Will Two Languages Confuse My Child? What the Research Actually Says

Somebody with a clipboard told you to stop.

Maybe a health visitor, maybe a teacher, maybe a well-meaning relative who had been in the country longer than you. The phrasing varies and the message does not: your child is behind, they are hearing too much, pick one language, and make it English.

A great many diaspora parents have received that advice. Most of them took it, because it arrived with authority attached and because no parent wants to gamble with their child's speech.

The advice was wrong. Not a matter of opinion, not "the research is mixed." Wrong.

---

---

But here is what nobody tells you next

The delay myth is a distraction. While parents were busy worrying about it, the heritage language was disappearing anyway — through a mechanism that has nothing to do with confusion and almost nothing to do with the early years.

Heritage languages are rarely retained past three generations. The pattern repeats across continents and communities with remarkable consistency: the grandparents speak it, the parents understand it, the grandchildren do not.

And here is the finding that should change how you think about this. Research on second-generation children has found that the large majority understand their heritage language — but only a minority actually speak it with their own siblings.

Sit with that. The language is not usually lost by being forgotten. It is lost in the gap between understanding and speaking.

A child who understands everything and answers in English looks, to a proud parent, like a bilingual child. They are not, quite. They are a child with an excellent receptive grasp of a language they have decided not to produce — and production is the thing that gets passed to the next generation. Comprehension dies with them.

Why the child stops speaking

It is almost never about capability. It is social.

Somewhere between five and seven, most children work out which language has social currency in the world they are actually living in. The playground, the classroom, the screen, the friends. The heritage language becomes the language of the house and the phone calls, and the house is a smaller world than it used to be.

Sociolinguist Joshua Fishman, whose framework for measuring language shift underpins most of this field, identified the tipping point plainly: shift accelerates sharply once the heritage language stops being the main language spoken at home. Not when it stops being spoken. When it stops being the default.

Siblings are the hinge. Once two children start speaking English to each other, the volume of heritage language in that house collapses, because siblings talk to each other far more than they talk to you.

The arithmetic, roughly

A widely cited rule of thumb among bilingualism researchers holds that a child needs something in the region of a third of their waking hours in a language to become an active speaker of it rather than just a comprehender. Treat that as a rough shape rather than a precise threshold — real children vary enormously, and the quality and variety of what they hear matters as much as the raw hours.

But run your own numbers anyway, because the shape is sobering.

A four-year-old in childcare or preschool in an English-speaking country is getting most of their waking language in English before anyone has done anything wrong. The heritage language is fighting for the breakfast table, the car, the bath, and bedtime. That is not a failure of effort. That is the arithmetic of living abroad.

Which means the question is not "am I doing enough." It is "where can I find more hours, and more people."

What the research says actually works

One study of families using the one-parent-one-language approach — Susanne Döpke's One Parent — One Language: An Interactional Approach (John Benjamins, 1992) — looked at which families succeeded, and found two things separating them from the rest.

The first: the parents were consistent, and — this is the uncomfortable one — they did not let the child answer in the other language. When the child replied in English, the parent did not simply accept it and carry on. Not harshly. Just persistently: a look, a repeat, a wait.

The second, and the bigger one: the children had people other than their parents to speak it with. Other relatives, neighbours, a community, a religious group, other children. Heritage language retention succeeds when the language is used in more than one context, with more than one person, for more than one purpose.

You cannot be an entire language community by yourself. Nobody can. The parents who succeed are not the ones who try hardest — they are the ones who found a second and third context.

What this means if the language is Swahili

If you are raising a child in Swahili outside East Africa, you likely do not have a Swahili school nearby, or a Saturday programme, or a street where it is spoken. So the contexts have to be built.

Give the grandparents a job, not a greeting. A weekly video call where the child is asked to do something in Swahili — report on their week, name what is on their plate, ask their grandmother a question they prepared — produces speech. A call where the child is greeted and then handed back produces comprehension and nothing else.

Find one other child. One. A cousin, a family friend, anyone their age they can be silly with in Swahili. A single peer is worth more than an hour of drilling, because it moves the language out of the authority relationship and into play, which is where children actually produce language.

Books they can say, not just hear. This is where the understanding-speaking gap gets closed for young children. A book with very simple, highly repetitive Swahili — the same few structures returning page after page — gives a two- to seven-year-old lines they can produce, not just follow. That is a different cognitive act, and it is the one that matters. Repetition looks boring to the adult reading it. It is the entire mechanism for the child.

Do not fight the sibling shift with rules. You will lose. Change what the language is for instead. If Swahili is the language of jokes, food, songs and the people they love most, it has a use. If it is the language of being corrected, it has none.

Protect one daily slot completely. Bath, breakfast, the walk to school — one window where the language is simply not negotiable and nobody has to decide each time. Consistency beats intensity, and the decision fatigue is what kills most families' efforts, not the difficulty.

What you are actually protecting

There is a tendency to treat heritage language loss as a sentimental problem — the sadness of grandparents who cannot talk properly to their grandchildren. That is real, and it is not the whole of it.

Researchers including Ellen Bialystok at York University have found that children who actively maintain two languages show measurable advantages in attention and cognitive flexibility over monolingual peers. Not confusion. The opposite of it.

And there is the thing research is poor at measuring. A child who can speak to their grandmother in her own language has an unmediated relationship with her. No translating parent in the middle, no summarised version, no jokes that do not survive the trip. They get the actual person.

That is what the clipboard was asking you to give up.

---

Your turn. Were you told to drop your language — and did you? I would like to know who told you, and what happened next. And if you are in the middle of the sibling shift right now, say so below, because you are not the only one and the comments are more useful than the article.

If you are choosing a name as well as a language, our guide to authentic Swahili names and their meanings gives the root of each one. And if pronunciation is the barrier for the English speakers around your child, start here.

Photograph: Barbara Olsen / Pexels — a mother crouched face to face with her toddler, mid-conversation.