Talking to Your Baby: The Research on Early Language Development in Bilingual HK Families

My husband is narrating a nappy change.

"Okay, we're lifting the legs now. Left leg, right leg. Look at those legs. Those are your legs. What a nappy. This is a very full nappy. We'll put a clean one on. Clean nappy, here we go."

He is doing this in Cantonese. This is not a man who normally narrates his daily activities. Parenthood has changed him.

He is doing exactly the right thing. I have told him so, which he seems to find both pleasing and faintly ridiculous given that we are standing over a changing mat discussing the legs of a person who cannot respond.

The research on infant-directed speech

"Parentese" — the slightly higher-pitched, slower, more melodic speech that adults naturally direct at infants — is not silliness. It is developmentally functional. Infants as young as two months old preferentially orient to infant-directed speech over normal adult speech. The exaggerated intonation, slower pace, and simpler grammatical structures all help infants extract the patterns of their language.

The landmark research by Patricia Kuhl at the University of Washington shows that language learning in the first year is statistical: babies are extracting the phonological probabilities of whatever language they're hearing. By around eight months, they begin specialising in the sounds of their home language and losing sensitivity to sounds from other languages. This process is experience-dependent — the more language-rich input, the better the statistical model the baby builds.

The "30 million word gap" research by Hart and Risley — later updated and somewhat contested in terms of exact figures, but supported in direction — found that children from language-rich environments hear far more words in the first three years than children from language-poor environments, and this predicts vocabulary and reading outcomes years later.

What matters most: the quantity of contingent, back-and-forth verbal interaction. Not passive exposure (TV doesn't count for this purpose), but responsive conversation — even with someone who can't yet respond. You talk, they babble, you respond to the babble as if it were speech, they learn that communication is a two-way process.

The HK multilingual question

The question I hear most often from HK parents: with Cantonese at home, Mandarin emerging as important, and English valued for education and career — which language should I use with my infant?

The research answer: it doesn't matter. What matters is that the input is meaningful, emotionally warm, and from a real human who is engaged with them.

Infants can handle multiple language inputs simultaneously without confusion. The idea that multilingual input causes language delay is a persistent myth I will address in a dedicated article, but briefly: children exposed to multiple languages from birth learn each language slightly more slowly than monolingual peers, but by school age they are generally fully proficient in their home languages, and they gain significant cognitive benefits from bilingualism that continue throughout life.

There is one nuance worth noting: if a parent speaks one language as their dominant, natural language and another as a second or third, the emotional richness and spontaneity of interaction will typically be stronger in the native language. Speaking to your baby in a language you feel comfortable and expressive in is likely more valuable than speaking in a language you find stilted, even if the second language is considered more prestigious.

What we actually do

My husband speaks to our children primarily in Cantonese. I mix — mostly Cantonese at home, but when I'm in researcher mode, sometimes English comes out because it's the language I read and write in professionally. My mother-in-law speaks Cantonese. Our helper speaks Tagalog to her own family and English to us and to the children.

My daughter, who is nearly two, has words in all three. Her first word was 媽媽, which my husband and I are still quietly competing over credit for. Her word for our helper involves a sound that might be approaching "Maya" or might be something entirely of her own invention.

What I notice: she talks more with people who respond to her talking. This seems obvious but it matters. When I put my phone down and actually face her and respond to every syllable she produces, her vocalisations multiply. When I am distracted and give partial responses, she tries harder briefly, then redirects to something else.

She is teaching me to be more present in every conversation. The research predicted this. It's still surprising, living it.

Practical points

Narrate your activities. You are not talking to yourself — you are providing a language-rich environment while also doing the dishes. "Here's the soap. We're washing our hands. Cold water, warm water. Now we're drying them."

Respond to babble. Treat it as a turn in conversation. Wait. Let them "finish." Respond. This teaches the back-and-forth structure of language before words arrive.

Read aloud, daily, from as young as you can manage. Point at pictures. Ask questions, even to someone who can't answer. The research on shared book-reading and vocabulary development is remarkably consistent.

Don't stress about which language. Stress about whether you are actually talking.

My husband has added the narrative commentary to bath time as well. Last week it included a detailed description of rubber duck properties.

Our son thinks this is wonderful. I have to agree.