At the playgroup there is always one child who seems to be narrating the room. Truck. Duck. Shoe. Ball. Their parent glows a little, the way you glow when something you made is working. And then there is your child, who has walked up to every adult in the room, waved, and said “hi” with the confidence of a maître d’, who tells the dog “night-night” at bedtime and announces “all done” with a flourish, but who has not, as far as you can tell, ever named a single object. You count on the drive home. Hi. Bye. Uh-oh. All done. No. It feels like a list of manners, not a vocabulary. And underneath the counting is the question nobody says out loud in the parking lot: is my kid behind, or are they just doing this differently?

Two children, two first vocabularies

In 1973 the psychologist Katherine Nelson published a monograph that followed eighteen children from their first ten words to their first fifty. She expected the lists to look roughly alike. They didn’t. One group of children filled their early vocabularies with the names of things: ball, dog, cup, shoe, car. Nelson called these children referential. Another group spent their first fifty words on something else entirely: hi, bye, no, stop it, go away, I want it, thank you. Their words were about people and the traffic between people. She called this style expressive.

The referential children had, mostly, been playing a naming game with their parents. The expressive children’s words looked like a different game altogether, one about managing the social world: getting attention, refusing, greeting, closing a scene. Both groups were learning language. They had simply walked in through different doors.

What the expressive child is actually doing

Here is the part that changed how researchers think about the finding. When Elena Lieven, Julian Pine and their colleagues looked closely at expressive children’s vocabularies in the early 1990s, they noticed that many of the items weren’t single words at all. “Allgone,” “whassat,” “Iwantit,” “night-night” were being produced as single, unanalysed chunks: frozen phrases, stored and used as one unit the way you might store a melody without knowing its notes.

This is a legitimate way to learn. Usage-based accounts of language acquisition, most associated with Michael Tomasello, argue that all children start with item-based chunks and only gradually discover that the pieces can be pulled apart and recombined. Some children just do more of their early learning in chunks. The crucial point in Pine and Lieven’s work is what happens next: the frozen phrases thaw. A child who has been saying “more juice” as one block starts saying “more apple,” “more book,” “more Daddy.” The slot has been discovered. The phrase has become a frame with a hole in it, and the hole is where grammar begins to grow.

So the child who says “hi” to everyone is not avoiding language. They are learning the social frames first and the labels second. The child naming trucks is learning labels first and will pick up the frames later. The work of Elizabeth Bates and her colleagues on this point is reassuring in a specific way: style is a dimension, not a diagnosis. Almost no child is purely one or the other, and children of both leanings arrive at ordinary sentences by the time they are three.

Where the style comes from

It is tempting to read your child’s style as a verdict on your parenting. The evidence is gentler and more interesting than that. Nelson and later researchers did find that parents of referential children tended to label more, while parents of expressive children used more directives and social routines. But this is a correlation, and it almost certainly runs both ways. A child who lights up at “peekaboo” and “bye-bye” gets more of it. A child who keeps pointing at the dog gets told “dog” forty times a day. You are both tuning to each other.

Temperament plays a role too. A sociable toddler finds greetings rewarding in themselves, and a word that produces a smile from every adult in a room is a word worth saying often. Birth order nudges the balance as well: Pine found that second-born children tended to carry more frozen phrases and personal-social words in their early vocabularies, which makes sense when much of what you hear is aimed at an older sibling. None of this is destiny. It is a starting lean.

One more thing worth saying plainly. Online you will meet the phrase “gestalt language processor” used to describe children who learn in chunks, sometimes with the implication that they need a special approach. The behaviour is real and well documented. The research term is holistic or expressive style, and the long-term studies treat it as a normal variation in how children get into language, not a condition.

When it is not just style

Style explains the shape of a vocabulary. It does not excuse a vocabulary that isn’t growing. Clinicians commonly use a rough threshold: fewer than about fifty words of any kind at twenty-four months, or no two-word combinations by then, is the point to seek an evaluation. Count the phrases. “All done” counts. “Night-night” counts. If the list is growing month on month, and your child understands far more than they say, points, follows simple directions and uses words to get you to do things, the style is almost certainly just a style. If the list is stuck, or comprehension seems thin, call a speech-language pathologist. Early evaluation costs you an afternoon. Waiting costs you time you cannot get back.

Your next moves

  • Count the phrases as words, today. Write down everything your toddler says with consistent meaning, including hi, bye, uh-oh, all done, night-night and no. That is the real vocabulary, and it is almost certainly longer than the list you were keeping in your head.
  • Hang a noun on the greeting. When your child says “hi” to the dog, answer “Hi, dog!” When they wave “bye-bye” at the bath, say “Bye-bye, water.” You are attaching a label to the frame they already love, which is the easiest place for a label to stick.
  • Give one frozen phrase a wobble. Pick a chunk they use, such as “more juice,” and change exactly one piece in your own speech: “more banana,” “more bubbles.” Keep the rest identical. You are showing them where the slot is.
  • Play the referential game for three minutes a day. Sit with a basket of five everyday objects and simply name each one as you hand it over, then wait. No quiz, no “say ball.” Just the word and the thing, repeated across the week.
  • Book the evaluation if the count is stuck. If your child is near two and the total list, phrases included, is under about fifty, or two-word combinations haven’t appeared, call a speech-language pathologist this week. You can always cancel if the words arrive first.

A small daily door into naming

Acorn was built around the referential game: three minutes a day of one word and one thing, with no ads and no upsells. If your toddler walked in through the social door, a short daily session like this adds the labels they aren’t getting from their own favourite routines, without turning your kitchen into a flashcard drill. And if they are already a confident namer, you can skip it entirely and spend your three minutes on the frames instead. Either way, the point is the same: there are two doors into language, and your child has already found one. If you’d like a gentle way to open the other, you can try Acorn at https://acorn.lumenlabs.works