We are often asked about the reason that words are spoken individually in the Speak for Yourself augmentative and alternative communication (AAC) app instead of allowing the user to compose their message and then speak the full sentence. The reason it is set up that way is so that the AAC user gets immediate auditory feedback on each word that he presses. This allows him to hear (and see) where one word ends and the next begins. The word segmentation breaks language down into more manageable parts and in many cases, (I believe) that this helps to increase verbal speech and also language comprehension skills.
Neurologically, the feedback loop produces convergence so that when an AAC user touches the buttons (motor movement), then hears the word being produced (auditory), and then sees something happen (visual), whether it’s someone responding to what he said (handing him milk) or seeing the words go up to the message window, he gets visual feedback. To experience convergence in your own brain, take a look at this video. Our senses do not process information in isolation. When your brain receives multisensory input in succession, it starts to group them together so that the movement becomes faster and more automatic. Donald Hebb is credited with saying, “Neuron’s that fire together wire together,” and Hebb’s Law explains how this neuroplasticity works. For those of you who love neurology (no sarcasm intended because I really do love neurology), here is an article with more details.
Another important reason to segment language has to do with our ability to know when one word stops and another one starts. Here’s an example:
The following sentence (did anyone know it was a sentence?)
私はクッキーを食べたいです
is pronounced: Watashiwakukkīotabetaidesu
Now, I’m going to tell you that it means, “I want to eat a cookie” (in Japanese). If you were hungry, could you pick out the Japanese word for “eat”?* Unless you already speak Japanese, chances are you would have a difficult time isolating it. When we speak a language, we segment the words and “imagine” the spaces in between them.
* The Japanese word for eat is “tabemasu,” which changed forms in the sentence above to combine with “want.”
The spaces between the words are something we are able to “hear” once we know a language. If you listened to someone speaking Japanese, unless you know some individual words, it sounds like a long stream of sounds and syllables. Once you can identify an individual word, you start to hear that word and then are able to identify where the breaks are in language. When you can do that, you can start to recombine words to form novel utterances. Instead of hearing “I want milk”, and memorizing that to know that he gets milk, a student who is using an AAC system that speaks each word realizes that milk doesn’t always have to be part of “I want” and can start to use it in new ways to say things like “chocolate milk is better” or “that milk spilled.” The more words he learns and is able to “pick out”, the more language he is able to segment and understand when others are speaking as well. Here’s a link with research citations describing this in more detail.
Here is a research article from Cornell talking about this phenomenon as well as Hebb’s Law and how it impacts an autistic individual who is hearing and filtering language. Here is one example:
“For example, if an autistic child hears “Don’t touch the stove” repeatedly, while not hearing other phrases that involve “don’t”, “touch”, or “stove”, then “Don’t touch the stove” may become fixated. A consequence of this is that, rather than using the word “stove” appropriately, the whole phrase will be uttered. We stress that the fact that the child does not hear “don’t”, “touch”, or “stove” in other phrases does not necessarily mean that such phrases are never uttered in the child’s presence. Rather, it means that, due to the child’s filter, the child does not pay attention to these phrases.” (this portion of the article is found on page 5)
Here’s another article from a University of Connecticut Master’s Thesis that looks at this word segmentation, particularly as it relates to autism:
“Specifically, one of the initial steps in learning language is determining word boundaries from within a continuous speech stream. Segmenting a continuous stream of speech into meaningful linguistic units (i.e., words) is a fundamental requirement of language learning, but is particularly difficult because auditory pauses do not reliably occur between words and therefore do not provide a useful cue to word boundaries (Cole, Jakimik, & Cooper, 1980).” (page 2)
In addition to all of the research, it is easier to understand for a listener when they hear each word, and it also holds the AAC user’s place in the conversation. If you hear each word, you know he’s saying something and you can hear him building the sentence. This enables the social conversational cues needed for listeners to not interrupt, but also to stay engaged. I have seen adult AAC users who do not speak each word, attempting to simultaneously keep the listener’s attention and compose their message. It takes some time for people to become good AAC listeners. When an AAC user is speaking their message word by word, it gives listeners the benefit of hearing individual words and then hearing the complete sentence. It allows listeners to have some context before the entire message is given. You may not see the benefit for messages like “I want milk” but when AAC users begin to get more complex and tell narratives, it makes all the difference.
Leave a Reply