have not yet mastered academically.
A Data-Based Reading of the Issue
A language policy is most credible when it answers four questions with data: who the learners are, what language teachers can teach in, what materials exist, and what language exams use. If any one of these four layers is missing, the policy may look orderly on paper but uneven in classrooms.
Reference Notes
[a] 40% don’t access education in a language they understand | Global Education Monitoring Report — UNESCO GEM Report article on the global share of learners without education in a language they understand.
[b] Languages in education | UNESCO — UNESCO language education page with global figures on multilingual education, learner language access, and world language diversity.
[c] Language of instruction | UNICEF FLN Hub — UNICEF resource hub on language of instruction and foundational literacy and numeracy.
[d] Teaching young children in the language they speak at home is essential to eliminate Learning Poverty — World Bank release summarizing its language-of-instruction policy approach.
[e] Languages matter: global guidance on multilingual education | UNESCO — UNESCO 2025 guidance page on multilingual education policy and practice.
[f] UIS Data Browser — UNESCO Institute for Statistics data portal, including early-grade first-language instruction data notes.
[g] 70% of 10-Year-Olds now in Learning Poverty, Unable to Read and Understand a Simple Text — World Bank press release on the 2022 global learning poverty update.
[h] Immigrant background and student performance: PISA 2022 Results (Volume I) | OECD — OECD PISA 2022 data on immigrant students and language spoken at home.
[i] Learning Poverty is a combined measure of schooling and learning. — World Bank explanation of the learning poverty indicator and its reading-based methodology.
[j] Implementation Guidelines for Language Instruction under National Curriculum Framework for School Education-2023 for Foundational and Preparatory Stages — CBSE circular on R1, R2, and early-grade language-of-instruction implementation.
[k] Guidance for the classroom-based assessment of multilingual learners: assessing languages, literacies and learning across the curriculum — UNESCO and UNICEF guidance on assessment in multilingual classrooms.
[l] Mother Tongue-based Bilingual Education — South Africa Department of Basic Education page on mother-tongue-based bilingual education.
[m] EAL/D students | V9 Australian Curriculum — Australian Curriculum guidance on English as an Additional Language or Dialect learners.
[n] Countries Library – Education by Country — Country-level education system profiles for comparing schooling structures, governance, curriculum, and assessment context.