Core Words and Fringe Words: Why AAC Boards Look the Way They Do

By Venkatesh Ransing · Effara

Evidence strength: Strong · July 20, 2026 · 8 min read

Most parents open their child's new AAC app expecting to find the words they would have picked: juice, cookie, mama, the dog's name, maybe bathroom. Instead the home screen is full of small, strange choices — go, more, not, that, want, stop, like, what. It can look like a mistake, or like something designed for a linguist instead of a four-year-old. It isn't. That screen reflects one of the most replicated findings in communication research, and understanding it will change how you use the system — and what you stop worrying about.

The finding: a few hundred words do most of the talking

Record everyday conversation — anyone's, anywhere — and count which words actually get used. The result is remarkably consistent: a small set of a few hundred high-frequency, reusable words accounts for roughly 80% of what people say, day in and day out. The same small set keeps surfacing across ages, situations, and even languages (AAC Institute, Core Vocabulary and the AAC Performance Report).

These are core words. Look at what's on the list and you'll notice what's missing: nouns. Core is dominated by verbs, pronouns, little location words, describing words, and question words — go, get, put, more, you, me, it, in, on, big, all done, what, where. The nouns you expected — juice, cookie, dinosaur — are fringe words: the personal, specific vocabulary that fills in the remaining ~20%, different for every person (AAC Institute).

One number to be careful with, because it circulates in tempting shortened form: this does not mean a handful of words gets you 80%. Roughly 50 words covers only about 40–50% of everyday communication; reaching ~80% takes a few hundred. The breadth is the point — which is why serious AAC systems carry real vocabularies rather than a page of favorites.

Why core words beat a board of nouns

The power of a core word is that it works everywhere. Go works for the swing, the car, the door, the bath, and get-me-out-of-here. More works for tickles, toast, bubbles, and one more episode. Stop and not work on anything a child objects to — which is to say, they hand the child refusal, one of the most important powers a voice has. A noun, by contrast, works only in the narrow slice of life where its object appears: ball is useful near balls (Speech in the City, Selecting Core Words).

There's a second, quieter reason cores matter, and it's about what a vocabulary teaches. A board made mostly of snacks and favorite objects quietly teaches that communication is for requesting — that the device is a vending machine. As SLP Stephanie Ekis puts it: "When we restrict vocabulary, we unintentionally teach students that communication is about getting things, not expressing themselves." Her contrast is worth keeping: "Nouns help students talk about things. Core and descriptive words help them talk about ideas — and ideas are where real communication lives" (Ekis, My Dynamic Therapy). With core words, a child can comment (like, silly, big), protest (stop, all done), ask (what, where), and joke — the full range of jobs communication does, not just the transactional one (AssistiveWare, Core word teaching strategies).

There's also a speed argument, from the research tradition that studies AAC performance: effective communication comes from fast, automatic access to the words used most often (Hill & Romich, via AAC Institute). If 80% of communication runs through a few hundred words, those words belong up front, in stable positions, where finding them can become a motor habit — which is why good systems keep core words in fixed locations that never shuffle (Thistle et al. 2018) (evidence: one small controlled study in typically developing children; the consistency principle is also clinically endorsed).

But fringe is not the junk drawer

Here's where honest evidence reporting cuts against the tidy version of the core-words story, and it matters for how you set up your child's system.

Recent peer-reviewed work has pushed back on treating word frequency as the whole answer. A 2024 study analyzing how toddlers actually build utterances concluded that frequency counts alone are an insufficient basis for choosing an AAC vocabulary — and, strikingly, that utterances combining both core and fringe words were grammatically and semantically richer than utterances built from either type alone (Binger et al. 2024) (evidence note: the analysis is of typically developing toddlers' spoken language, applied to AAC by extrapolation). A narrative review focused on young children reached a compatible caution: published core lists may under-emphasize exactly the kinds of words that dominate a young child's early expressive vocabulary — the names of the people, things, and routines they care about — and so shouldn't be treated as the definitive guide for the earliest communicators (Laubscher & Light 2020).

In other words: the field's own research says core-only dogma is as wrong as the noun board. Fringe words are not a reward tier or an afterthought — they're where your child's motivation, identity, and specificity live. Go is powerful, but a child who loves dinosaurs deserves dinosaur; nothing recruits a new AAC user like discovering the device can say the thing they actually care about. The evidence-backed design is both: a strong, stable core doing the grammatical heavy lifting, plus an easy path to personal fringe — your people, your foods, your obsessions — added freely as life happens (AssistiveWare, 4 things every robust AAC system has; Binger et al. 2024).

Where the word lists come from

You don't have to take any single list on faith — it's worth knowing how they're made, because it explains their family resemblance.

Core lists come from vocabulary research: recording real communication across settings and counting, then checking which high-frequency words recur across many speakers rather than being one person's habit (AAC Institute; Laubscher & Light 2020). Different teams, different decades, different populations — and substantially overlapping lists, which is exactly what you'd expect if the underlying pattern is real.

A concrete, freely available example: the Universal Core — 36 words selected by the Center for Literacy and Disability Studies at UNC Chapel Hill by systematically combining existing core lists, the vocabularies built into speech devices and apps, written-language frequency data, and the demands of school communication. It's published under a Creative Commons license and designed for learners with significant communication needs, usable by touch, eye gaze, or partner-assisted scanning (UNC CLDS, Universal Core Vocabulary). Thirty-six well-chosen words is not a limit — it's a proven starting layer that real systems build well beyond.

What this means in practice

Four takeaways, each one something you can act on this week:

Don't "fix" the board by swapping core for snacks. The strange little words are the engine. Add fringe alongside them — never at their expense.
If the board overwhelms, hide — don't rearrange or delete. Most systems can mask words while keeping every position fixed, so the board simplifies visually without breaking the motor map your child is slowly building (AssistiveWare, Choosing a grid size; Thistle et al. 2018).
Model the core; feed the fringe. Core words are what you'll mostly tap while you talk (go, more, stop, like — one or two per sentence, no performance required); fringe words are what you'll mostly add as your child's interests announce themselves. The how-to lives in our modeling article (Related reading).
Judge the system, not the child, by the vocabulary. A child "not using the device" whose board holds six snacks hasn't failed at AAC — they've been given very little language to succeed with. A robust system offers core, personal fringe, many communicative functions, and room to grow (AssistiveWare, 4 things; Farrall, What is Beginning AAC?).

What the evidence does and doesn't show

Robust: the core/fringe frequency pattern itself — a small, stable set of words doing the large majority of conversational work — is one of the better-replicated findings in this space, across ages, contexts, and languages (AAC Institute). The access-speed logic (frequent words deserve fast, stable access) and the fixed-location principle are supported experimentally and endorsed clinically (Thistle et al. 2018).

Honest limits: "core-first design produces better language outcomes than alternatives" has not been tested in head-to-head trials — the design case is an inference from vocabulary research plus clinical consensus, and this article says so rather than dressing it up. The freshest peer-reviewed work complicates rather than crowns the core story: frequency alone is an insufficient selection basis, mixed core-plus-fringe utterances are richer, and core lists may shortchange the words young children reach for first (Binger et al. 2024; Laubscher & Light 2020). That's not a reason to abandon core — it's the reason the recommendation throughout is core and fringe, never either alone.

If Effara is relevant here

For transparency: we make Effara, an iPad AAC app, and its vocabulary is built on exactly the logic above — a fixed core-word board whose positions never move (new words appear by unmasking, not rearranging), plus a deliberately easy path for families to add personal fringe with photos and voices from home. Everything in this article applies to any robust AAC system.

Sources

AAC Institute. Core Vocabulary and the AAC Performance Report. aacinstitute.org
Speech in the City. Selecting Core Words. speechinthecity.com
Ekis, S. What Are We Afraid Of? Why Restricting AAC Vocabulary Holds Students Back. My Dynamic Therapy. mydynamictherapy.com
AssistiveWare. Learn about core word teaching strategies. assistiveware.com
Thistle, J. J., Holmes, S. A., Horn, M. M., & Reum, A. M. (2018). Consistent symbol location affects motor learning in preschoolers without disabilities: Implications for designing augmentative and alternative communication displays. American Journal of Speech-Language Pathology, 27(3), 1010–1017. PubMed
Binger, C., Magallanes, P., San Miguel, V., Harrington, N., & Hahs-Vaughn, D. (2024). How toddlers use core and fringe vocabulary: What's in an utterance? American Journal of Speech-Language Pathology, 33(4), 1718–1747. PubMed
Laubscher, E., & Light, J. (2020). Core vocabulary lists for young children and considerations for early language development: A narrative review. Augmentative and Alternative Communication, 36(1), 43–53. PubMed
Center for Literacy and Disability Studies, UNC Chapel Hill. Universal Core Vocabulary. med.unc.edu
AssistiveWare. 4 things every robust AAC system has. assistiveware.com
AssistiveWare. Choosing a grid size. assistiveware.com
Farrall, J. What is Beginning AAC? janefarrall.com