Why Our Kids Don’t Speak Punjabi — And What the Research Actually Says

There is a conversation that repeats itself in Sikh homes from Toronto to Southall to Melbourne, and it usually begins with an observation.

Walk past a group of Chinese-heritage teenagers at a mall food court and you will often hear them switch into Mandarin or Cantonese with each other — not because a parent told them to, but because that is simply one of the languages they talk in. Nobody is supervising. Nobody is being reminded.

Now picture most Punjabi Sikh homes. A parent says “Punjabi vich bol” for the fourth time that day. A child answers in English anyway. A small, tired friction settles between them — and over years, that friction becomes a relationship in which one person is always correcting and the other is always resisting.

Parents usually land on the same theory: maybe the problem isn’t the children. Maybe the adults have quietly switched to English at home.

The research says that theory is essentially correct, that it has been documented for over thirty years, and that there is a specific number attached to it.

The short version

If you read nothing else, read this.

A child’s Punjabi is predicted by how much Punjabi they hear at home, not by how often they are told to speak it. In a study of 1,899 families, the parents’ language pattern was the deciding factor.

The danger zone: when one parent mixes Punjabi and English and the other has moved to English, only about 36% of families end up with a child who speaks the heritage language.

The strong zone: when both parents use the heritage language at home, that figure is about 97%.

Nagging actively backfires. Pressure over language is linked to shame, family conflict, and families giving up on the language altogether to relieve the stress.

The single highest-value habit: answer your child in Punjabi. Every time. That’s it.

Start today: for one week, track your own Punjabi use — not your child’s. Change nothing else.

Is Punjabi actually in decline in the diaspora?

Wherever Sikhs have settled, the official language statistics tend to look reassuring. Canada offers the clearest picture, since it holds the largest Sikh population outside India and collects detailed language data.

In the 2021 Census, Punjabi became the second-largest non-official mother tongue in the country at roughly 667,000 speakers, overtaking Cantonese. Punjabi mother-tongue speakers grew about 33% in five years (Statistics Canada, 2021, as summarised by Environics Analytics, 2022; Canadian Language Museum, 2022).

Here is the problem with that number.

It is an immigration story, not a transmission story.

Statistics Canada was explicit that growth in South Asian languages spoken at home was driven by newcomer arrival, with Punjabi home use rising 49% between 2016 and 2021 — many times the overall population growth rate (CIC News, 2022). Punjabi also had one of the highest proportions of any large mother-tongue group speaking neither English nor French at home (Environics Analytics, 2022), the signature of a very recently arrived population.

The figures look healthy because Punjab keeps sending more first-generation speakers. They say very little about the children already here.

The British Sikh community shows the next chapter. Being one to two generations deeper, it has been studied more closely. Research tracking three generations of Punjabi–English bilinguals in England found that by the second and third generation, speakers had shifted into a distinct contact variety of English rather than maintaining the parental pattern (Sharma & Sankaran, in The Phonetics and Phonology of Heritage Languages).

This is also the exact trajectory Italian, Polish and Greek communities in Canada followed a generation ago. Those languages are now in visible decline at home, attributed directly to aging speakers and the absence of new arrivals (CIC News, 2022). Their census numbers looked fine too — right up until they didn’t.

In short: Immigration is masking transmission. Healthy headline numbers are hiding what is happening inside individual families.

Do Chinese families really do it better?

Partly. But not for the reasons usually assumed — and the difference is the most useful thing in this article.

The romantic version, that Chinese diaspora families have simply solved this, does not survive contact with the literature:

Research on Chinese immigrant children in North America consistently finds English vocabulary growing rapidly while heritage language vocabulary shows early stabilisation and then decline, even in homes with substantial Mandarin input (Koh, Chen et al., System, 2021).

Studies of Chinese heritage families document the same struggles Sikh parents recognise: limited time and energy, and children losing motivation (Chen & Zhang, 2021).

A 2026 study of Chinese families in the Netherlands found parents deliberately choosing not to enforce Mandarin — prioritising family harmony and adjusting expectations downward rather than fight the battle (Wang et al., 2026).

That is not a community that escaped this problem. It is a community making the same painful trade-off.

So why the food-court difference? Four structural advantages, not better parenting:

Peer density. Enough same-age children that the language becomes a friendship language rather than only a parent language. Buildable — but it has to be designed, not simply gathered.

Perceived economic value. Mandarin is widely framed as a career asset; Punjabi as sentimental. Partly within our control.

Grandparent presence. Longitudinal research found grandparent presence significantly predicts heritage language maintenance (Verdon et al., 2014). Most Sikh families already have this.

Scale of infrastructure. Large, standardised weekend school systems built over decades. Buildable, slowly.

In short: Three of the four advantages are buildable. None of them is “remind your child more often.”

The 36% problem — the study that reframes everything

Annick De Houwer, a leading researcher on childhood bilingualism, surveyed 1,899 families where at least one parent spoke a language other than the majority language, focusing on children aged 6 to 10.

Every single child spoke the majority language. What varied enormously was whether they also spoke the minority language.

The deciding factor was not the child’s aptitude, motivation or intelligence. It was the pattern of what the parents spoke to them (De Houwer, 2007).

Share of families with at least one child speaking the heritage language:

97% — Both parents spoke only the heritage language at home

93% — One parent heritage-only, the other heritage plus majority

74% — One parent heritage-only, the other majority-only (the classic “one parent, one language”)

36% — One parent mixing both, the other majority-only

Look at that last row. For many families it is a photograph of their own kitchen. When one parent mixes Punjabi and English and the other has essentially moved to English, roughly two-thirds of children do not end up speaking the heritage language at all.

De Houwer found that the popular “one parent, one language” method is “neither a necessary nor a sufficient condition” for success (2007, p. 420). What matters is total, consistent heritage-language input. The strongest results came from what researchers call mL@H — minority language at home, where the heritage language is simply the household’s default.

In short: A child’s Punjabi is not determined by how often they are told to speak Punjabi. It is determined by how often they are spoken to in Punjabi. The instruction is not the input.

Where a heritage language is actually won or lost

Joshua Fishman built the foundational framework for reversing language shift, the Graded Intergenerational Disruption Scale. Across its eight stages, one is repeatedly identified as the pivot point: Stage 6 — the intergenerationally continuous home, family and neighbourhood (Fishman, 1991; 2001).

Everything else in a language’s ecosystem — classes, media, institutions, recognition — sits on top of that foundation and cannot replace it. Fishman cautioned that measures introduced before Stage 6 is secured tend to produce short-lived results, and that leaning on institutions too early leads to disillusionment (Valdés, 2017, summarising Fishman).

This is worth holding in mind when thinking about weekend Punjabi classes. They do genuinely valuable work: script and literacy, formal vocabulary, Sikh history, and the not-small matter of putting a child in a room full of other Sikh children each week. Where children already speak some Punjabi at home, these programs compound quickly. Where they don’t, the class is starting closer to zero each week.

In short: The decisive hours are the ordinary ones at home. Classes multiply what the home produces; they cannot substitute for it.

Why Punjabi is harder than Mandarin — and it isn’t the children’s fault

There is a factor rarely named, and it has nothing to do with the diaspora.

Punjabi arrives in our homes already carrying a status wound.

Sociolinguistic research from Pakistan documents Punjabi being devalued as a language of the rural and uneducated, consciously abandoned by urban middle classes, and lacking prestige, power and institutional support (Gillani & Mahmood, 2014; Abbas et al., 2021).

Studies found educated parents in Lahore preferring Urdu with their own children rather than Punjabi (Rahman, 2016).

One study is titled after a participant’s own words: “I Am Not Punjabi, My Parents Are” (Acta Linguistica Asiatica, 2021).

A professor in Quetta, quoted in 2021 research, described barely encountering a fluent Punjabi-speaking student, observing that students were “really afraid” to admit belonging to Punjab.

Even in Indian Punjab, the dominant aspirational path for a generation has been English-medium schooling.

So when a Punjabi Sikh family lands in Toronto or Surrey or Birmingham, they are not starting from neutral. Many parents already carry an inherited association between Punjabi and backwardness, and English and arrival. That association does not need to be spoken to be transmitted. Children are extraordinarily good at detecting which language their parents are proud of.

This is the real asymmetry with Mandarin. Those parents are not pushing harder — they are offering a language their children can see has status in the world.

Researchers call the fix impact belief: parents who genuinely believe bilingualism is an asset, and that their own actions matter, are the ones who consistently follow through (De Houwer, 1999; King & Fogle, 2008).

In short: No technique overcomes a language the parents are quietly embarrassed by. This one is fixed at the level of visible pride.

What does not work

Nagging. When anxiety around language becomes intense, families may abandon heritage language maintenance entirely, believing that is the fastest route to relief (Sevinç, 2020). A systematic review found that heritage language management which feels too effortful produces shame, disappointment, frustration and tension at home (Hollebeke et al., 2020). Stephen Krashen described heritage learners as facing a “no-win” situation — expected to perform, and where failure is especially painful.

Using Punjabi only for discipline. If Punjabi is the language of “kine vaari kiha?” while English is the language of jokes, praise, comfort and bedtime, the child has been handed a precise emotional map of both languages. They will choose accordingly — and they will be right to.

Waiting until they are older. Lily Wong Fillmore’s study of 1,100 families found children aged 3 to 5 particularly vulnerable, with families whose children attended predominantly English preschools reporting negative shifts in home language patterns at several times the rate of others (Wong Fillmore, 1991). The window is early and it closes quietly.

Assuming comprehension equals ability. Many diaspora children are receptive bilinguals — understanding a great deal, producing almost nothing. This is a distinct category in the research, and it does not resolve on its own.

8 things that actually work

At a glance:

Change your own input pattern

Use tested conversation repair strategies

Protect two or three domains completely

Give grandparents a defined role

Design a Punjabi peer setting

Make Punjabi the language of warmth

Use short immersion bursts

Measure yourself, not your child

If you have limited energy, start with 3, 4 and 8 — they cost the least and move quickly. Numbers 1 and 5 carry the most weight, and both take longer.

1. Change the input pattern, not the instruction

If both parents can make Punjabi the household default, do it — that is the 97% row. If that is impossible, the next best documented pattern is one parent using Punjabi consistently and only, which lifts outcomes to around 74% from the 36% floor (De Houwer, 2007).

The point is not purity. Consistency substitutes for volume when volume is limited.

2. Use the conversation repair strategies that research has tested

Elizabeth Lanza identified parental discourse strategies that create a “monolingual context,” prompting a child to switch languages. These have been tested across multiple studies (Lanza, 1997; Juan-Garau & Pérez-Vidal, 2001; Döpke, 1992):

Minimal grasp — a warm, curious “Hmm? Ki keha?” instead of answering the English. This is not ignoring your child and not a punishment. It is a gentle signal that this conversation runs in Punjabi.

Expressed guess — offer the Punjabi version and invite confirmation: “Oh, tu keh riha si ki school vich bahut garmi si?”

Request for translation — “Punjabi vich kiddan kehnde?” — asked as a puzzle, not a test.

Parents who use these get more heritage-language production. Parents who move on or switch to English themselves get less. The mechanism is not discipline — it is giving the child a reason, in the moment, to reach for the Punjabi word.

3. Protect domains instead of trying to convert everything

Fishman’s insight was that threatened languages survive through compartmentalisation — owning specific functions completely rather than competing everywhere and losing everywhere.

Pick two or three and defend them absolutely:

The dinner table

The car

The fifteen minutes before bed

All conversation with Dadi ji

Everything else can be English. A narrow, non-negotiable domain beats a broad, constantly-renegotiated aspiration.

4. Build a language nest — with the grandparents already in the house

The most celebrated revitalisation success of the modern era is Māori kōhanga reo, launched in 1982. Fluent elders led the sessions, immersion was total, and parents used the language at home as well. It grew from five centres to over 300 within three years (Tāwhiwhirangi, in National Geographic, 2024; King, 2018). Crucially, many parents who were not fluent learned alongside their children.

Most Sikh families already have the essential ingredient in the living room. But grandparents need a role, not just proximity — and explicit permission to stop meeting grandchildren in broken English. If Bibi ji switches to English to be kind, that kindness is costing the child the language.

5. Design a Punjabi peer setting — carefully

Children adopt the language their friends use. Full stop. The food-court teenagers are not obeying parents; they are talking to each other in a language their whole friend group shares.

But there is a warning here, and it is the reason this strategy is so often attempted and so rarely works.

Peer groups converge on the language of least effort. Put five children together where three are comfortable in English and two are stronger in Punjabi, and the group will settle into English — nearly every time. The Punjabi speakers adapt downward, because at that age being fluent and funny matters far more than being correct. Gathering Punjabi children in a room does not create a Punjabi peer group. Left undesigned, it usually creates an English one.

Online spaces sharpen this. Multiplayer games, group chats and video calls are effectively English-only ecosystems, with vocabulary that has no everyday Punjabi equivalent. Do not fight for those domains. You will lose, and spend relationship capital losing. Hold the domains you can.

So the design matters far more than the gathering:

Match fluency, don’t mix it. Pair children of similar Punjabi ability. One strong speaker in a weak group adapts down; two comparable speakers hold the line.

Anchor with an older speaker. A fluent teenager, young adult or grandparent sets the norm. Children follow the most socially interesting person present, not the most correct one.

Give the language protected ground. Kirtan and sewa are the natural candidates — the activity is already rooted in Punjabi, so the norm does not have to be imposed from outside. Name those spaces explicitly as Punjabi-only, and hold them.

Don’t assume a camp is an immersion setting. In practice, instructors and children alike drift into English at camps and retreats, precisely because everyone wants the message to land. That is an understandable instinct and often the right call in the moment — but it means Punjabi has to be designed into specific, named activities rather than assumed across a whole program. One session genuinely held in Punjabi is worth more than a full camp that intended it and drifted.

The honest version: peer Punjabi has to be built, not hoped for. But when it holds, it is the most durable form of the language a child will ever have — because it is the only one not attached to authority.

6. Make Punjabi the language of warmth first

De Houwer’s work on harmonious bilingualism emphasises that conflict over language choice damages family wellbeing, while parental attitudes and discourse patterns drive successful development (De Houwer, 2015; 2020).

The translation: a child should meet their best moments in Punjabi. Jokes. Nicknames. Praise. Comfort. Songs. Silliness.

If a language only ever arrives attached to obligation, the child is not rejecting Punjabi — they are rejecting obligation.

7. Use immersion bursts — they work faster than most parents expect

A case study of an English–Bulgarian child documented genuine loss of Bulgarian production, followed by reactivation of the “lost” language after a ten-day trip to Bulgaria (Hristova-Nikolova, 2015). Ten days.

Trips to Punjab, extended stays with relatives, a summer with grandparents, a Gurmat camp that genuinely holds Punjabi as its working language — these are disproportionately powerful compared to a weekly class. Given limited resources, buy immersion rather than curriculum.

8. Measure yourself, not your child

For one week, don’t track whether your child spoke Punjabi. Track whether the adults did — honestly, in percentages, meal by meal.

Most parents who try this are startled. This is the only variable in the entire research literature you directly control, and it happens to be the one with the largest documented effect.

Understanding Gurbani: how spoken Punjabi builds the foundation

For Sikh families, there is something more here than culture. Every Punjabi word a child gains is a step toward understanding Gurbani for themselves — and that is a capability we can deliberately build.

It helps to see it as three skills that stack, each one making the next possible:

Decoding — recognising Gurmukhi akhar and reading aloud. This is what most children learn first, and it opens the door.
Comprehension — knowing what the words mean, which grows from a working Punjabi vocabulary of thousands of words. This is built mostly at home, in ordinary conversation, years before a child notices it is happening.
Interpretation — Gurbani grammar, Sant Bhasha vocabulary, the metaphorical register. This is its own study, and a rewarding one. Institutes such as SikhRI run dedicated Grammar of Gurbani courses precisely so English-speaking adults in the diaspora can take it up.

The Sikh Research Journal makes a useful point here: even those fluent in modern Punjabi go on to study further in order to reach the words, phrases, allusions and metaphors held in Sri Guru Granth Sahib Ji. This is not a discouragement — it is a lifelong invitation, and one that adults take up alongside their children.

So what does spoken Punjabi give a child?

It gives them a running start. A child with a few thousand Punjabi words comes to Sri Guru Granth Sahib Ji already prepared to meet it. The words are unfamiliar but not foreign. The grammar feels tilted rather than alien. A line opens partway, the child senses there is depth waiting there — and wants to go further. That wanting is the whole thing. Every stage after it is teachable.

This is why the ordinary Punjabi of the kitchen and the car matters so much. It is not separate from a child’s Sikhi; it is the ground it grows in. Guru Sahib gave us Bani in the languages people around them lived and spoke, so that its wisdom could be understood, questioned and lived. Each word we hand our children is a word they will one day recognise in front of the Guru.

And here is an advantage most communities would envy.

Nearly every community trying to sustain a language must invent a daily reason to use it. Sikh families do not. Nitnem is daily. Kirtan is weekly at minimum. Ardas, langar, sangat, camps — an entire ecosystem of scheduled, meaningful Punjabi use is already built into the rhythm of ordinary Sikh life.

In short: We are not short of infrastructure. The opportunity is to bring our children into it — speaking with them in Punjabi, not only around them.

Your four-week reset plan

Week 1 — Measure. Track only your own Punjabi use. Change nothing else.

Week 2 — Claim one domain. Choose dinner or the car, and make it Punjabi-only for the adults. Children may still answer in English. Do not correct them.

Week 3 — Add minimal grasp. “Hmm? Ki keha?” Warmly. Once or twice per conversation, not every sentence.

Week 4 — Recruit. Give grandparents an explicit job, and ask them not to switch to English. Set up one recurring Punjabi peer contact — cousins, a family, a camp.

Notice what is absent. There is no instruction to tell a child to speak Punjabi. Not once. The research does not support it, and it costs the very relationship it was meant to protect.

What is actually at stake

One finding deserves more attention than any statistic above.

Tseng and Fuligni surveyed over 620 adolescents from immigrant families. Teenagers who spoke a different language from their parents reported less cohesion and less discussion with both mother and father. The highest levels of closeness and real conversation were reported by adolescents who communicated with their parents in the heritage language (Tseng & Fuligni, 2000).

Later research replicated the pattern. Studies of Chinese American and Chinese Canadian families link parent–child language mismatch to reduced parental warmth, more family conflict and higher adolescent depressive symptoms — concluding that adolescents who cannot engage in meaningful, reciprocal conversation with their parents are less likely to feel supported by them (Costigan & Dokis, 2006; Weaver & Kim, 2008).

Home language is not only a cultural question. It is a measurable predictor of whether a family can still talk to each other when it matters most.

What is quietly at stake in the daily choice of language is this: the possibility that at seventeen, when something is genuinely wrong, a child has the words to tell their parents — and that at eighty-five, a grandmother is fully understood by her own grandchild.

The correction is not a curriculum. It is not another program. It starts with the least dramatic act available to any parent: the next time your child speaks to you, answer them in Punjabi.

Frequently asked questions

My child understands Punjabi but refuses to speak it. Is that normal? Yes, and it has a name: receptive bilingualism. It is very common and it does not resolve on its own. The research-backed response is not more pressure but more input, plus the repair strategies in section 7 above.

Will speaking Punjabi at home hurt my child’s English? No. In De Houwer’s study of 1,899 bilingual families, every single child spoke the majority language regardless of the home pattern. Separate research on bilingual preschools found no loss in the home language and greater gains in English over time (Winsler, Díaz, Espinosa & Rodríguez, 1999). The majority language takes care of itself; the heritage language does not.

Is it too late if my child is already 12? Harder, not too late. The strongest window is early (Wong Fillmore, 1991), but the immersion research shows meaningful reactivation in adolescents and adults. At this age, motivation and peer context matter more than parental instruction — focus on sections 5 and 6.

My spouse doesn’t speak Punjabi. Are we stuck? No. The pattern where one parent uses the heritage language consistently and only produced around 74% success — roughly double the mixed pattern’s 36%. Consistency from one parent is worth a great deal.

If I get Punjabi-speaking children together, won’t they just speak English to each other? Very likely, yes — unless the setting is designed. Mixed-fluency peer groups converge on English because it is the language of least effort, and the stronger Punjabi speakers adapt downward rather than stand out. Match children by ability, include an older fluent speaker, and choose activities that genuinely cannot be done in English. Gathering alone is not enough.

Should I correct my child’s Punjabi mistakes? Sparingly. Correction raises anxiety, and anxiety is associated with families abandoning the language altogether. Instead of correcting, recast: repeat what they said, naturally, in correct Punjabi, and keep the conversation moving.

Is weekend Punjabi class enough on its own? It is valuable but not sufficient. A few hours a week supplements a child’s linguistic life rather than substituting for it. Classes compound powerfully where the home is already producing Punjabi.

Should we teach Gurmukhi script or spoken Punjabi first? Speaking and listening generally come first. Script without vocabulary produces accurate reading with limited comprehension. Building the spoken base gives a child something to attach the letters to.

References

Abbas, F., et al. (2021). “I Am Not Punjabi, My Parents Are”: Degradation of the Language of Dominant Majority. Acta Linguistica Asiatica.

Canadian Language Museum. (2022). Ups and Downs in Canada’s Languages: Census 2021.

Chen, X., & Zhang, Y. (2021). Heritage language maintenance of Chinese immigrant families: Perceptions, practices, and challenges.

CIC News. (2022). Statistics Canada report on language diversity due to immigration in Canada.

Costigan, C. L., & Dokis, D. P. (2006). Relations between parent–child acculturation differences and adjustment within immigrant Chinese families. Child Development.

De Houwer, A. (1999). Environmental factors in early bilingual development: The role of parental beliefs and attitudes. In G. Extra & L. Verhoeven (Eds.), Bilingualism and Migration. Mouton de Gruyter.

De Houwer, A. (2007). Parental language input patterns and children’s bilingual use. Applied Psycholinguistics, 28, 411–424. doi:10.1017/S0142716407070221

De Houwer, A. (2015). Harmonious bilingual development: Young families’ well-being in language contact situations.

De Houwer, A. (2020). Harmonious bilingualism: Well-being for families in bilingual settings. In A. C. Schalley & S. A. Eisenchlas (Eds.), Handbook of Home Language Maintenance and Development. De Gruyter Mouton.

Döpke, S. (1992). One Parent, One Language: An Interactional Approach. John Benjamins.

Environics Analytics. (2022). Census 2021: Canada’s Linguistic Diversity.

Fishman, J. A. (1991). Reversing Language Shift. Multilingual Matters.

Fishman, J. A. (Ed.). (2001). Can Threatened Languages Be Saved? Reversing Language Shift, Revisited. Multilingual Matters.

Gillani, M., & Mahmood, M. A. (2014). Punjabi: A tolerated language, young generations’ attitude. Research on Humanities and Social Sciences, 4(5), 129–137.

Hollebeke, I., Struys, E., & Agirdag, O. (2020). Can family language policy predict linguistic, socio-emotional and cognitive child and family outcomes? A systematic review. Journal of Multilingual and Multicultural Development.

Juan-Garau, M., & Pérez-Vidal, C. (2001). Mixing and pragmatic parental strategies in early bilingual acquisition. Journal of Child Language, 28(1), 59–86.

King, J. (2018). Māori: Revitalization of an endangered language. In The Oxford Handbook of Endangered Languages.

Koh, P. W., Chen, X., et al. (2021). Early Chinese heritage language learning in Canada. System.

Lanza, E. (1997). Language Mixing in Infant Bilingualism: A Sociolinguistic Perspective. Oxford University Press.

National Geographic. (2024). The Māori saved their language from extinction. Here’s how.

Sevinç, Y. (2020). Anxiety as a negative emotion in home language maintenance and development. In Handbook of Home Language Maintenance and Development.

Sharma, D., & Sankaran, L. Intergenerational transmission of laterals in Punjabi–English heritage bilinguals. In The Phonetics and Phonology of Heritage Languages. Cambridge University Press.

Sikh Research Institute. Grammar of Gurbani course materials. sikhri.org

Sikh Research Journal. On comprehension of Gurbani across displacements of time and location. sikhresearchjournal.org

Tseng, V., & Fuligni, A. J. (2000). Parent–adolescent language use and relationships among immigrant families. Journal of Marriage and Family, 62(2), 465–476.

Valdés, G. (2017). From language maintenance and intergenerational transmission to language survivance. International Journal of the Sociology of Language, 2017(243), 67–95.

Verdon, S., McLeod, S., & Winsler, A. (2014). Language maintenance and loss in a population study of young Australian children. Early Childhood Research Quarterly.

Wang, Y., et al. (2026). “We don’t want to force it”: Emotional well-being as the organising principle in family language policy among Chinese families in The Netherlands. Journal of Multilingual and Multicultural Development.

Winsler, A., Díaz, R. M., Espinosa, L., & Rodríguez, J. L. (1999). When learning a second language does not mean losing the first. Child Development.