Growing up with languages: Implications for multilingual education

What is known about the language development of multilingual children in very linguistically diverse societies, and what are the implications for multilingual education? This is the topic of a lecture by Prof. Ajit Mohanty at Jawaharlal Nehru University, India, summarized in this post. (You can watch the lecture yourself here.) I like this lecture because Mohanty promotes translanguaging, or the child’s use of their integrated language repertoire to make sense of the world across so-called language boundaries… and reconciles this phenomenon with domains of language acquisition (specific languages a child knows to various degrees, and involving various competencies) and language socialization (the child’s developing awareness of distinct languages, dialects, and registers in their social interactions). Mohanty, a former Fulbright scholar at Columbia University, wrote The Multilingual Reality: Living with Languages (2018) and has developed multilingual education policy for Odisha (a state in India) and Nepal.

This is the first time I’ve summarized a talk rather than an article, in this case one of the talks in a series called the Plurilingual Lab Speaker Series at McGill University, Canada. This series examines (1) questions about language learning and teaching, (2) challenges and benefits of multilingual education, (3) relevant issues of social justice, (4) decolonization of language education in classroom teaching, and (5) how language policy at the local, national, and international level can better meet societal needs.

Mohanty, A. (2021, March 25). Growing up with languages: Implications for multilingual education. Plurilingual Lab Speaker Series, McGill University, Montréal, Canada.

Languages spread across the globe (not just English), and every country contains a tapestry of multilingualism rather than the myth of one, two, or several national languages. Two points must be made: (1) the dozens or hundreds of languages in every national society are not socially equal, and (2) no language is used for exactly the same purposes as any other. The first point is troubling, but the second is not. A social experiment that would reproduce every single language to be a mirror image of all the others, so that each can be used on its own in every context, is not only redundant but dystopian. If the multilingual person is not two or three or four monolinguals in one brain, why should we characterize a multilingual country in this way as well?

Due to both the human-driven unequal status of languages in society and natural social forces, each language in a country has domains in which it is more or less likely to be found, whether in formal education, religious rituals, or various forms of media for different populations, generations, and audiences. Mohanty discusses the relationship between these two phenomena—linguistic inequality bad, domain specificity normal—in this lecture, showing how they interact with one another and suggesting how we can make the bad one better by focusing on the normal one through a translanguaging approach to education.

Growing up with languages

In the second half of the talk, Mohanty talks about education, but in the first half he talks about multilingual children. Before a multilingual child in a very linguistically diverse society even enters school, what is their life like?

That child is born into a society where different languages of unequal status exist in different realms (i.e., diglossia). At the same time, people experience languages interwoven together—in thought more than in oral speech, in oral speech more than in writing—”as a network or totality of communicative acts” (i.e., heteroglossia). According to Mohanty, the novelist Salman Rushdie once reported hearing five languages in one sentence while in India. As another example, here is a dialogue I found between two older adults in Singapore, one of whom was just widowed, from Li Wei’s (2018) article “Translanguaging as a Practical Theory of Language.” You can see Hokkien in bold, Teochew in square brackets < >, Mandarin underlined, Malay in double quotation marks, Cantonese in single quotation marks, Singlish in Italics, and English in the rest:

Li Wei (2018), pp. 13-14

As babies grow into toddlers and then into small children, they go from their immediate family and neighbourhood to informal local area gatherings (social and religious) to complex local institutions with wider connections (markets, banks, health centres) to wider regional, national, and international settings (hospitals, trains, major city centres). At each layer, more languages are added, as children encounter the regionally dominant language, the national language, the language(s) dominant in that part of the world (e.g., Arabic in the Middle East), and more distant foreign languages. Of course, India has layers and layers of indigenous languages, regional languages, and nationally dominant languages, in addition to foreign languages. Many schoolchildren in India speak at least 3-5 languages, but their proficiency in each of these is not balanced; in some languages, they only know bits and pieces for specific purposes.

A similar situation exists in countries with hundreds of indigenous languages and 10-30 regionally dominant ones—think of the nations in Africa, Southeast Asia, and South Asia. Though Mohanty didn’t mention it, I assume a child’s strongest (most generative) proficiency tends to be in the regionally or nationally dominant language, whether or not it is their home language (that could be a minority language). This is what I have observed in my studies of Filipino immigrant children in North America; for example, in the case of a typical newcomer immigrant child of Ilonggo heritage, their Tagalog is predictably stronger than either their Ilonggo/Hiligaynon or their English. But if that child speaks an indigenous Philippine language and lives in Ilonggo territory, I would predict him to speak Ilonggo better than his home language, Tagalog decently (though somewhat behind his Tagalog L1 peers), and English not so well. Home language loss exists everywhere (even in non-English-dominant countries)… it trends towards the dominant regional and the national languages.

This means indigenous students will struggle in classes taught in the regional language (which policymakers mistake for their mother tongue), and regional language speakers (mistakenly called “dialect speakers” even though their languages are not dialects of the national language!) will struggle in classes taught in the national language, just as national language L1 speakers struggle in English Medium Instruction. These are all examples of subtractive language education in officially target-language-only classrooms, where students are not allowed to speak their other languages (but do so anyway to cope), teachers are not allowed to use these languages (but do so anyway to cope), and the result is a great deal of wishy-washy learning. In the end, children develop the greatest proficiency in the regionally/nationally dominant language… more due to naturalistic immersion and assimilation in the world outside school than due to formal instruction. And if this language is not the language of instruction (i.e., if it is English), students will use the regional/national language, which still counts as “their” language, to do a great deal of metalinguistic talk about the grammar of English but struggle to have a conversation in English.

Those are domains of language acquisition as I see them following Mohanty’s talk. In terms of people’s social perception of distinct languages, Mohanty describes how children in multilingual societies progress through a generalizable developmental pattern:

  1. From not seeing languages as differentiated, to
  2. Broad differentiation (e.g., Language A at home, Language B at school, Language C in the market), to
  3. Contextual, interlocutor and functional differentiation (e.g., knowing when to use which language(s), in what way, and to what degrees, with potentially the same person, at different times; for example, when and how to use Language A, or A + B, or A + B+ C with a parent according to different contextual factors). I can relate to this… as my mother and I can converse in English only, in Tagalog only, in Taglish, or in an arrangement in which she addresses me in Tagalog and I reply in English. Many of our language choices are subconscious and dynamically shifting from moment to moment in a conversation as we translanguage (i.e., use our whole language repertoires to communicate). There are, however, proficiency and domain factors restricting how each interlocutor can use her languages—e.g., she will speak more Tagalog than me; she has a wider range of words in Tagalog; I will use applied linguistics terms in English she won’t understand; it is socially ridiculous and never attempted for me to talk to her in Tagalog only while she speaks back in English only, etc.

As all children, regardless of nature and degree of multilingualism, increase their awareness of language norms with regard to different interlocutors, contexts, and tasks, they also develop an awareness of the “why” of language choices, which languages or dialects are of higher/lower status in society, and the racialized, gendered, and classed identities associated with particular ways of speaking. Moreover, as children grow, they also start to develop their own language preferences and performances. The more heterogeneous (varied) a child’s early language environment, the earlier the development of their intuition about distinct languages and dialects and the norms of multilingual communication, an awareness that can even precede the formal teaching of languages at school. Mohanty’s PPT states: “The processes of multilingual socialization transform the sociolinguistic synergy in multilingual societies to a psychological reality for every child.” Also, social class impacts language socialization, as we shall see with three (presumably white-collar) siblings in an Outer-Circle English-speaking country (India) who knew English better than any Indian language.

These three siblings were Mohanty’s research participants: a 12-year-old boy, a nearly-8-year-old boy, and a nearly-6-year-old girl. By his description, the three siblings’ language development was quite similar. They all spoke bilingually in English and Odia (an Indian language) with parents/grandparents, and English only with all others—even visitors knowing Odia, and English only among themselves, even though they knew Odia. This may have come as a surprise to some audience members, as it cannot be chalked down merely to snobbery or linguistic imperialism. It is a question of what social meanings Odia came to have in the lives of these siblings. Ultimately, because Odia was used only in the domain of very intimate social-affective communication between children and elders, it was not used in domains where this social situation did not apply.

Moreover, since the children were English-dominant and weak in Odia (hence the limited and specific domain of the latter), they were most comfortable speaking English with one another, and possibly might have felt judged by people outside the family for their lack of Odia proficiency, which could have been one reason why they did not use the language except with their parents and grandparents. This imbalance in proficiency was apparent in their speech; although language gaps are not the only reason why people code-mix, the three siblings’ code-mixing featured many uses of English words when they did not know words in Odia. Now, had they spoken Odia more fluently, and with more people besides elders in their immediate family, Odia likely would not have had that special restricted intimate value; it would have just been another language available for their practical use in a wide range of contexts (and hence no particular context). Note that the restricted domain did not diminish the value of Odia—in fact, it may have served to elevate it.

Another potentially surprising finding is that the children didn’t seem to have enough exposure to Indian languages to be able to name them. Regular visitors came to the home and spoke Bengali, Gujarati, Hindi, Malayalam, Marathi, Nepali, Spanish, Telugu, or Urdu, and the children had trouble identifying what language was being spoken when one of these languages was spoken. This cannot only be attributed to age (the oldest was 12). Rather, it seemed unnecessary for the children to know; there was no practical need for them to do so, as they were presumably addressed in English (and they apparently weren’t curious). This resulted in lack of naming/recognizing competence, which is one of Blommaert and Backus’ (2013) four levels of language proficiency:

  1. Widespread age-appropriate competence in almost any domain;
  2. Partial competence in specific domains;
  3. Knowing “bits and pieces”/phrases/words in a language;
  4. Recognizing competence (i.e., don’t speak the language but recognize it as Russian/Spanish/Portuguese/etc. when you hear it).

Inability to name a language from one’s country of origin due to limited experiences is something I can relate to, as I often hear something on the streets of Hong Kong or Vancouver that I am almost certain is a Filipino language, without knowing what it is. It could even be advanced Tagalog or Filipino (Tagalog-based, containing a mix of Philippine languages) but with too many words I don’t know. Mohanty’s point, I think, like that of Blommaert and Backus, is that—unless you are a very curious person who takes pleasure in language learning—you know the languages you know to the extent you need (this includes learning to merely recognize languages, and explains why we can all identify a dozen of world’s most powerful and prestigious languages).

Now, I turn to what this “growing up with languages” means for multilingual education; the title of the talk is “Growing Up with Languages: Implications for Multilingual Education.”

Implications for multilingual education

It is an oversimplification merely to claim that “everyone deserves to be taught in their mother tongue” for as long as they want, through their K-12 education and even into higher education. This assumption is problematic for several reasons. First, what is the mother tongue? To answer, I will quote a memorable passage from a book chapter by Terrence G. Wiley, who quoted his colleague Munene Mwaniki. Mwaniki explained:

I was born in Eastern Kenya but now [at the time of relating this story] live in South Africa. The common language in my native Kenyan region is Kimbeere. That was actually my father’s tongue. However, the larger regional language—mainly because the geopolitical reasons associated with standardization and national language policies on language of education was—Kikuyu. Thus, our local Kimbeere is a minority language that is only about 50% mutually intelligible with Kikuyu. My initial schooling was in Kikuyu, which was considered to be my “mother tongue.” In fact, my mother’s tongue was Kikamba. She was a language minority in our Kimbeere-speaking community. So when I went to school, I encountered my alleged “mother tongue” Kikuyu as more or less a second language. By fourth grade (Standard 4 in the Kenyan educational system) I had to learn Swahili and English. From Standard 4 … English was the medium of instruction, with Swahili being used for Swahili lessons. It is instructive that a national-examination taken at the end of 8th grade (Standard 8) is written in English, save for the Swahili lessons. So if you ask me, what my mother tongue [i.e., first language/L1] is, I suppose it is Kimbeere, which is not my mother’s tongue. (Wiley, 2020, pp. 275-276)

Note that Mwaniki describes his most comfortable language growing up as the language he was most immersed in until he started school, Kimbeere. This was the dominant community language, not his mother’s tongue, but by coincidence his father’s tongue. A government with good intentions offered instruction in the regionally dominant language, Kikuyu, in the early grades (which probably helped out Kikuyu L1 speakers), but inevitably there was a transition to the two national languages, Swahili and English.

Mohanty (2018) points out that India has 300 or 400 languages, 20+ official ones (i.e., major national languages), 120+ in print media, about 90 in radio/TV, about 35 in films, and around 30 in primary education (this last figure is a generous estimate; we can say “30 languages” if we count special programs for specific communities). The number of languages taught in schools has declined, from 81 in 1970 to 41 in 1998 and about 30 today (by the above generous estimate). According to Mohanty, there is a roughly 3-level hierarchy, with (1) English at the top, (2) major national languages, and (3) minority languages. Mohanty calls the divide between (1) and (2) the English-Vernacular divide, and the divide between (2) and (3) the Vernacular-Tribal divide. Every year, linguistic diversity is lost, as students who speak minoritized languages shift towards nationally dominant languages and English like the three siblings he studied. This results in a “vicious cycle” for minority languages (see the figure below). Because of educational and social neglect, these languages become weak and underused, a fact observable in the younger generations, and this demonstrable weakening and reduction in domains of use becomes justification for further neglect. (Look into the work of Jasone Cenoz and Durk Gorter, on Basque language revitalization in Spain, presented in another Plurilingual Lab talk, to see why minority languages will not save themselves… they require governments, activists and users to maintain them.)

Image from Mohanty (2021) Plurilingual Lab lecture

The educational neglect of minoritized languages leads to illiteracy, poor educational performance, language minoritized students dropping out of school, subtractive language learning in forced submersion programs, “capability deprivation,” and loss of diversity (Mohanty, Panda, Phillipson & Stutnabb-Kangas, 2009). What happens when children are taught in languages they don’t understand (even valued ones that they want to learn, including the national language and English)?

Mohanty listed the costs: non-participation, teacher-dominant classroom interactions, coercive relations of power, rote learning and memorization (if any learning takes place at all!), and large scale educational failure. (He asked the audience what they would do if he addressed them in a language they didn’t know… pointing out that they would be sleeping, or leave the Zoom meeting.)

So does this mean we need to implement Mother Tongue education for everyone until the age of 18 or 22? Even with good government intentions and resources (rightfully) poured into education in at least the main regional languages in the early years, the problem is that many students will still be taught in a language that isn’t their mother tongue (see Mwaniki’s quote above) in South Asian, Southeast Asian, and African countries… in addition to many countries where national discourses make people believe that their country lacks diversity in its native languages. (For example, I wonder what my Master of Education students from China, who are K-12 teachers, mean when they tell me “all my students speak Chinese.” However, I feel unable to probe into this further, as I speak no Chinese languages and have no recognizing competence to identify Chinese languages beyond Mandarin and Cantonese.)

Also, recall domains of use—not all languages are written; not all have vocabulary for science, math, and other academic subjects, partly due to the above “vicious cycle.” One argument against Mother Tongue education beyond the early years would be that it would be going for “balanced” bilingualism, knowing everything in Language A that you know in Language B, and knowing everything in Language B that you know in Language C… and we know that is redundant: the individual language repertoire is an integrated, all-terrain vehicle, not a bicycle or tricycle with two or three equal wheels (García, 2009).

No—because language use in society is heteroglossic, classrooms should be heterglossic too. And because some classrooms have a great degree of individual diversity when it comes to domains of language use (even if students share the same handful of languages when asked to list them), translanguaging really is the only feasible option for letting all students make meaning with their whole language repertoire. It is, of course, important to ensure that they make an effort to bridge their repertoires and those of others, and whether/how they do this can be assessed by examining small group conversations (Mendoza, 2020).

To conclude: we can at least combine Mother Tongue education in the early years, in regional languages, with a translanguaging classroom language policy and an inclusive classroom culture with language-brokering. Now, the next question is, how do we want to make use of the formal classroom setting to extend students’ domains of language use beyond what people naturally develop, and how can this be done to achieve opportunity of access to dominant languages as well as revitalization of minority languages, with equal negotiation between different stakeholders (as they take into account their shared and different views)? These are some of the questions for policymakers, curriculum developers, and teachers that I hope to address in this blog.

While supporting translanguaging in education, we need to promote the multilingual awareness Mohanty spoke of in his talk, building on instincts and skills that multilingual children are already developing and demonstrating in their natural language socialization. While language shift is hard to stop, the human damage that goes along with it—illiteracy, poor educational performance, dropping out of school, subtractive language learning in forced submersion programs, “capability deprivation,” and loss of diversity (Mohanty et al., 2009)—may be far more preventable. It is under principals’ and teachers’ control to care for students as people, and care about the quality of their educational experiences. Besides, the languages that “died” don’t really die—they infiltrate and change the dominant languages, shaping them through language contact. This is why language shift towards dominant languages is not entirely the same as loss of language diversity, nor does it necessarily have to mean bad human experiences, if we are conscientious in our actions.

Dominant languages cannot be kept “pure” from minoritized languages’ influence any more than the minoritized languages cannot help being taken over by the dominant ones. Some people will hate the first fact; others will hate the second. But as many translanguaging scholars remind us, it is people, and their well-being, and their educational experiences, and the quality of their lives and social experiences, that matter more than languages. When we value the whole of students’ language repertoires, taking on a “translanguaging stance” (Seltzer & García, 2020) in education, it guides us as to what to do with the languages.


Blommaert, J., & Backus, A. (2013). Superdiverse repertoires and the individual. In I. de Saint-Georges & J.-J. Weber (Eds.), Multilingualism and multimodality: Current challenges for educational studies (pp. 9-32). Leiden, Netherlands: Brill | Sense.

García, O. (2009). Bilingual education in the 21st century: A global perspective. Malden, MA; Oxford, UK: Wiley-Blackwell.

Li Wei. (2018). Translanguaging as a practical theory of language. Applied Linguistics39(1), 9-30.

Mendoza, A. (2020). What does translanguaging-for-equity really involve? An interactional analysis of a 9th grade English class. Applied Linguistics Review. 1-21. Early view.

Mohanty, A. K. (2018). The multilingual reality: Living with languages. Bristol, UK: Multilingual Matters.

Mohanty, A., Panda, M., Phillipson, R., & Skutnabb-Kangas, T. (Eds.) (2009). Multilingual education for social justice: Globalising the local. New Delhi, India: Orient Blackswan.

Seltzer, K., & Garcia, O. (2020). Broadening the view: Taking up a translanguaging pedagogy with all language-minoritized students. In Z. Tian, L. Aghai, P. Sayer & J. Schissel (Eds.), Envisioning TESOL through a translanguaging lens: Global perspectives (pp. 23-42). Cham, Switzerland: Springer.

Wiley, T. G. (2020). Afterword: On contested theories and the value and limitations of pure critique. In J. MacSwan & C. Faltis (Eds.), Codeswitching in the classroom: Critical perspectives on teaching, learning, policy, and ideology. New York, NY; Abingdon, UK: Routledge.

Published by annamend

Assistant Professor in the Department of Linguistics, University of Illinois at Urbana-Champaign

%d bloggers like this: