AAC: Augmentative and Alternative Communication Guide
AAC — Augmentative and Alternative Communication — is an area of clinical practice and a category of assistive technology that supplements or replaces spoken or written communication for individuals with complex communication needs. An estimated 5 million Americans may benefit from AAC (Beukelman & Light 2020, Augmentative and Alternative Communication, 5th ed., Brookes Publishing). AAC spans unaided modes (sign, gesture, facial expression) and aided modes (picture-exchange, speech-generating devices, app-based systems). Communication is a fundamental human right under the UN Convention on the Rights of Persons with Disabilities, Article 21, and the ASHA Position Statement on AAC (2016).
Presume competence — the foundational principle
Every AAC intervention begins from the assumption that the user has more to communicate than they currently express. Romski & Sevcik (2005, Infants & Young Children 18(3):174-185) established that gatekeeping vocabulary based on perceived cognitive level is an evidence-rejected practice. Light & McNaughton (2012) extend the principle: robust vocabulary from day one, not a staged unlocking. Failure to presume competence is the most cited reason for AAC abandonment, which reaches 50-90% of users within the first year (Johnson, Inglebret, Jones & Ray 2006, AAC Journal).
Core vocabulary — the 80/400 rule
Fewer than 400 words account for approximately 80% of everyday spoken communication (Banajee, Dicarlo & Stricklin 2003, AAC Journal 19(2):67-73; Yorkston, Dowden, Honsinger et al. 1988). The top 23 core words — I, want, go, more, help, stop, like, not, that, it, you, on, in, up, down, is, the, a, to, and, do, have, here — appear in the majority of toddler utterances. AAC displays must prioritize these high-frequency core words over low-frequency fringe vocabulary (specific nouns like zebra or pizza). Fringe vocabulary remains accessible but never displaces core from primary grid positions.
LAMP — Language Acquisition through Motor Planning
LAMP is an AAC approach, developed by Halloran & Halloran at the Center for AAC & Autism, grounded in the principle that consistent motor patterns develop automaticity. Every word occupies a fixed motor-planning location in the grid across all contexts and sessions — the location for more is always the same coordinate, reached by the same sequence of taps. Changing symbol locations destroys learned motor plans.
Thistle & Wilkinson (2015, AAC Journal) experimentally confirmed that consistent symbol location supports motor learning in AAC, especially for preschoolers. Potts & Satterfield (2012, Autism Center of Excellence) documented vocabulary and social-communication gains using LAMP. The practical consequence for AAC app design is unambiguous: core-vocabulary grid positions are immutable. AI prediction should never reorder the grid.
Does AAC delay speech? The definitive answer
No. Millar, Light & Schlosser (2006) meta-analysis of 27 cases documented that 89% of individuals showed speech-production gains with AAC, 11% showed no change, and 0% showed a decrease. Romski & Sevcik (2005, Infants & Young Children 18(3):174-185) definitively debunks the myth, establishing that AAC supports — and often facilitates — natural speech development. Ganz, Earles-Vollrath, Heath et al. (2012, J Autism & Developmental Disorders) meta-analysis of 24 single-case studies confirms large effects of aided AAC on communication for students with ASD.
The practical implication is that early AAC introduction is clinically warranted and should not be withheld pending further attempts at speech development. Iacono, Lyon & West (2021, J Developmental & Physical Disabilities, mega-review of 84 reviews) confirms aided AAC is effective across intellectual and developmental disability populations.
Expert perspective
"The goal of AAC is not to reject speech but to provide a reliable, immediately accessible communication channel from the beginning. Individuals who use AAC communicate — and, in many cases, also develop spoken language. AAC opens the door; it does not close it."
Communicative competence — the Light & McNaughton framework
Light & McNaughton (2014, AAC Journal 30(1):1-18) define communicative competence across four domains that together define meaningful AAC use:
- Linguistic — Knowledge of the native spoken language + of the AAC system's linguistic code (symbols, grammar).
- Operational — Technical skills to use the AAC system — motor access, navigation, feature control.
- Social — Appropriate interaction skills — turn-taking, topic maintenance, partner adaptation.
- Strategic — Adaptive strategies to overcome limitations — repair strategies, conversation control, partner coaching.
Aided-language input — modeling is mandatory
Sennott, Light & McNaughton (2016, Research and Practice for Persons with Severe Disabilities 41(2):101-115) systematic review of 9 single-case studies with 31 participants established that communication partners modeling on the AAC device produces meaningful gains across pragmatics, semantics, syntax, and morphology. Kent-Walsh, Murza, Malani & Binger's (2015, AAC Journal) meta-analysis of partner instruction produced a large effect (d=1.34) on communication outcomes. Operational takeaway: adults around an AAC user must themselves use the device to narrate and comment throughout the day.
Access methods — one size never fits all
- Direct selection (touch). Most common; requires isolated finger control and sufficient motor range.
- Switch scanning. For users with significant motor impairment. Row-column scanning most efficient for large vocabularies (Lesher, Moulton & Higginbotham 1998, AAC Journal).
- Eye-gaze/head-tracking. Effective for users with severe motor limitations (Borgestig et al. 2016, Disability & Rehabilitation: Assistive Technology). Dwell time must be calibrated individually.
- Multi-modal hybrids. Eye-gaze + switch scanning outperforms single-method access for many users with CP (Mcnaughton, Rackensperger, Dorn & Wilson 2022, Assistive Technology).
Touch targets for AAC users must be larger than the 44px minimum of general UI design. Yadav et al. (2021, Human Behavior & Emerging Technologies) found 7mm targets are missed by children aged 7-10 approximately 30% of the time — IncluVoice uses 64px minimum targets for this reason, consistent with WCAG 2.2 SC 2.5.8.
AI in AAC — autonomy constraints
AI-based prediction can reduce motor actions by as much as 57% (Vertanen, Kristensson, Patel et al. 2022, ACM CHI) and accelerate composition 29-60% over traditional methods. However, Caron, Holyfield & Light (2024, AAC Journal) and Sennott, Light & McNaughton (2016, Seminars in Speech & Language) are explicit: AI must augment, never override, user autonomy. The user controls the message, not the algorithm. IncluVoice prediction is advisory only — it never changes symbol locations, never selects without user confirmation, never silences or edits the user's utterance.
How IncluShift implements AAC
IncluVoice opens instantly with zero authentication — an AAC device is the user's voice, and authentication between a nonverbal student and their "Stop" button is a clinical anti-pattern. Core-vocabulary motor-plan locations are immutable in code (motor plan registry throws if consumers attempt to reassign). Partner modeling is supported through adult-controlled mirroring. AI prediction never reorders the grid. Multi-modal access includes direct selection, switch scanning, and eye-gaze compatibility. See IncluVoice.
Key research citations
Beukelman & Light (2020). Augmentative and Alternative Communication (5th ed.). Brookes Publishing.
Sennott, Light & McNaughton (2016). AAC modeling intervention review. Research and Practice for Persons with Severe Disabilities, 41(2), 101-115.
Millar, Light & Schlosser (2006). AAC impact on speech production. JSLHR. [89% gained, 0% decreased]
Romski & Sevcik (2005). AAC myths and realities. Infants & Young Children, 18(3), 174-185.
Light & McNaughton (2014). Communicative competence. AAC Journal, 30(1), 1-18.
Banajee, Dicarlo & Stricklin (2003). Core vocabulary determination. AAC Journal, 19(2), 67-73.
Ganz et al. (2012). AAC meta-analysis for ASD. J Autism & Developmental Disorders.
Kent-Walsh, Murza, Malani & Binger (2015). Partner instruction meta-analysis. AAC Journal. [d=1.34]
Related
This page provides educational information about AAC research and practice. IncluVoice is an AI-powered communication support tool; it is not a medical device or substitute for speech-language pathology evaluation. For AAC assessment and prescription, consult a licensed speech-language pathologist.