Years ago a brilliant little girl changed the way I viewed AAC. Amber was four years old, preliterate and nonverbal. She came to speech therapy with a AAC device that she carried like a briefcase.
She always had wonderful and clever things to say but one day she stopped me in my tracks. She came to speech with some tootsie rolls that I held onto for her. As we were singing and playing and doing speech-y stuff, she interjected by pressing the word “chalk” (stored on her device with other art supplies; the icon was a picture of a white, cylindrical object against a black background). Immediately after, she pressed the word “let” (stored as a verb with a little family as the icon).
Chalk. Let.
Chalk let.
Chalklet.
Holy moly. She’s saying “chocolate.”
This gave me incredulous pause. And then it hit me- of course this is how she thinks. This is how language development works! When beginning verbal communicators speak they aren’t referencing icons or graphemes. They move their articulators to produce sounds that form words. Amber was tapping into this same developmental language learning process – just with a different articulator. Instead of moving her lips, jaw and tongue to create sounds and words, Amber moved her finger (incidentally, this inspired my AAC tattoo).
“Chalk let” was likely not the first time she reappropriated words and manipulated sounds to create her message and it definitely was not the last. Over the years, her mother and I collected examples of Amber’s use of programmed sounds and words:
- “Apple saws” for “applesauce”
- When asked about the fireworks on the 4th of July, “color foal”
- “Turn up” for “turnip”
- “No L” for “Noel” (in response to her mother spelling the word and the synthesized speech mispronouncing it as “nole”)
- “Some foggy putting” for “some figgy pudding”
- “Are” with one hand over her eye, pretending to be a pirate
- “Square L” for squirrel
- “Chick fell a” for “Chik Fil-A”
- “Tummy 8” for “tummy ache
I learned from Amber to start listening to the output on my clients’ devices instead of reading or watching them generate language. I need to hear what she’s saying and what she’s hearing.
Now stuff like this happens in the language of neurotypical verbal children. My kids (twin 6 year olds) tell me that they have a “head egg” when their heads hurt. This is because the word “ache” isn’t in their vocabulary yet but the word “egg” is, so they actually hear “head egg” when someone else says it, then their output matches their input. Most parents have similar stories to this. My own parents still laugh at me getting excited to “go to see the rabbits” when in fact we were going to Cedar Rapids.
Beginning AAC communicators have the ability (and right) to develop language using the same processes and strategies and by making the same mistakes. The more I allowed myself to listen to Amber the more I heard what she heard and what she was trying to say. This phenomenon is a clear case for giving beginning communicators access to speech output devices. We all know that there is power in sign language and low-tech systems but letting kids play with sounds will allow children to mirror some natural language development processes.
Excitingly, Amber isn’t completely unique. My colleagues at Bridgeway Academy started hearing similar vocabulary reappropriations from their clients (“water small” for “waterfall,” “draw” for “straw,” “happy feel to” for “Happy Feet 2,” “sleep be” for “sleepy”). One of my students called me “Win see” and another “more on!” My disability awareness and advocate BFF, Robert Rummel-Hudson told me his daughter Schuyler at 13 years old and just started to experiment with swearing used the word “apple” to describe a teacher she didn’t like!
Now, for the sake of clarity and literacy, I tend to first accept the utterance, but then share the proper production of a word, just like we do with invented spelling for beginning writers.
A couple months ago one of my clients asked me to “sing rowboat,” then adding “life is but a tree.” I shared this on Facebook and a colleague in the AAC field (Russell Cross) was curious about how many other people had experienced similar things. The responses from across the globe were overwhelming. Here are a few:
- “Giraffe +ing” for “drafting”
- “Love bye” for “lullabye”
- “Pizza hunt” for “Pizza Hut”
- “Peach therapy” for “speech therapy”
- “Ice see you” for “ICU”
- “3 blonde mouse” for “3 blind mice”
- “”Jacket” for “Jacqui”
- “Bowl” for “bolt”
- “When she’s” for “Wendy’s”
- “Mad Gas Car” for “Madagascar (we’ve seen this one at Bridgeway Academy too!)
- “Alike” for “I like”
- “Me key House” for “Mickey Mouse”
- “Speak well” for “speak whale” (from Finding Nemo)
- “March yellow” for “marshmallow”
- “Bible” for “bubbles” (this was after the team hid the word “bubbles” from the child’s device; he still found a way to communicate his desire and it’s probably a good example for why NOT to hide words- it doesn’t change the fact that someone is interested in something!)
- “Soda” for “Sodor” (where Thomas the Tank Engine lives)
- “Be Lake” for “Blake”
- “You corn” for “unicorn”
- “House come” for “how’s come”
This phenomenon is clearly not isolated to a few kids. Even some literate adults report doing similar things (“off ten” for “often” or “touch e” for “touchy”). These examples give tangible rationale for giving preliterate children robust vocabularies, speech output and most importantly, presumed competence!
Here’s a bonus video of a fabulous girl named Katya saying “ah chew” to pretend to sneeze
Lindsey's AAC Blog 
Hi Lindsey, I am presenting on just this subject at the ISAAC in Australia in August “(A glimpse into the sausage factory” Thursday 2pm). I have been tracking the Realize Language data of a little german girl, Selma for 18 months and have many dozens of examples. More to the point there must be even more examples that I didn’t catch because they were too far off, or I was confused by the apparent meaning. The basis of this part of my talk is Levelts explanation of articulation which I have adapted for device users. I am looking at how Selma choses items express herself – sometimes semantically sometimes phonetically (sometimes one of 3 other methods).
The one thing I would disagree with you above is that I think you are not taking the idea far enough. You talk about this language feature as “mistakes” – I think it is a strategy. A strategy that is not only “allowable” but often necessary not just to acquire language but to articulate it too. It is interesting that there is loads of literature on language development in aac but almost none in methods of articulation. (Shout-out here to ELENA DUKHOVNY whose “Phonological Encoding in Aided Augmentative and Alternative Communication” is a must read on the subject).
Your examples nicely show that kids are processing successive chunks of what the need to say.
This idea is food for thought on many things including how we model (you mention listening to the device – my contention is that if we read at the same time we can’t “hear” what is said, so where do we sit? Does anyone actually teach this kind of word selection. I know I never have in any systematic way.
LikeLike
Hi Paul-
I totally agree that this is a strategy and merely meant to liken these as “mistakes” to invented spelling. I am confident that the use of phonology and using stored words as morphemes or sounds is a strategy.
I’d very much love to read about your adaptation of Levelt’s model to AAC!
I have read Dukhovny’s work and know that we need a lot more research and attention to this. I’d love to see more AAC displays designed with this in mind.
It’s nearly impossible to not read what the kids so say, so I find myself pulling in a third party to whom I do not show the text and ask them to just listen.
I have modeled and taught some of this vocabulary use but not much and certainly not systematically. I do teach and model adding -ing and -s as early morphemes; I want kids to know how to add these word endings because we cannot always rely on the SGD having present progressives/gerunds and plurals prestored.
Thanks for your thoughts. I certainly hope this essay can be a springboard for people’s thinking.
LikeLike
Hi Lindsey, I am presenting on just this subject at the ISAAC in Australia in August “(A glimpse into the sausage factory” Thursday 2pm). I have been tracking the Realize Language data of a little german girl, Selma for 18 months and have many dozens of examples. More to the point there must be even more examples that I didn’t catch because they were too far off, or I was confused by the apparent meaning. The basis of this part of my talk is Levelts explanation of articulation which I have adapted for device users. I am looking at how Selma choses items express herself – sometimes semantically sometimes phonetically (sometimes one of 3 other methods).
The one thing I would disagree with you above is that I think you are not taking the idea far enough. You talk about this language feature as “mistakes” – I think it is a strategy. A strategy that is not only “allowable” but often necessary not just to acquire language but to articulate it too. It is interesting that there is loads of literature on language development in aac but almost none in methods of articulation. (Shout-out here to ELENA DUKHOVNY whose “Phonological Encoding in Aided Augmentative and Alternative Communication” is a must read on the subject).
Your examples nicely show that kids are processing successive chunks of what the need to say.
This idea is food for thought on many things including how we model (you mention listening to the device – my contention is that if we read at the same time we can’t “hear” what is said, so where do we sit? Does anyone actually teach this kind of word selection. I know I never have in any systematic way.
(PS I looked at your examples above and saw “dropped/trapped”. I was just thinking today whether “looks like” might be a further natural selection method (currently I use phonetic, semantic, automatic, and deduced). Sometimes Selma says stuff what I just can fathom, until I half close my eyes and see that the word she chose has the same sillouette as the one she was trying to express.)
LikeLike
I remember when my son used to ask when we were going to Pigs bird (Pittsburgh).
LikeLike
Oh my gosh, YES. I have a client who uses this strategy all the time (e.g., “itch + you” = issue). However, because unfamiliar listeners tended to more literally interpret the words, we added a button that says “It sounds like…” on his REPAIR STRATEGIES page so he could let listeners know that he was using this strategy.
LikeLike
Another thought here, Lindsey – Why doesn’t this have a name.
LikeLike
Yes!
Robot for Robert as one example….
“Yahoo” …this one took a while. Yahoo was doing different things… .became clear when yahoo meant the name “Zahour”.
Many more examples from my son who is AAC user.
And going to preset words e.g. “spaghetti” on food page
Back spacing to
..spa
Add nish
To say
“Spanish”
There should be a term for that too!
LikeLike
@Natasha – bit late to reply, but what you describe is a very common thing that is especially interesting because it is specific to device users. Creating a word by going to the left rather than just to the right…..Leftwords production
LikeLike