It usually starts within a few weeks of school. You ask a question in your language, and the answer comes back in the school language. You ask again. Same result. Within a couple of months your child understands everything you say and answers almost nothing in kind, and the family language has quietly become something that only travels in one direction.
This is one of the most common turns in bilingual family life, and it is also one of the most misread. Parents interpret it as rejection, or as their own failure to be consistent enough. The research points somewhere less personal: a child is doing an efficient calculation about which language gets things done, and the calculation shifts the moment school starts. Here is what the data shows about how often this happens, what it means, what makes it worse, and what actually brings the language back into use.
Why does my child suddenly refuse to speak our language?
Because the balance of usefulness changed, not because the language lost value in their eyes. Once school starts, the majority language becomes the language of friendship, status, teachers, playground negotiation and almost every new word the child learns. The home language stays confined to a small circle of speakers and a narrow set of everyday topics.
Researchers describe exactly this asymmetry: the heritage language is typically used at home with few speakers on routine topics, while the majority language is used with many speakers, in many contexts, for many purposes. Add the social pressure children pick up fast about which language is normal in public, and answering in the school language is simply the path of least resistance. The blog Language on the Move, run by sociolinguists, has documented the pattern in family after family.
How common is it for a child to stop using the home language?
Common enough to be measured at population scale. A study by Verdon, McLeod and Winsler using the Longitudinal Study of Australian Children tracked children who spoke a language other than English at the first wave of data collection. By the second wave, 91.5 percent were still using it. By the third wave, at ages four to five, 86.6 percent were.
Put the other way round: roughly one child in seven had stopped using the home language before the end of the preschool years, and that is before school-age pressure really arrives. The paper is published in Early Childhood Research Quarterly, and an accepted manuscript is freely available from Charles Sturt University.
Is my child losing the language, or just not using it?
Almost always the second, at first. A child who understands everything and answers in another language is a receptive bilingual, and receptive knowledge is a real asset, not a consolation prize. The grammar, the sound system and a large vocabulary are all still there. What is missing is production practice, and production is the part that decays fastest without use.
The distinction matters because it changes the timeline. Reactivating a language a child still understands is a matter of months of consistent use, not years of relearning. What you want to avoid is the slide from "does not speak it" to "no longer follows a fast conversation", because that is where the input starts dropping too and the loop becomes self-reinforcing. Research on heritage language development describes this drift clearly: as children socialise more in the societal language, exposure to the home language falls, and attrition follows the drop in input.
Does insisting work?
Rarely, and it carries a cost worth knowing about. Refusing to respond until the child switches, correcting every sentence, or making the language a condition of getting what they asked for tends to make the home language feel like an obstacle placed between the child and their parent. Children respond by using it even less.
The stakes are not only linguistic. Research on what is called shared language erosion, where a child's majority language races ahead while the parents' does not, links the widening gap to weaker parent and child communication and more conflict. The mechanism is simple: the conversations that need nuance become the ones nobody can have properly. That analysis is open access, and a broader review of language, culture and adaptation in immigrant children reaches similar conclusions. Preserving the relationship is the higher priority, and it also happens to be the better language strategy.
What actually keeps a home language alive?
Reasons to use it, and more than one person to use it with. Those two levers do more than any amount of insistence. A 2025 study in Frontiers in Psychology on resilient heritage language maintenance found that maintenance rests on the interplay of family practice, cultural connection and pragmatic choices rather than on parental determination alone. The full text is open access.
| What usually fails | What tends to work | Why |
|---|---|---|
| "Say it properly in our language" | Answer their content, model the sentence back naturally | Keeps the conversation going instead of ending it |
| Correcting each mistake | Recasting: repeat what they said, correctly, and move on | Gives the correct form without turning talk into a test |
| Home language only for chores and instructions | Home language for the fun slots: jokes, stories, games, video calls | Links the language to reward rather than to obligation |
| One parent as the only speaker | Grandparents, cousins, community groups, holidays | More speakers means more contexts and higher stakes |
| Waiting for the child to initiate | Fixed daily anchors: breakfast, bath, bedtime story | Predictable slots protect input when everything else is busy |
Why does the number of speakers matter so much?
Because a language spoken by one person is a language a child can route around. Research on heritage language proficiency has found a distinct "number of speakers" effect: children exposed to a heritage language by more people reach higher proficiency than children with equivalent hours of exposure from a single speaker. Variety of voices, accents, vocabulary and conversational styles does work that repetition from one person cannot.
Annick De Houwer's large survey of 1,899 bilingual families, published in Applied Psycholinguistics, pointed the same way: home input patterns where both parents use the minority language give the highest chance that the child will use it too. We looked at what that means when only one parent speaks the language in a separate article on single-speaker households. The number of speakers study is open access and the De Houwer paper is on Cambridge Core.
What if I am the only speaker in the house?
Then the practical problem is not motivation, it is arithmetic: one adult cannot generate enough varied conversation on a working weeknight to counterbalance seven hours of school. The answer is to widen the circle rather than to push harder, and most of that widening is low-tech. A standing weekly video call with a grandparent in the other country is worth more than an hour of drilling, because the child has to be understood by someone who genuinely cannot switch.
Some families use a talking companion to hold one of the daily anchors when the parent is not available for it. Ted, our interactive plush, holds a real conversation in more than sixty languages and can be set by the parent to speak only the home language, which turns the bedtime story or the twenty minutes before dinner into production practice rather than passive listening. It is not a substitute for a human speaker and we would not pretend otherwise: the research is unambiguous that people, and specifically more than one person, are what carries a language. What it can do is keep the language in daily circulation on the evenings when nobody else is available. Parents who want to know what is recorded, stored or shared before adding anything that listens will find it set out on our security and certifications page.
One more thing worth holding on to: the amount of exposure needed is smaller than most parents fear, and we broke the numbers down in our piece on how many hours a week a child actually needs. Refusal at five is not the end of a bilingual childhood. It is a phase with a known shape, and it responds to reasons far better than it responds to pressure.
Frequently asked questions
Why does my child answer me in a different language?
Because the school language has become the language that gets the most done in their daily life, and they know you understand it. This is a use pattern rather than a loss of ability, and it typically appears within the first months of school entry.
Will my child forget our language completely?
Not quickly, and not inevitably. A child who still understands you retains the grammar, the sounds and much of the vocabulary, and reactivation usually takes months rather than years. The risk grows only if reduced speaking leads to reduced listening, because that is when input falls too.
Should I refuse to answer unless my child uses our language?
Most researchers and clinicians advise against a hard refusal, because it tends to make the home language feel like a barrier between parent and child. Recasting works better: answer what they said, repeat it naturally in the home language, and keep the conversation moving.
How much exposure does a child need to keep a second language?
Studies on heritage languages suggest proficiency drops sharply when the language accounts for a small share of a child's daily input, with problems concentrated well below a quarter of exposure. The number of different speakers matters as much as the number of hours.
Is it too late if my child stopped speaking our language years ago?
Usually not. Receptive knowledge is durable, and adults who spoke a language in early childhood relearn it faster than beginners. The realistic goal is often a comfortable conversational level rather than the fluency of a child raised entirely in the language.