Skip to content

Support Early Language Development

Principle · All principles
Evidence · 26 claims (17 for, 6 mixed, 3 against) · 28 studies (10 quant-synthesis, 6 causal, 6 review, 4 associational, 1 qualitative, 1 theoretical), q1–q4 · 8 of 28 report an effect size · 20 claims rest on one study

Conditional relationship

For a young child (birth to about five) who is building first-language vocabulary, grammar and conversational skill, frequent exchanges in which an adult responds to what the child says, points at or attempts, lets the child take a turn, and expands or recasts the child's words are expected to produce more growth in the child's own talk (expressive vocabulary, number and variety of utterances) than the same amount of language the child only hears: adult talk directed at the child with little room for the child to reply, overheard conversation, or background media. The outcome is the child's oral language measured shortly after a period of weeks of such interaction (immediate-performance on vocabulary and language measures); later reading and school achievement are a separate, further claim.

What tests this, and how far. No claim in the wiki compares responsive, turn-taking interaction with passive exposure in general. The general statement rests on associational evidence: conversational turns predict language development and brain measures better than the number of words heard or family socioeconomic status [+M], reported second-hand and with no sample or effect size. One routine has been tested experimentally: dialogic reading, in which the adult prompts the child to talk about the book, praises and expands the child's answers and gradually hands more of the telling to the child. Against ordinary shared reading of the same books, it raises expressive vocabulary and oral language [+S] (two meta-analyses). So the relationship is tested for one setting (shared book reading with children of about two to five) and carried from there to everyday talk, play and care routines as an untested proposal.

The relationship is conditional on the child's age and starting language (the dialogic-reading gain is substantially smaller for four- to five-year-olds and for children at risk of language impairment), on who delivers it (gains obtained by researchers did not replicate when children's own teachers did the reading in ordinary classrooms), on how often it actually happens (centres that ran dialogic reading less often got smaller oral-language effects), and on the outcome (no effect on phonological processing). It does not say that the amount of talk is irrelevant, nor that homes whose language practices differ from a researcher's expectations are deficient: the claim page records that the original "30 million word gap" estimate has not fully replicated and that part of the gap reflected that mismatch.

How the converted siblings sit beside this: Scaffolding and Fading owns the general relationship of contingent support raised after a failed attempt and lowered after a success; expanding and recasting a child's words is that move applied to talk, and this page does not repeat its model. Explicit Vocabulary Instruction owns the teaching of a selected set of words, which becomes a larger share of the design as children reach four and five. Phonics owns code-focused teaching, which this relationship does not reach. The theories behind it are Language Acquisition Theory (the social-interactionist account) and Sociocultural Theory (support in the zone of proximal development).

Default design, while the relationship is untested

The earlier page's guidance, kept as a concrete default a designer, parent educator or early-years lead can act on and revise. Steps 4 to 6 rest on the tested dialogic-reading routine; the rest are untested proposals, labelled.

  1. Follow the child's lead and answer it. Notice what the child is looking at, pointing to, babbling or saying, and respond to that, in words, within a few seconds; then pause long enough (a slow count of five is a working proposal) for the child to take another turn. With infants, gestures, pointing and sounds count as turns. Untested as an intervention; the reason for it is the associational turn-count finding (conversational turns [+M]) and the child-directed speech element.
  2. Talk with, not at, the child. Count exchanges, not words: a stream of adult talk the child cannot reply to, overheard adult conversation, or a screen playing in the background is not a substitute. Untested proposal; no claim here tests passive exposure against interaction.
  3. Pitch language just above the child's, and expand it. Use short sentences near the child's own length plus one step (child: "dog"; adult: "a big dog, running"). Recast errors rather than correcting them ("I goed" → "You went to the park?"). Do not hold language simple for longer than the child needs (the earlier page warns that sustained over-simplification limits exposure to what the child is ready for). Untested as a separate rule; expansion is one component of the dialogic-reading routine in step 4.
  4. Run shared reading as a conversation (dialogic reading). On each page or two, prompt the child (complete a sentence, recall what happened, an open-ended "what's happening here?", a wh- question, or a question linking the book to the child's own life), praise and expand the answer, and ask the child to say the expansion. Over rereadings of the same book, hand more of the telling to the child. Basis: Dialogic reading improves language outcomes [+S] (meta-analyses; moderate effect on expressive vocabulary against typical shared reading).
  5. Dose: often and short, for weeks. A working default is one dialogic session of about 10 minutes a day, at least five days a week, for 6 to 8 weeks, with the same book read two or three times across a week. These numbers are untested proposals; what the evidence says is only that frequency matters: centres with lower compliance got smaller oral-language effects [~M].
  6. Keep the group small enough for every child to talk. One adult with one child, or with up to four or five children, so that each child gets several turns per book; a whole-class read-aloud gives most children few. Group size is an untested proposal (the hub claim's Discussion says whole-class read-alouds dilute turn-taking); the meta-analyses pooled one-to-one and small-group delivery.
  7. Train the adults by doing, with feedback. Show the routine (live or on video), have each adult practise it with their own child or group while a coach watches, give feedback on two or three things (did the child talk on most pages; was the child's answer expanded), and check again after a week or two. Basis, carried from parent training on child behaviour: coaching with practice outperforms information alone [+M]; and for language specifically, the teacher-delivered replication failure inside the dialogic reading hub [~S] says the routine is lost when adults are only told about it.
  8. Widen the partners. Beyond the main caregiver, give the child regular conversation with other adults and with peers (play, meals, routines), so the child meets varied language and partners. Untested; from the earlier page.
  9. Progression across the early years (untested arc). Under about 18 months: responsive naming and turn-taking in care routines and play, books as objects to point at and talk about. About 2 to 3 years: daily dialogic reading with completion and recall prompts, plus talk during play. About 4 to 5 years: open-ended and "your own life" prompts, longer stories and retelling by the child, with explicit teaching of selected words (Explicit Vocabulary Instruction) because the dialogic gain is smaller at this age. Within any 6- to 8-week block, move a child to harder prompts when the child answers most prompts on a familiar book unaided; return to completion prompts and shorter books when the child mostly stays silent.

When a condition is not met: if no adult can give children one-to-one or small-group time, schedule short rotating small groups and drop whole-class reading as the main language activity; if the adults cannot be observed and coached, give them one routine (one prompt type and expansion) rather than the whole sequence, and check its frequency.

Fitting the design to a situation

The three facts that most change the decision:

  • The child's age and current language. An infant who points but does not yet speak, a two-year-old with a few hundred words, a four-year-old who tells stories, and a child with a diagnosed language delay need different prompts, group sizes and adult support, and the tested effect shrinks at four to five and for children at risk. If unstated, ask: how old are the children, what do they say now (single words, short phrases, stories), and has anyone flagged a delay or hearing loss?
  • Who will carry out the interaction, and whether they can be coached. Parents at home, an early-years teacher with a full room, a home visitor, or volunteers each need different training and a different amount of the routine. If unstated, ask: who spends time talking with these children, how many children per adult, and can anyone watch them do it once and give feedback?
  • Which language outcome counts. Expressive vocabulary and the child's own talk are where the tested gain lies; receptive vocabulary, phonological awareness, letter knowledge and later reading need other or additional teaching. If unstated, ask: what should the children be able to say or do differently, how will anyone know, and when?
If the brief says… Then change… Basis
Infants, or children not yet using words Respond to gestures, sounds and pointing as turns; name what the child attends to; use books as things to point at and talk about rather than to read through; keep exchanges to a few turns at a time across care routines untested proposal; the associational turn-count finding (conversational turns [+M]) and the child-directed speech element. The dialogic-reading trials did not include infants
Typically developing children of about 2 to 3 Make daily dialogic reading the core routine (completion and recall prompts, expansion, repeated books), one-to-one or in small groups, and coach the parents or staff who do it Dialogic reading improves expressive language [+S] (moderate effect against typical shared reading; larger for younger children)
Children of about 4 to 5 Shift to open-ended and "your own life" prompts and retelling by the child; add short, explicit teaching of a few selected words from each book; do not expect the same gain from reading alone Dialogic reading improves language outcomes [~S] (effect substantially smaller at 4–5); Explicit Vocabulary Instruction
Children with language delays or disabilities Use one-to-one or very small groups, more sessions, simpler prompts, and work with a speech-language professional; do not rely on a parent-led reading routine alone, and judge progress on the child's own talk (utterances, variety of words) Potentially positive effects for children with disabilities [+M] (two trials, 52 children with language delays); Dale et al. 1996 [+M] (gains on 2 of 6 measures); parent-led dialogic reading showed no significant differences [-M] (19 children); the hub's smaller effect for children at risk [~S]
Children who are deaf or hard of hearing Give early, full access to a language the child can perceive (including a sign language), with the same responsive turn-taking in that language; involve specialists early untested here; from the earlier page (early sign-language exposure is time-sensitive). No claim in the wiki was read for it
The home language differs from the setting's language, or families speak several Ask caregivers to talk and read in the language they speak most fully; do not advise dropping the home language; in the setting, support both where staff can untested here; Preserve Heritage Language While Acquiring an Additional Language; the turn-count claim's note that some of the original "word gap" reflected different, not deficient, home practices (conversational turns [~M])
Delivered through parents or caregivers at home Coach rather than inform: show the routine, have each parent practise it with their own child while someone gives feedback, follow up once or twice; expect the families most at risk to need the most support to change routines Parent coaching outperforms information only [+M] (carried from parenting and child behaviour outcomes, not language); parent-implemented interventions improve outcomes [+M] (carried from children with autism); the hub's finding that families at greatest risk did not change home reading [~S]
Phone-only contact with families, or no in-person sessions Send one short, concrete prompt a week (one thing to say or ask during a routine the family already has, such as bath time or a bus ride), and, where possible, one short video of the routine; do not rely on texts alone for a child with a delay READY4K! text messages [+M] (over 1,000 preschool families; small effect on an early-literacy screen, not on oral language). Remote coaching of the dialogic routine is untested
classroom with one teacher for 15–20 children Rotate small groups of 4–5 for 10-minute dialogic sessions while others play; train assistants or volunteers to run some groups; check once a week that sessions happen and children talk on most pages the teacher-delivered replication failure in the hub [~S]; frequency of implementation moderates effects [~M]; group sizes are untested proposals
Scarce facilitator time (volunteers, a busy shift, a home visitor with an hour a month) Teach one move, not the sequence: one prompt per page plus expanding the child's answer, on a pocket card; spend coaching time on watching one session and giving one piece of feedback; protect frequency over length untested proposal; frequency of implementation moderates effects [~M]
The brief names its own routine (a curriculum's story time, a home-visiting script, its own observation checklist) Keep it; change only what the evidence says matters: the child talks on most pages or turns, the adult expands what the child said, and it happens most days. Use the brief's own checklist and categories to record this; add no new test untested proposal; the hub claim's three ingredients (dialogic reading [+S])
Single session only (one parent workshop, one library story time) Spend it on one modelled reading, each adult practising once with their own child with feedback, and a take-home card with two prompts; do not claim a change in children's language from it untested proposal; coaching evidence is from multi-session programmes (parent coaching [~M])
Caregivers with low literacy or low confidence reading aloud Use wordless picture books, photo books or the family's own stories; the prompts and expansion do not need the adult to read the text; praise what the caregiver already does untested proposal; the dialogic routine is built from prompts and responses, not decoding
Adults in work settings (childcare staff, health visitors) Train through practice with the children they work with and feedback, not lectures; give time in the rota for small-group reading experiential training of early childhood workers [+W] (an argument, untested); parent coaching [+M] carried
Goal includes phonological awareness or letter knowledge Add explicit code-focused activities; dialogic reading alone will not produce them no discernible effects on phonological processing [-M]; Phonics
Recording home talk (audio recorders, turn counters) or other private material The family decides whether to record and who hears it; use any count only as feedback to the family on turns, never as a judgement of the family or a ranking of homes untested proposal; the turn-count claim's caution about word-count framing (conversational turns [~M])
A gap-closing target (a programme expected to close a vocabulary gap) State a smaller, measurable goal (the children's own talk and the words taught), and do not promise to close a gap with a vocabulary programme alone Vocabulary-focused programmes are probably insufficient to close the vocabulary gap [~M] (school-age linguistic comprehension programmes, carried to early years)

Seventeen rows; fifteen link a claim, but only the 2–3, 4–5, delays, classroom, phonological and named-routine rows rest on dialogic-reading claims tested with young children's language. The parent, phone, single-session and work-setting rows carry evidence from parent training on child behaviour, autism interventions, an early-literacy text programme or an untested argument, and say so; the infant, home-language, recording, scarce-time and gap-closing rows are untested proposals that cite a claim only for its direction or its limit; the deaf and low-literacy rows are untested.

Observation, state and explanation

What can be observed is an exchange under stated conditions: who started it, how many turns each side took, what the child said or did (words, gestures, sounds), how long the adult waited, what the adult did with the child's turn (ignored it, repeated it, expanded it, corrected it, asked another question), and the setting (book, play, meal, one child or a group). "Has a delay", "is shy", "the home is language-poor" and "the adult is not responsive" are inferred states. One observation of a child in an unfamiliar setting, with an unfamiliar adult, in a language that may not be the child's strongest, shows little about what the child can do at home.

Response Plausible explanations Observation that could change the interpretation
Child stays silent when prompted during reading Prompts pitched too high; unfamiliar adult or group; the book's language is not the child's strongest; a hearing or language difficulty; the child is used to listening during reading Try completion prompts on a familiar book, one-to-one, with a familiar adult and in the home language; ask caregivers how much the child talks at home; arrange a hearing check if silence is general.
Child answers but in one word every time Prompts invite one word (naming questions); the child's current level; adult does not expand or ask for the expansion Change to open-ended and recall prompts for a week and count words per answer; check whether the adult repeats back an expanded version.
Adults report doing the routine, but children's talk does not change Sessions happen less often than reported; adults read with questions bolted on but do not wait or expand; the outcome measure does not capture the child's talk Watch or record one session (with consent) and count child turns and expansions; compare with the logged frequency.
A child talks freely at home but rarely in the setting Unfamiliar language or partners; group too large to get a turn; the setting's talk does not connect with home topics Observe in a small group with a familiar peer, use topics from home, and ask the family what the child talks about.
Gains on words from the books, none on a general language test The gain is specific to taught or read words; too little time for general growth; the test measures receptive language while the routine trains production Test with items from the books and with an expressive measure; extend the block and retest.

These rows are proposals, untested as diagnostics. A probe is itself an interaction and changes the child's response, so record each one, and keep open the explanations that remain rather than choosing the one that fits a label.

Evidence and qualifications

Learners, outcomes and comparisons differ across these claims (expressive vocabulary against typical reading, language measures against story time, child behaviour after parent training), so do not rank them by effect label. None measures whether gains last after the intervention ends, or whether they reach later reading.

The evidence by question

Further evidence, not yet read against this model

Claims bearing on the page that were read but not used as core evidence above.

Objective and learner-valued goal

The designer's objective is usually growth in the child's own spoken language (more words used, longer and more varied utterances, taking part in conversation) over a period of weeks. A young child does not state a learning goal; what the child values shows in what they choose to talk about, point at and ask for again: a favourite book, a pet, a game. The caregiver's goals are separate and also matter: to enjoy time with the child, to get through bedtime, to prepare the child for school, to keep the family's language. Agreement and divergence are both observations. A parent who values a calm bedtime story may resist a routine that turns every page into questions; a family that values its home language may stop taking part if advised to switch to the setting's language.

For example, a home visitor's objective may be more expressive vocabulary at the two-year check, while the parent wants the child to sit still for a story. A routine that starts from the child's favourite book, asks one question per page and stops when the child loses interest serves both; record the child's talk and the parent's view of how the routine went, so that the objective is not read as the family's goal.

What would revise this model?

The expectation should weaken if trials comparing responsive, turn-taking interaction (outside book reading) with the same amount of adult talk directed at the child found no difference in the child's own language, or if dialogic-reading gains disappeared at a delay of months. If gains depend on the routine being delivered by researchers or trained specialists, and not by parents or ordinary staff even with coaching, the design should move effort to specialist small-group delivery for the children who need it most. If later work shows the turn-count association is explained by other differences between homes, the general statement on this page rests on dialogic reading alone and should be narrowed to it. Do not protect the model by calling every null result "low fidelity" after the fact; record frequency and what adults did before the outcome is known.

Child talk during sessions, vocabulary on taught words, general expressive language, receptive language, phonological skills and later reading are separate outcomes. The present evidence sets no dose, no group size and no age at which to change routines.

Claims

Examples

Key Sources

  • Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young American children. Paul H. Brookes.
  • Clark, E. V. (2009). First language acquisition (2nd ed.). Cambridge University Press.
  • Werker, J. F., Pegg, J. E., & McLeod, A. J. (1994). Cross-language speech perception: Initial capabilities and developmental change. Developmental Psychology, 30, 702-721.
  • Arduini-Van Hoose, N. (2020). Language development. In Educational psychology. Retrieved from https://edpsych.pressbooks.sunycreate.cloud. CC BY-NC-SA 4.0.
  • Biscevic, I., Pasalic, A., & Memisevic, H. (2018). The effects of executive functions and theory of mind on semantic fluency in preschool children. Problems of Education in the 21st Century, 76(1). https://www.scientiasocialis.lt/pec/. doi:10.33225/pec/18.76.21
Page metadata
Field Value
type principle
id supporting-early-language-development
aliases preschool-curricula-ef-tom-training
title Support Early Language Development
description For young children building a first language, frequent responsive exchanges in which an adult answers and expands the child's own turns are expected to grow the child's expressive language more than language the child only hears; one routine, dialogic shared reading, is tested against ordinary reading, and the general relationship rests on associational evidence.
status review
canonical True
generated by claude/unspecified on 2026-10-15
trust unverified — no one has confirmed this page's content
sources 1 — biscevic-2018

Frontmatter is the page's machine-readable record — see the Schema & Guide.