You speak Portuguese. Your partner does not. You started out determined, talking to the baby in Portuguese every day, and for a while it was working. Then your child turned four, started school, and began answering you in English every single time. You keep going. They keep switching. And you start wondering whether the whole thing is quietly failing.
This is the most common bilingual family setup in the world and the hardest one to make work. It is also the one where the research is most useful, because the numbers tell you exactly where the difficulty sits and what actually moves it. The short answer is yes, it is possible. The longer answer is that the strategy most parents are told to use is not the one with the best odds.
What are the actual odds when only one parent speaks the language?
Lower than most parents expect, and this is worth knowing early rather than discovering at age seven. In the largest study of its kind, only about 36% of children became active speakers of the minority language when just one parent used it and the other used the community language.
That figure comes from Annick De Houwer's survey of 1,899 families in Flanders, published in Applied Psycholinguistics. It remains the reference point for this question because of its size: most bilingualism studies work with a few dozen families.
Look at the top of the chart and the pattern is obvious. When both parents use the minority language at home, 93% to 97% of children end up speaking it. The variable that matters most is not method. It is how many people in the house use the language.
Is one parent one language the best strategy?
It is the most famous strategy, and it is not the strongest one. In the same dataset, strict one parent one language produced active bilinguals in about 74% of families, while the approach where one parent used both languages did slightly better at 79%.
That surprises most people, because OPOL is usually presented as the gold standard. The likely reason is simple arithmetic: when a parent is willing to use the minority language even in mixed company, the child hears more of it and hears it in more situations. Strict OPOL can accidentally confine a language to one person and one set of contexts, which makes it easy for a child to file it under "Mum's language" rather than "a language".
None of this means OPOL is a mistake. A 74% success rate is genuinely good. It means the rigid version of it is not worth defending at the cost of exposure.
How much exposure does a child actually need?
Enough that the language is part of daily life rather than a visit. Researchers studying bilingual acquisition generally require children to have at least 20% to 30% of their waking hours in a language before they count as growing up with it, which works out at roughly 15 to 25 hours a week.
For catching up to monolingual peers on vocabulary, the bar is higher. Elin Thordardottir's work with Montreal five year olds, published in the International Journal of Bilingualism, found that children with 40% to 60% exposure to a language performed like monolinguals on receptive vocabulary. Below that, scores dropped. And expressive vocabulary, actually producing the words, needed more exposure than comprehension did.
This is the number that explains the frustration in most one-parent households. Your child understands you perfectly and answers in the other language. That is not defiance and not failure. It is the predictable signature of an exposure level that is high enough for comprehension and not yet high enough for production. Our piece on how much daily bilingual practice actually helps goes deeper on the hours question.
What can you do if your partner does not speak the language?
Stop trying to be the only source and start building a second and third one. The De Houwer numbers say the constraint is the number of voices, so that is the thing to attack.
| What to add | Roughly how much | Why it works |
|---|---|---|
| Grandparents or relatives on video call | 2 to 3 hours a week | Adds a second voice and a real reason to speak it |
| A weekend playgroup or Saturday school | 2 to 4 hours a week | Peers make the language socially normal, not parental |
| Audio, songs and stories in the language | 30 to 60 minutes a day | Cheap volume, and it works during the commute or breakfast |
| Your partner learning 20 phrases | Everyday use | Signals the language belongs to the household, not one parent |
| A holiday in a place where it is spoken | 1 to 2 weeks a year | Nothing else produces the same jump in willingness to speak |
That last point about your partner is the most underrated. A non-speaking partner who learns to say good morning, sit down and I love you in the language changes its status in the house entirely. They do not need fluency. They need to be visibly on the team.
The one-line version: the odds are set less by which method you follow than by how many people your child hears using the language, so spend your effort adding voices rather than enforcing rules.
What if your child refuses to speak it back?
Expect it around school age, and do not treat it as a verdict. When a child starts school the majority language gains enormous social weight, and answering in it is the path of least resistance rather than a rejection of yours.
Three things help, none of them pressure. Keep speaking it yourself without demanding replies, because comprehension is a real asset and it keeps the door open. Use gentle prompts rather than corrections, such as repeating their answer in your language and moving on. And find one context where the language is simply the only option, whether that is a grandparent who does not speak the majority language or a week with cousins abroad.
What tends not to work is refusing to understand, or making the language a condition of getting something. It converts a relationship into a transaction, and children opt out of transactions. If you also worry that two languages might be slowing your child's speech down, the evidence on that is reassuring and we covered it in does learning two languages delay speech.
Where do books, audio and talking toys fit?
They are volume, not substitute. No recording gives a child what a responsive human conversation gives, because the value is in the back and forth. But when you are one adult trying to supply an entire language, volume is exactly the thing you are short of.
Passive audio has real limits. Songs and audiobooks build vocabulary and accent but do not push a child to produce language, and production is the harder half. Anything that gets a child talking back has an advantage here.
That is the gap Ted, our interactive plush companion, was designed for. It holds a real conversation in over 60 languages, adapts to your child's level, and you choose the language and topics in the app, so it can speak your family's second language when nobody else in the house can. It is not a replacement for the grandparent on the video call, and we would not claim it is. It is a way of adding minutes on the days when the grandparent is asleep in another time zone. Parents can see what is stored and how the controls work on our FAQ page.
Frequently asked questions
Can a child become bilingual if only one parent speaks the language?
Yes, but the odds are the lowest of any family setup. Around 36% of children become active speakers when one parent uses the minority language and the other uses the community language, compared with 93% to 97% when both parents use it at home. Adding other speakers is what closes the gap.
How many hours a week does a child need to learn a second language?
Researchers generally treat 20% to 30% of waking hours, roughly 15 to 25 hours a week, as the minimum for genuine bilingual acquisition. Matching monolingual peers on vocabulary usually requires 40% to 60% exposure, and speaking needs more exposure than understanding.
Is one parent one language the best method?
Not necessarily. In a study of 1,899 families, strict one parent one language produced active bilinguals in 74% of cases, while one parent using both languages reached 79%. Total exposure and the number of speakers matter more than the method itself.
Why does my child understand but refuse to speak the language?
This is called receptive bilingualism and it is very common once school begins. It usually reflects exposure that is sufficient for comprehension but below the higher threshold needed for production, combined with the social pull of the majority language. It is often reversible with more exposure and a context where the language is the only option.
At what age is it too late to start a second language?
It is not too late in early childhood. Younger children reach native-like pronunciation more easily, but the amount of exposure matters more than the starting age, and older children often learn grammar and vocabulary faster. The practical answer is to start now with whatever hours you can find.
Further reading: Barbara Zurer Pearson's chapter on children with two languages, and research on quantifying language exposure in infants and children.