Last updated 2026-07-11
TL;DR
Teaching a child to initiate communication comes down to two things: build the desire to communicate, then hand them a reliable, low-pressure way to act on it. Hanen, PECS, and naturalistic developmental behavioral interventions all lean on the same move: set up moments where the child needs something and has a tool ready to ask for it. Then wait, respond fast, and repeat it daily.
Why some kids wait to be spoken to instead of starting things themselves
Most children start communicating long before their first word. They reach, point, and babble to pull an adult's attention where they want it. By 12 months, typical development includes pointing to share interest and pulling a caregiver toward something wanted [1]. When that doesn't happen, or stalls, it's rarely stubbornness. Usually the child hasn't yet connected the idea that communication is a tool they can pick up and use.
For late talkers, the gap is often plain developmental lag. For autistic kids, the barrier is frequently two things at once: reduced joint attention (the impulse to share an experience with another person) and motor planning challenges that make it physically hard to start a word or gesture [2]. Some kids have the words inside and just haven't learned that words get things done. Others have learned that waiting works fine, because adults read every need before it's expressed.
There's also a learned passivity that builds up quietly. When adults constantly prompt, fill in words, or hand over what a child wants before any attempt to ask, the child learns that starting is optional. Well-meaning adults sometimes teach kids not to bother.
What counts as "initiation" for a young or nonverbal child
Initiation doesn't require a spoken sentence. For a young or minimally verbal child, it's any self-generated communicative act: reaching toward a caregiver, handing over an object, pointing, using a picture card, pressing a button on a speech-generating device, or catching your eye and vocalizing at the same time [3].
Speech-language pathologists sort initiation into a handful of functional categories:
| Type of initiation | Example | Why it matters |
|---|---|---|
| Requesting | Child points at juice or says "more" | Most common first target; high motivation |
| Commenting | Child holds up a bug to show you | Builds joint attention and social bond |
| Protesting | Child pushes away unwanted food | Reduces frustration and behavior |
| Greeting | Child waves or says hi unprompted | Social entry skill |
| Calling | Child says your name to get attention | Essential for safety |
The American Speech-Language-Hearing Association describes communicative intent as the child's ability to deliberately send a message, and notes that intentional communication typically emerges between 8 and 12 months in typical development [3]. When it's delayed or absent past 12 to 18 months, that's one of the clearest early flags for a speech-language evaluation. If you're new to this process, the piece on early intervention walks through how evaluations and services actually work.
When it's worth raising with someone
The American Academy of Pediatrics recommends developmental surveillance at every well-child visit and formal developmental screening at 9, 18, and 30 months, with autism-specific screening at 18 and 24 months [4]. Their red flags: no back-and-forth sharing of sounds, smiles, or facial expressions by 9 months, no pointing or waving by 12 months, no words by 16 months, and no two-word spontaneous phrases by 24 months.
Not initiating is different from not talking. A child can have words and only ever use them in response, never to start an exchange, and that still deserves attention. So does a child who used to initiate and then stopped: regression warrants prompt evaluation regardless of age [4].
Nobody diagnoses off a checklist, and this article isn't trying to. If you're worried, the right move is a referral to a certified speech-language pathologist. The guide on finding a speech therapist covers what to look for and what an evaluation actually involves.
What actually works at home
The research here is clearer than on many early childhood questions. Naturalistic developmental behavioral interventions, which include JASPER, ESDM, and PRT, show the strongest evidence for building communication initiation in autistic children [5]. They share a structure: follow the child's lead, find high-motivation moments, then systematically prompt and reward attempts to start.
A few strategies carry the most practical backing. One is what the research literature calls "communication temptations" [6]: sabotage the routine a little. Put the crackers in a jar the child can't open, set up a favorite toy and don't turn it on, hand over a cup with no milk in it. Communication becomes the easiest solution. Then wait, at least five full seconds, before offering help. That silence feels awful to adults. It works anyway.
Getting things out of arm's reach helps too. If everything the child wants is already within grabbing distance, there's no need to ask. Put preferred items on a high shelf, in a clear bin they can see but not open, or in a bag you're holding, so access depends on communication rather than proximity.
Time delay is worth using on purpose, not just as an accident of a busy moment. Position yourself near the desired item, make eye contact, and wait, no prompt, no model, just expectant waiting. This has good evidence for increasing spontaneous requests in children with autism [6]. Start with a three-second delay and stretch it out gradually.
Whatever the child offers, a look, a reach, a sound, honor it fast. Hand over the item, name it, respond warmly, aiming for something like a three-second response window. Slow or missing responses teach a child that starting doesn't pay off. And if your child uses an AAC device or a low-tech picture board, model on it yourself throughout the day, not as a prompt directed at the child but as ordinary use. Children learn AAC initiation partly by watching adults use the system first [7].
The Hanen approach
The Hanen Centre's "It Takes Two to Talk" program trains parents, more than therapists, to support early communication. It has one of the better evidence bases for improving communication in late talkers and young autistic children, including a randomized controlled trial in the Journal of Speech, Language, and Hearing Research showing gains in parent responsiveness and child communication [8].
The program runs on a concept called OWL: observe, wait, listen. Instead of leading, the parent watches what the child is interested in, waits for the child to do something, then listens and responds. It sounds simple and is hard to do consistently, since it means resisting the instinct to direct, prompt, and fill every silence.
Hanen also teaches "Four Steps to Encouraging Initiation," moving from setting up an opportunity, to waiting with expectant body language, to modeling a communication attempt, to responding when the child tries. A certified Hanen program is usually delivered by an SLP, though the books and some group programs are available to families directly.
Where PECS and picture systems fit in
The Picture Exchange Communication System (PECS) was built specifically to teach initiation, which is worth saying plainly since many picture systems actually teach labeling or responding instead. PECS Phase I starts with the child picking up a picture card and handing it to a communication partner to request something. The child moves toward the adult, not the reverse, and that physical act of exchange is the initiation [9].
The research on PECS is substantial. A Cochrane review found improvements in initiating requests and some evidence of increased speech, though it rated the overall quality of evidence as low-to-moderate [9]. What's not in dispute is that PECS Phase I, done correctly, produces spontaneous requesting fairly quickly in many children.
One nuance matters: PECS isn't meant to be permanent for most kids. It's a bridge. Many SLPs use it to establish the concept of intentional initiation, then move the child toward a speech-generating device or toward speech as abilities grow. For kids with motor planning difficulties that affect pointing or handling cards, apraxia of speech can complicate picture-based initiation, and an SLP who knows the child should adjust the motor demands of the system accordingly.
Autistic kids and late talkers need different approaches to initiation, and the environment and the therapist's technique both matter more than most parents expect. Here's how it breaks down.Autism versus late talking: why the approach differs
Late talkers without another diagnosis usually already want to communicate. They just don't have the words yet. Build vocabulary, give them something worth saying, and initiation tends to climb on its own, because the joint attention and social drive were there all along.
Autistic children are a more mixed group. Some have strong social motivation and just need a motor pathway: a device, a card system, a reliable gesture, something to start with. Others have reduced joint attention, meaning the underlying impulse to share an experience or get someone's attention hasn't developed yet and needs to be taught directly alongside any communication system [2].
JASPER (Joint Attention, Symbolic Play, Engagement, and Regulation) was built for this second group. It's a therapist-delivered intervention that builds joint attention through play before layering on communication, and a randomized trial in JAMA Pediatrics found it significantly increased initiating joint attention in minimally verbal autistic children aged 5 to 8 [10].
If your child doesn't look up to show you something interesting, or points to request but never to share, that gap deserves as much attention as whatever communication system you're using. A device won't solve initiation if the child has no interest in sharing the experience with you in the first place. For more on approaches specific to this population, see autism spectrum speech therapy.
The environment either invites communication or shuts it down
Kids in highly predictable, highly accommodating homes have fewer reasons to speak up. When every need gets anticipated and every routine runs on autopilot, there's no gap left to fill. Creating that gap on purpose, sometimes called "engineering the environment," is one of the most recommended home strategies in both the Hanen literature and NDBI manuals [5].
A few things that actually work: keep preferred items visible but out of reach, so a clear bin on a shelf gives the child something to point at (locked away, there's nothing to point to; freely available, there's no reason to). Set up activities that need two people, like wind-up toys, bubbles, or a swing that needs pushing, so the payoff requires pulling you in. Cut screens during practice windows, since passive screen time asks nothing of the child communicatively. And turn snacks into practice by offering small portions piece by piece, which naturally creates repeated requesting.
One thing parents often miss: noise and chaos suppress attempts to start. A dysregulated or overwhelmed child, or one stuck in a loud room, initiates less. Calm has to come before communication, not after.
Can echolalia turn into real initiation?
Yes. Echolalia, repeating words or phrases heard elsewhere, isn't a wall blocking communication. For many children it's the bridge to it [11]. The echolalic phrase is simply the tool the child already has, and the work is helping them repurpose it for new jobs, including starting a conversation.
A child who says "do you want a snack?" while handing you a cracker has learned that the phrase connects to food, even though they're borrowing your words instead of saying "I want a snack." That's still initiation. It's still functional. Treat it as communication.
With steady modeling and quick responses to attempts, many children shift from echolalic toward more original language over time, though rarely in a straight line. Some kids mix echolalic and generative speech for years, and that's fine functionally. The echolalia meaning article is worth reading next if you want more on this.
What therapists actually do in sessions
A skilled SLP working on initiation spends most of a session creating opportunities and then waiting them out, not drilling vocabulary. With young kids, sessions often look more like structured play than teaching.
Mand training, borrowed from applied behavior analysis, starts with a preference assessment to find high-interest items, then teaches the child to request them through whatever modality works (speech, picture, device, sign). The request has to come from the child, not from a prompted imitation. Incidental teaching arranges the environment so the child approaches and starts the exchange first, with the SLP adding only the smallest prompt needed to shape a better response. And functional communication training steps in when frustration or behavior has been doing the communicating: teaching "help" or "break" as a way to initiate can cut behavioral escalations sharply.
A good SLP also coaches parents and caregivers, since initiation generalizes best when the adults around a child respond consistently. One or two sessions a week won't move much if nobody carries the approach into daily life [5]. If you're still looking for services, online speech therapy is a real option now for many families, especially for coaching-heavy work.
Why a child initiates with one person but not others
This is common, and it doesn't mean the skill isn't real. It just hasn't generalized yet, which is a separate teaching target from learning the skill in the first place.
Kids naturally gravitate toward whichever adult has the most responsive, predictable track record, which is a smart move from a learning standpoint. If Dad always hands over juice on request but Grandma needs three tries before she responds, the child learns to save the effort for Dad.
Building generalization means getting multiple caregivers to use the same response strategy so the child's experience stays consistent, starting with similar partners (say, two parents) before branching out to less familiar adults, and practicing in new settings on purpose, since a request skill learned at home won't necessarily show up at the playground without some deliberate exposure there too.
Little Words was built around exactly this problem. The app gives kids a consistent AI communication companion that responds predictably, and some families use it between therapy sessions to keep initiation practice going across different settings. The quiz at littlewords.ai/start takes about three minutes if you want to see whether it's a fit. And if a child's initiation is extremely rigid and partner-specific, it's worth having an SLP rule out childhood apraxia of speech, since motor planning difficulties can make communication feel safer with just one practiced partner.
How long progress takes
Anyone who gives you a precise timeline without knowing your child is guessing. That said, there's some data to anchor expectations: in PECS studies, many children reached spontaneous requesting at Phase I within a few weeks of consistent implementation [9]. JASPER trials measured joint attention changes over 10 to 24 sessions [10]. NDBI programs typically use a 12 to 20 week block to measure outcomes [5].
What the research does support is that consistency matters more than any other single variable. A technique used daily in natural routines beats the same technique used only twice a week in therapy. Parent training that reaches 80 to 90 percent implementation fidelity produces the best outcomes, and getting to that fidelity takes practice.
Progress is also lumpy rather than steady: a child might show nothing for three weeks, then generalize across three new communication partners in a single day. Often the learning has been happening quietly before it shows up in behavior.
If you've worked on initiation consistently for eight to twelve weeks with no movement at all, bring it up with an SLP. It might mean the target is too hard, the motivating items aren't preferred enough, or the prompting hierarchy needs adjusting.
Most kids start pointing, reaching, and looking at you while making sounds somewhere between 8 and 12 months, and by their first birthday they're usually pointing just to share something interesting, not only to ask for things. If a child isn't pointing or waving by 12 months, the AAP considers that worth mentioning to a pediatrician, and no initiation at all by 16 to 18 months means it's time for a speech-language evaluation. For a nonverbal child, the fastest research-backed approach is PECS Phase I, which can get a child spontaneously requesting within days to weeks when it's done consistently. Consistency is what makes or breaks it, and you have to start with items the child genuinely wants. A toy or snack they're lukewarm about won't motivate anything. Parents often can't tell if their child is truly initiating or just answering a prompt. Here's a simple test: put the desired item in view, stand near it looking neutral and expectant, and wait 10 to 15 seconds without saying a word. If your child does something communicative in that silence, that's real initiation. If they only respond after you ask "what do you want?", that's a prompted response, and it's a different skill entirely. Sign language, including key word signing like Makaton, works well as a bridge for kids with decent hand motor control. Evidence shows signs don't hold speech back and can actually support it by giving a child a way to communicate while their spoken language catches up. Teach signs for the things they want most, model them often, and respond right away when they attempt one. It's completely normal for a child to initiate only to request things at first. Requesting carries the highest built-in motivation because the payoff is immediate, so commenting and social initiation tend to show up later, once requesting is solid. If your child requests reliably, you've got something to build on. To grow commenting, set up small exciting moments (a bug crossing the path, a silly noise) and model a comment yourself without expecting one in return. If a child initiates at home but not at school, that's not a lost skill, it's a generalization gap. They've learned initiation works in one place but haven't yet found out it works everywhere. Get the SLP and teachers on the same page about prompting and available preferred items, and consider adding a generalization goal to the IEP or IFSP. Screen time that only asks a child to watch offers no chance to practice initiating anything. The AAP recommends skipping digital media besides video chat under 18 months and capping it at an hour of high-quality, co-viewed programming between ages 2 and 5. Turning screens off specifically during play, meals, and communication practice time costs you nothing and opens up real opportunities. "Communication temptation" is a technique where you engineer a situation that makes a child want to communicate: a favorite snack in a jar they can't open, a wind-up toy you let run down without restarting, an empty cup handed over with no drink in it. Then you wait, silently, and let the need do the work. It's one of the better-supported naturalistic strategies out there. Worried AAC will replace speech? It won't. This fear comes up constantly, and the research consistently shows AAC doesn't suppress speech and often helps it along, mostly by lowering frustration and giving the child a taste of successful communication. ASHA's stance is that AAC should be offered right away, not held back until a child has failed at speech first. Getting a child to initiate with people outside the family takes deliberate practice. Start with familiar secondary adults, a grandparent or family friend who knows how to respond, then slowly introduce less familiar people. A new person in a familiar setting is easier than a new setting altogether. Aim for the same behavior showing up across at least three different partners before you call it generalized. Joint attention, sharing focus on something with another person (spotting a bird, then looking back at you to share the moment), sets the stage for commenting and social initiation. Some children request just fine but rarely initiate purely to share. Programs like JASPER target joint attention directly and have solid evidence for boosting social initiation in autistic children. Teaching initiation looks different depending on age, though the underlying principles don't change. Toddlers respond best to play-based, naturalistic approaches built around sensory or food motivators. School-age kids are driven more by social motivation, and the targets shift toward peer interaction, asking for help, and protesting appropriately. A school-age child who never learned to initiate may also have picked up passive habits that take longer to unlearn. To track progress, the simplest method is counting unprompted communication acts (requests, comments, protests) during a fixed window, say a 10-minute play session, once a week. Look at the trend over four to six weeks rather than reacting to daily ups and downs. Many SLPs also use parent-report tools like the MacArthur-Bates Communicative Development Inventories alongside this.Sources
- CDC, Developmental Milestones and Act Early: By 12 months, typical development includes pointing to share interest and using gestures to communicate
- ASHA, Autism Spectrum Disorder Practice Portal: Reduced joint attention and social communication initiation are core features of autism spectrum disorder
- ASHA, Late Language Emergence Practice Portal: Communicative intent and intentional communication typically emerge between 8 and 12 months in typical development
- American Academy of Pediatrics, Developmental Surveillance and Screening: AAP recommends formal developmental screening at 9, 18, and 30 months and autism-specific screening at 18 and 24 months; red flags include no pointing or waving by 12 months
- Schreibman et al. (2015), Naturalistic Developmental Behavioral Interventions, Journal of Autism and Developmental Disorders: NDBIs including JASPER, ESDM, and PRT show strong evidence for building communication initiation in autistic children; consistency and parent implementation fidelity predict outcomes
- ASHA, Augmentative and Alternative Communication Practice Portal: Communication temptations and time delay increase spontaneous requests; aided language stimulation supports AAC initiation and AAC does not suppress speech
- ASHA, Augmentative and Alternative Communication Practice Portal: Aided language stimulation and modeling on the device are recommended to support AAC initiation; AAC does not suppress speech development
- Carter et al. (2011), 'It Takes Two to Talk Hanen Program RCT', Journal of Speech Language and Hearing Research: Randomized controlled trial showed Hanen It Takes Two to Talk improved parent responsiveness and child communication initiation in late talkers
- Cochrane Database of Systematic Reviews, Picture Exchange Communication System (PECS) for autism: Cochrane review found PECS improved initiating requests with some evidence of increased speech; overall quality of evidence rated low-to-moderate
- Kasari et al. (2014), JASPER RCT for minimally verbal children, JAMA Pediatrics: JASPER significantly increased initiating joint attention over 10 to 24 sessions in minimally verbal autistic children aged 5 to 8
- Prizant & Duchan (1981), 'The Functions of Immediate Echolalia in Autistic Children', Journal of Speech and Hearing Disorders: Echolalia serves communicative functions including initiation in autistic children and is not simply meaningless repetition
- American Academy of Pediatrics, Media and Young Minds Policy Statement: AAP recommends avoiding digital media other than video chat for children under 18 months and limiting to one hour for ages 2 to 5