Children acquire language at a rate that would be impossible to achieve by deliberate instruction, and the mechanisms are partly understood.
The trajectory
First words typically appear around the end of the first year, with enormous individual variation.
Vocabulary then grows slowly and accelerates, with many children showing a period of rapid acquisition in the second year.
Two-word combinations follow, then grammatical complexity develops rapidly over the following years.
Which is a general pattern, and the timing varies widely among children who develop typically.
What supports it
Amount of language directed to the child, which has been studied extensively.
The frequently cited findings about differences in words heard by socioeconomic group have been criticised methodologically and reanalysed, with the differences appearing smaller than originally reported.
What has held up better is that conversational turns — back and forth exchanges — predict outcomes more strongly than word counts.
Which suggests interaction rather than exposure is the mechanism, and it is a more actionable finding.
Child-directed speech
The distinctive register adults use with infants — higher pitch, exaggerated intonation, simplified structure, repetition.
Which appears across cultures in varying forms and has been shown to attract and hold infant attention.
Studies find infants learn words better from it than from adult-directed speech.
Joint attention
Both adult and child attending to the same thing.
Which appears to be the context in which word learning occurs most efficiently, since it resolves the problem of knowing what a word refers to.
Following the child's attention rather than directing it is associated with better outcomes, which is a small practical point with reasonable support.
Reading
Books contain vocabulary and structures that conversation does not.
Which is why shared reading is associated with vocabulary development beyond what talking alone provides.
Interactive reading — asking questions, relating to the child's experience, letting them lead — has better evidence than reading the text alone.
Bilingualism
Children exposed to two languages acquire both, with vocabulary in each initially smaller and total vocabulary comparable.
Which means apparent delay in one language frequently disappears when both are counted.
Bilingualism does not cause language delay, which is a persistent belief that research does not support.
Advice to families to abandon a home language is not supported and can damage family communication.
When to seek advice
Health services publish milestone guidance, and the specific indicators warranting assessment generally include no babbling by a certain age, no words by around eighteen months, no two-word combinations by around two years, loss of previously acquired words, and not responding to sound.
Which are prompts to seek assessment rather than diagnoses.
Early speech and language intervention is effective, and waiting to see is generally worse than asking.
Any concern should go to a health visitor, doctor or speech and language therapist, and referral routes exist in most systems including self-referral in some.
Gesture
Pointing and other gestures precede and predict spoken language.
Which means a child who gestures to communicate is developing communication even without words, and gesture use at a given age predicts later vocabulary.
Responding to gestures with words is one of the mechanisms by which the association may operate.
Screens and language
Background television reduces the quantity and quality of parent-child talk in observational studies.
Which is one of the clearer findings about screen effects on young children, and it concerns adult device use as much as the child's.
Video calls with responsive interaction appear to differ from passive viewing.
Dummies and bottles
Prolonged use has been associated with speech development in some studies, with the evidence being mixed.
Which is why guidance generally suggests limiting use during the day once talking begins, and a health visitor can advise on the individual situation.
Late talkers
A proportion of children with delayed early vocabulary catch up without intervention.
Which has led to debate about whether to intervene or monitor, and the practical position in most guidance is that assessment is appropriate since identifying who will not catch up is not possible in advance.
Intervention is generally low-risk and effective, and delay in accessing it is the greater risk.
Hearing
Undetected hearing loss, including fluctuating loss from middle ear fluid, is a common contributor to language delay.
Which is why hearing assessment is generally the first step when language is delayed.
Newborn screening detects permanent loss and does not detect subsequent or fluctuating loss.
What to do daily
Narrating what you are doing, describing what the child is attending to, and waiting for responses.
Which builds the back-and-forth exchange that the research identifies as the mechanism.
Nursery rhymes and songs
Repetitive, rhythmic and predictable, which supports phonological awareness and memory.
Which is why they persist across cultures, and singing to a child requires no ability to sing.
Which is why they appear in every early years programme.
The repetition that adults find tedious is precisely what makes them work.