Lead
"Our child still isn't talking." Somewhere past the first birthday, that sentence takes up residence in the back of a parent's mind. On the walk home from the playground, where a same-age child was saying "woof-woof" and "vroom," a wave of anxiety arrives without warning. Open social media and there it is — a post about a thirteen-month-old already stringing two words together. Come home, watch your own child, and there's still just "aah" and "ooh."
This anxiety does not come from a failure of attentiveness. It comes, in fact, from paying close attention to a child's development. But before letting that attention tip into urgency, there are a few things worth thinking through carefully.
This article covers the medical range for first words, the difference between expression and comprehension, the social signals that precede language, and what it means to choose not to rush.
The window for first words is wider than most parents expect
Start with the numbers.
The normative data from the MacArthur-Bates Communicative Development Inventory (CDI) — drawn from parent reports on 1,803 children aged 8 to 30 months — plots expressive vocabulary development at the 10th, 25th, 50th, 75th, and 90th percentiles [1]. The median for reaching 10 expressive words falls around 13 months; the median for 50 words falls around 17 months. At the same age, the gap between the 10th and 90th percentile in vocabulary size is roughly five to six times [1]. What gets described as "you should have X words by Y months" is nothing more than the median of that distribution with its width removed and its center converted into an alarm.
Japanese national data tell the same story. According to the 2023 Survey on Infant and Child Physical Development published by the Cabinet Office's Children and Families Agency, the proportion of infants whose caregivers report them "saying single words" passes 90% at around 16 to 17 months — meaning that at the one-year mark, roughly 70% of children are not yet producing single words [2]. At the first birthday, a non-talking child is not an outlier. They are, by the numbers, in the majority.
That said, "a wide normal range" does not mean "nothing to attend to." Rescorla's well-known longitudinal study of children classified as late talkers — expressive vocabulary in the bottom 10 to 15% at age two — found that while a substantial proportion caught up to peers by school age, the group as a whole showed measurably lower language and literacy scores at age 17 compared with children who had not been late talkers [3]. Being somewhere within a wide normal range and keeping the option of professional evaluation available for children at the far edge of that range are not contradictory positions. They can and should coexist.
Expression and comprehension are different axes
Language development has two dimensions that move independently: expression (what the child produces) and comprehension (what the child understands). Watching only expression means missing half the picture.
The CDI normative data show that receptive vocabulary consistently leads expressive vocabulary, with the gap widest between 8 and 16 months [1]. At 13 months, the typical child might say around 10 words but understand around 50 [1]. That gap is the normal state of affairs, not a warning sign.
You can observe comprehension without waiting for words. A child who takes the wrapper to the bin when asked "can you put that in the trash?" is processing language at the sentence level. A child whose eyes move to the correct object when you say "where's the dog?" has word-to-referent mapping in place. A child who responds to the words on a picture-book page is accumulating vocabulary even if none of it has surfaced as speech yet. None of these require the child to produce a single sound.
Children who are slow to express often experience what the literature calls a vocabulary spurt — an accelerating accumulation of words over a short window. In a longitudinal study of 18 infants by Goldfield and Reznick, 13 of the 18 showed a period of rapid acceleration, and most spurts consisted primarily of a surge in nouns [4]. Using expression alone as the yardstick for language development means reading from a dial that only shows part of the gauge.
The social signals that arrive before words
Before language becomes a concern, there are earlier indicators worth watching: does the child make eye contact, respond to a parent's facial expression, follow a gaze, and point?
Pointing in particular has a robust body of evidence behind it as a predictor of language development. A meta-analysis by Colonnesi and colleagues, drawing on 25 studies with roughly 734 children, found a concurrent correlation of r = 0.52 and a longitudinal correlation of r = 0.35 between pointing and later language development [5]. Declarative pointing: the "look at that" gesture used to share interest with another person, as opposed to imperative pointing which demands an object — was more strongly associated with language outcomes than imperative pointing [5]. Experimental work by Liszkowski and Tomasello demonstrated that by 12 months, infants use pointing in ways that already reflect awareness of what the other person knows and does not know, making it a genuine act of shared attention rather than a simple reflexive reach [6].
Pointing is also checked at the 18-month checkup (the 18-kagetsu kenshin, a universal pediatric visit in Japan). The M-CHAT: Modified Checklist for Autism in Toddlers — a caregiver-completed questionnaire used to screen for autism spectrum disorder in children 16–30 months old, used widely as a screening instrument for children 16 to 30 months, evaluates around 20 caregiver-reported items including pointing, gaze following, and interest in other people [7]. Looking for "does this child point?" before "does this child talk?" is, for certain decisions, the more informative sequence. Individual variation is large, however, and no single indicator can be read in isolation. The 18-month checkup is a natural and appropriate place to bring any concerns.
When the 18-month checkup flags something
If the 18-month checkup identifies a language delay, the options presented typically follow a progression.
First, watchful waiting: monitoring until the next scheduled checkup (age two or three in most municipalities), or a reassessment in a few months. Watchful waiting is not the same as doing nothing — it is a committed schedule of observation.
Next, municipal programs: parent-and-child groups, community drop-in centers. Less intensive than formal therapy, but they offer ongoing contact with public health nurses and developmental psychologists.
Further along, early intervention services (ryoiku): available in many municipalities regardless of formal diagnosis.
Finally, specialist evaluation: developmental pediatrics, pediatric neurology, speech-language pathology.
Which step to take depends on the child's overall picture and the family's circumstances. Hesitation about early intervention is common among parents, but Rescorla's follow-up data make a clear case for keeping the option open: if a portion of late-talkers show lingering language gaps at age 17, then having access to early evaluation and support is a reasonable protective measure [3]. That is not the same as urgency. If there is not yet enough information to decide, getting a second professional opinion before committing to a path is not too late.
When in doubt, the pediatrician or the local public health center (hokensentā) is the most efficient first stop.
Not rushing — but keeping the observation going
This article will not recommend exercises to "make your child talk faster." There are two reasons for that. One is that the evidence for most such interventions is thin. The other is that parental anxiety about a child's speech rate can itself degrade the quality of everyday interaction — which is, in the end, the environment in which language develops.
What can be done instead is to record what the child produces, without judging whether it counts as a real word. "Sounds like 'nmah.'" "'Bub' is turning into 'boo-boo' (car)." "'Nigh-nigh' showed up just before sleep." Write down the sound and the situation. No verdict required.
The timing of this matters more than it may seem. A longitudinal study by Majnemer and Rosenblatt demonstrated that parental recall of developmental milestones — when asked retrospectively — is compressed toward later dates by several months [8]. The difference between remembering when the first word appeared and having written it down the day it happened is, in terms of data quality, substantial. When a public health nurse at a checkup asks about the progression of language development, a running log is primary source material, not a reconstructed estimate.
The log is also more than a clinical document. "First word in April, but before that, these sounds, in these situations, and these changes" is a narrative of development. The single-point "first word in April" is not. A record like this — kept in Memori, in a paper notebook, in a phone's notes app, in any form at all — becomes a record of a child's linguistic history, not just a milestone date.
The stance this article argues for is: not rushing, while keeping the observation rigorous. Those two things are not in tension.
Summary
The window for first words is wide [1,2]. Comprehension and social signals precede expression, and pointing in particular is robustly associated with later language development [5,6]. Some late talkers show residual language differences into adolescence — but this is an argument for maintaining access to evaluation and support, not an argument for rushing [3].
There is nothing wrong with wanting your child to talk. The suggestion here is simply to direct that wanting inward first: observe carefully, record what you see, and let the observation accumulate before making it into pressure. That is a quiet form of the ethics of not rushing.
When in doubt, consult the pediatrician or the public health center. And keep the log — it holds the record of the uncertainty too.
References
- Fenson L, Dale PS, Reznick JS, Bates E, Thal DJ, Pethick SJ. Variability in early communicative development. Monogr Soc Res Child Dev. 1994;59(5):1–173; discussion 174–185. PMID: 7845413.
- Children and Families Agency, Cabinet Office, Japan. Summary Report of the 2023 Survey on Infant and Child Physical Development. 2024. https://www.cfa.go.jp/policies/boshihoken/r5-nyuuyoujityousa
- Rescorla L. Age 17 language and reading outcomes in late-talking toddlers: support for a dimensional perspective on language delay. J Speech Lang Hear Res. 2009;52(1):16–30. doi:10.1044/1092-4388(2008/07-0171). PMID: 18723598.
- Goldfield BA, Reznick JS. Early lexical acquisition: rate, content, and the vocabulary spurt. J Child Lang. 1990;17(1):171–183. doi:10.1017/S0305000900013167. PMID: 2312640.
- Colonnesi C, Stams GJJM, Koster I, Noom MJ. The relation between pointing and language development: A meta-analysis. Dev Rev. 2010;30(4):352–366. doi:10.1016/j.dr.2010.10.001.
- Liszkowski U, Carpenter M, Tomasello M. Pointing out new news, old news, and absent referents at 12 months of age. Dev Sci. 2007;10(2):F1–F7. doi:10.1111/j.1467-7687.2007.00552.x. PMID: 17286836.
- Robins DL, Fein D, Barton ML, Green JA. The Modified Checklist for Autism in Toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders. J Autism Dev Disord. 2001;31(2):131–144. doi:10.1023/A:1010738829569. PMID: 11450812.
- Majnemer A, Rosenblatt B. Reliability of parental recall of developmental milestones. Pediatr Neurol. 1994;10(4):304–308. doi:10.1016/0887-8994(94)90126-0. PMID: 8068156.