Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the autism research overview.
Short answer. No. The research is consistent across decades and across systems — Picture Exchange Communication System (PECS), speech-generating devices (SGDs), sign language, and other augmentative and alternative communication (AAC) approaches do not impede spoken language development. Across systematic reviews and randomised trials, AAC use is associated with neutral or positive effects on speech production for autistic children (Schlosser & Wendt, 2008; Millar, Light & Schlosser, 2006; Kasari et al., 2014). The "device delays talking" concern is one of the most persistent worries among parents and the research is one of the clearest places where the worry is not borne out.
Schlosser and Wendt's (2008) systematic review of AAC interventions in autism — covering PECS, SGDs, and manual signing across 27 studies — found no evidence that AAC suppresses speech. Across the studies, 11% of children showed substantial increases in speech production after AAC introduction, the majority showed modest increases or stability, and none showed declines attributable to AAC.
Millar, Light, and Schlosser's (2006) earlier review reached the same conclusion: introducing AAC was associated with positive or neutral effects on speech production in 89% of cases. Kasari and colleagues' 2014 randomised trial directly compared parent-mediated NDBI with and without an SGD; children in the SGD-augmented arm showed greater gains in spontaneous communicative utterances and spoken words than the speech-alone arm.
The mechanism is not mysterious. AAC reduces the communicative pressure on a still-developing speech system, which often produces more communicative attempts overall, more shared engagement, and more contexts for spoken language to emerge. Children who can communicate effectively — by any means — typically show more communicative growth than children who cannot.
Three factors keep the "AAC delays speech" concern alive in parent communities and even some clinical settings.
It feels obvious that giving a child an easier alternative would reduce the motivation to do the harder thing. The intuition transfers smoothly from contexts where it's true (e.g., calculator use and arithmetic fluency) to communication, where it isn't. Communication and arithmetic operate on different developmental architectures: communicative motivation is not a fixed quantity that gets "used up" by AAC.
Some clinicians trained before the modern AAC evidence base internalised a "speech first" rule that delayed AAC introduction until speech failed to emerge through other means. This sequencing was never well-supported and has since been replaced in evidence-based practice by recommendations to introduce AAC alongside speech-supportive intervention as early as the child's communicative profile suggests it could help.
Some of the children for whom AAC is most useful are also the children whose spoken language develops more slowly. Parents — and sometimes clinicians — see the AAC and the slower speech and infer causation. The slower speech is the reason AAC was introduced, not the result of it.
The current evidence supports introducing AAC as early as a child's communicative profile suggests it could help, often well before age 3. Schreibman et al. (2015) and the AAP clinical report (Hyman, Levy, Myers, 2020) both reflect this consensus. Wetherby and Prizant's work on Early Social Interaction emphasises that communication is the underlying target — by any modality — and the modality choice is a means to that end, not the end itself.
In practice this means a child showing communicative intent (reaching, pointing, sustained eye contact, leading an adult by the hand) without spoken words is often a candidate for early AAC introduction. The goal is to give that intent a reliable channel. When spoken words emerge, they layer onto the communicative system AAC helped scaffold.
AAC is not one thing. Three commonly-used approaches differ in mechanism and in the research questions they raise.
PECS uses physical exchange of picture cards to teach functional requesting and commenting. Bondy and Frost's protocol has substantial evidence for increasing spontaneous communicative initiations in young autistic children with limited spoken language. PECS is often the first AAC introduced in clinical settings because it can be delivered without specialised technology and emphasises spontaneous initiation.
Modern SGDs — typically tablet-based apps such as Proloquo2Go, TouchChat, or LAMP Words for Life — produce spoken output when a child selects symbols. Kasari and colleagues' (2014) RCT supports SGD use in NDBI contexts. SGDs offer broader vocabulary, faster scaling as the child's language grows, and the modelling benefit of hearing the spoken word produced when the symbol is selected.
Some children develop functional communication faster through signing than through PECS or SGDs, particularly children with strong motor imitation. Signing has its own evidence base in autism, though access to fluent communication partners (other signers) is the rate limit on its long-term utility for many families.
The choice among PECS, SGD, and signing is typically made by a speech-language pathologist (SLP) based on the child's profile — fine-motor capacity, visual preference, environmental support — and is often revisited as the child develops.
The research does not establish that every child needs AAC, or that AAC must be introduced at a specific age. The decision is individualised and made with an SLP who understands the child's communicative profile.
The research also does not settle the long-term outcomes for children who use AAC throughout development. Many continue to use AAC alongside spoken language; some transition fully to spoken language; some continue using AAC as their primary modality. All three trajectories are legitimate communicative outcomes — a child who communicates effectively through AAC has acquired communication, which is the underlying goal.
The "device will delay talking" worry is not supported by the research. If a clinician is recommending AAC introduction, the evidence-based question is not whether it will delay speech — it won't — but which system fits the child's profile, how it will be implemented, and how progress will be measured over a 90-day window.
The frame the research suggests is that AAC is communication, not a substitute for communication. The goal is for the child to be able to communicate, not for them to communicate in a specific way. Parents who hold that frame typically find AAC easier to integrate into family life than parents who experience it as a concession.
---
Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full autism research overview for the complete framework.