When the Language of School Is Not the Language of Learning

A child can be sitting in a classroom, listening carefully to the teacher and still miss the lesson—not because the concept is beyond the child’s ability, but because the language carrying that concept is unfamiliar.

This creates a problem that enrolment statistics cannot easily reveal. A school may have a building, a teacher and a textbook, yet the child may still face an additional task before learning can begin: understanding the language in which the lesson is being delivered.

India faces this problem on a large scale. According to NCERT data cited in the education corpus, nearly 44% of children entered school in 2022 speaking a language different from the medium of instruction. For such children, the linguistic demands of schooling can overlap with the demands of learning itself.

The policy challenge is therefore more complicated than choosing one language over another. India has to make learning accessible without closing off the opportunities that additional languages—particularly English—can provide.

The difficulty begins with India’s diversity

India’s linguistic diversity makes a single national classroom model unrealistic. The 2011 Census recorded more than 1,300 mother tongues and 121 constitutionally recognised languages.

This diversity is an educational reality, not merely a cultural statistic.

For a young learner, language determines how instructions are understood, how questions are asked and how unfamiliar ideas are connected to existing knowledge. A student who is still acquiring the language used by the school may therefore have to spend part of the learning process working out what the teacher or textbook is saying.

That is why NEP 2020 recommends that, wherever possible, the medium of instruction in the early grades should be the child’s home, mother, local or regional language, at least until Grade 5 and preferably beyond Grade 8. It also encourages bilingualism, including the use of the mother tongue alongside English.

The principle is straightforward. Conceptual learning should not be unnecessarily obstructed by linguistic unfamiliarity.

But translating that principle into classrooms is considerably harder.

Language proficiency itself remains a problem

There is an important complication here.

It would be misleading to assume that simply switching to a regional language automatically resolves India’s learning problems. Many children struggle to read and communicate adequately even in languages they already encounter regularly.

ASER 2018 found that 27% of Class 8 children in rural India could not properly read a Class 2-level text in their regional language or Hindi. By 2022, the figure had risen to 30.4%. English proficiency was also weak: 73.8% of Class 8 children could not read simple English sentences in 2016, while the corresponding figure remained 53.3% in 2022.

National Achievement Survey findings tell a similar story. In 2017, only 48% of Class 8 students could read a simple paragraph in their regional language or Hindi, 47% could write an essay or letter and 42% demonstrated a good grasp of grammar. In 2021, these figures improved to 56%, 49% and 44%.

These numbers cannot establish that the medium of instruction caused the learning deficit. Teacher quality, pedagogy, school conditions and foundational learning all matter.

But they reveal something equally important: the country cannot assume that children automatically possess strong proficiency in the language through which education is delivered.

That makes multilingual policy a question of quality as much as quantity.

If students are weak in their first or dominant school language and also have limited command of English, adding another language does not by itself create multilingual competence. The system first has to ensure that children acquire languages well enough to use them for learning.

English cannot simply be wished away

This is where the language debate becomes more difficult.

English occupies a special position in Indian education because it has consequences beyond the classroom. It is used extensively in higher education, professional education, national examinations and parts of the labour market. It is also associated by many families with access to institutions and careers that may otherwise be harder to reach.

The education corpus notes that English-medium schooling has become especially attractive because families see English proficiency as a route to better opportunities. It also records that this perception has been particularly important among traditionally disadvantaged groups.

This produces a genuine parental dilemma.

A parent may recognise the educational advantages of beginning with a familiar language and still insist that the child learn English early. The motivation need not be cultural preference. It can be a calculation about the child’s future.

That calculation cannot simply be dismissed as misguided.

If a child reaches secondary school with strong understanding in a regional language but inadequate English proficiency, the immediate learning experience may be better while later opportunities remain constrained. Conversely, introducing English without adequate support can make early conceptual learning harder.

The policy question is therefore not English or Indian languages.

It is how effectively the system can build both capabilities.

Bilingualism offers a more practical route

There are already examples of approaches that attempt to avoid this binary.

NEP 2020 encourages bilingual education. In Haryana, anganwadi educators use Hindi and English materials together, including both scripts while teaching basic concepts. The education corpus also records bilingual Science and Mathematics textbooks in Assam in English, Assamese, Bodo and Bengali, and Telugu-English bilingual textbooks in Andhra Pradesh.

These approaches matter because they change the role of the additional language.

Instead of asking a child to abandon the language they understand in order to acquire English, the classroom can use the familiar language to support conceptual understanding while English is developed progressively.

That requires appropriate materials.

NCERT released digital books in 104 regional languages and dialects in 2024. Such resources can reduce one of the practical constraints on multilingual education: the shortage of quality learning material in languages other than English.

But materials alone are insufficient.

A multilingual textbook can provide two languages on a page. A teacher still has to know how to move between those languages while explaining a difficult idea.

Translation can expand access to content; it cannot replace multilingual teaching ability.

Less widely represented languages face a harder problem

The challenge becomes more acute for children whose home language is not one of the major regional languages.

Odisha provides an important example. Its multilingual education programme operates across 21 tribal languages in 17 districts and supports nearly 90,000 children.

The significance of such programmes lies not simply in preserving linguistic diversity. They recognise that the child’s existing language can be relevant to the process of acquiring formal academic knowledge.

Digital initiatives can extend this effort. The education corpus identifies DIKSHA, BHASHINI, Adi Vaani and community-developed language technologies as part of a wider attempt to expand access to Indian-language content.

Yet technology has limits.

A platform can make material available in a language. It cannot automatically solve the problem of a classroom in which the teacher does not know that language, cannot assess the child’s understanding through it or lacks the training to shift between languages.

The constraint is therefore partly technological, but fundamentally human.

Language can also reproduce inequality

The medium of instruction matters because children do not arrive at school with identical linguistic resources.

Consider two students entering the same class. One has grown up hearing and speaking English regularly. The other encounters English mainly through school. Both may be taught the same lesson, but the linguistic effort required to follow it can be very different.

Families with greater financial resources can compensate outside school. They can pay for tutoring, provide English-language exposure and purchase additional learning materials.

Families without those resources depend much more heavily on the school itself.

This means language policy can interact with socio-economic inequality. A system that assumes that every child can independently acquire an unfamiliar instructional language risks transferring part of the responsibility for educational success from the institution to the household.

That is especially problematic when private tutoring becomes the mechanism through which children overcome language-related difficulties.

The issue is therefore not simply whether a particular language is prestigious or culturally important. It is whether the education system gives children sufficiently similar conditions from which to begin learning.

But mother-tongue instruction has limits too

There is an equally important danger on the other side.

The evidence does not justify treating mother-tongue instruction as a universal solution. Children can have weak foundational literacy even when assessed in a regional language or Hindi. The ASER and NAS evidence demonstrates this clearly.

A familiar language can remove one obstacle. It cannot compensate for poor pedagogy, inadequate materials, insufficient reading practice or weak teacher preparation.

There is also the problem of transition.

Students eventually encounter universities, competitive examinations and employment environments where English may be important. If English acquisition is delayed without a strong plan for developing proficiency later, an intervention designed to improve early learning could create a different barrier at the next stage.

The answer is therefore not to replace English with the mother tongue.

Nor is it to replace the mother tongue with English.

The more defensible objective is progressive linguistic capability: children should first be able to understand and express increasingly complex ideas, while gradually acquiring the additional languages they will need for higher education, mobility and work.

The question should be “which language, when and for what?”

This changes the policy discussion.

Instead of asking which language should dominate the education system, policymakers should ask:

Which language is most appropriate for a particular stage of learning?

When should another language be introduced?

What level of proficiency should the child achieve?

What resources and teacher capacity are required to make that possible?

And, critically: How will the system ensure that children from poorer households are not left to solve these problems privately?

The early years may justify greater use of the language children understand best because conceptual development is especially important at that stage. English can be taught alongside it rather than treated as an alternative.

As students become older, the balance can change.

This approach also requires assessment systems to become more sophisticated. A child who fails to answer a question may lack knowledge of the concept. But the child may also understand the concept and struggle to express it in the assessment language.

Those are different educational problems.

If the system cannot distinguish between them, it risks treating a language barrier as evidence of intellectual weakness.

Policy needs to move beyond the three-language arithmetic

The debate over multilingualism often becomes a numerical exercise: two languages versus three.

That can obscure the more important question of proficiency.

The education corpus records that NAS does not assess third-language proficiency, creating an important data gap in evaluating the effectiveness of the three-language approach.

Without reliable evidence on what students can actually do in their additional languages, policy can end up measuring exposure rather than capability.

A child who studies three languages but cannot use any of them confidently for meaningful communication has not necessarily benefited from multilingual education.

Conversely, a child who develops strong literacy in one language while acquiring usable proficiency in another may have a stronger educational foundation even if the formal language count is lower.

This is why the ultimate metric should not be the number of languages appearing on the timetable.

It should be what children can understand, communicate and learn through those languages.

The real choice is between weak capability and strong capability

India does not need to choose between linguistic rootedness and global opportunity.

It needs to stop assuming that one must come at the expense of the other.

A familiar language can give a child a stronger starting point for conceptual learning. English can widen access to higher education, professional opportunities and national and global networks. Additional Indian languages can support communication across regions and strengthen cultural and social connections.

The educational system’s task is to develop these capabilities without overwhelming children or shifting the burden onto families.

That requires trained teachers, appropriate bilingual materials, reliable assessment, language-specific resources and policies that recognise differences between States and linguistic communities.

Most importantly, it requires a change in what success means.

A language policy should not be judged by how many languages a child is officially taught.

It should be judged by whether the child can learn through a language, communicate through it and eventually use it to expand the range of opportunities available to them.

The central question is therefore not which language should win the classroom.

It is whether the classroom can ensure that language becomes a resource for learning rather than an additional barrier that children must overcome before learning can begin.