New research out of the Massachusetts Institute of Technology has found that having a back-and-forth conversation is more critical to language development than purely hearing words.

So what does this mean for modeling?

As you might be aware, modeling is an intervention strategy used to support a new AAC user to use their communication system. To implement the strategy, the communication partner talks to the person learning to use an AAC system, with the AAC system itself. The idea is that by seeing AAC in use by other people, the user will learn how it can be used themselves to communicate.

This strategy is quite popular, and has a really solid rationale behind it –it mimics typical language development. And any time we are able to legitimize our intervention to the ‘blueprint’ of typical language development, we’re on the right track.

Think about it. How does anyone learn to speak any language? Well, for babies, they are surrounded by people speaking a language. And then, seemingly by magic, after they have listened and heard words for around 1 year, they begin to speak the language themselves.

A hand selecting a card on a mat

However this study is really interesting, as it suggests that purely modeling, that is, purely listening to words and being talked at – may not be enough.

“The important thing is not just to talk to your child, but to talk with your child” – Rachel Romeo

This research found that having a back-and-forth conversation is more critical than merely hearing words.

So, how do we translate this to people learning to use AAC?

How do you facilitate a back-and-forth conversation with a beginning AAC user who may be expressing very little (or anything) at all?

Throughout my practice as a speech pathologist, I have always seen the benefit of having scaffolded conversations with even the earliest AAC user.

Scaffolding conversations has been described as the practice where “a communication partner takes a very active role in supporting the child’s communication through…actively problem-solving what the child might be trying to say, assisting them to work out what they could say…(or even) attributing meaning to what has been communicated with AAC” even when the true meaning is unknown. – Gayle Porter

Let’s think about this for a moment. Are scaffolded conversations part of typical language development as well? Of course they are. We caN liken scaffolded conversations to the process of babbling.

In comparison to modeling, I feel like babbling is a concept rarely considered or discussed for the person learning to use AAC – but I think it makes perfect sense. We define babbling as a period of time where children play with sounds, explore making noise, and learn that they can have an effect on their partner by making sounds out of their mouths. We see typically developing kids string nonsense sounds together (“ba-ba-ba” “go-i-na-ba-da”) and the people around the child do nothing but respond with similar coos and sounds, or attribute meaning to whatever they’ve blurted out. This period is crucial, as it allows the child to play with communication, without any negative consequences.

Think for a moment about our AAC users.

Do we give AAC users the same, consequence-free time to just play with, and explore their AAC systems? String together meaningless words with an AAC device, and have people respond with nothing but positivity?

This is where I see the power of having scaffolded conversations.

Think of them as giving complete control to the person learning to use their AAC system, and going on a wild goose-chase deep into the depths of an AAC system – and responding with nothing but positivity, interest, and attribute meaning.

I’ve documented a recent scaffolded conversation I’ve had with a person I support. Millie* is learning to use a PODD book, and had so much fun just exploring her communication system and conversing with me.

I noticed that Millie might have been interested in interacting with me, because she looked at me. I went over to Millie and asked if she had something to say. Millie responded by looking at me, and, after giving Millie some pause time, I saw a subtle head movement, which I interpreted as a “yes”. So, we got out her PODD book.

Millie, through habit, pointed to “More to say” on the first page, then, by chance, Millie pointed to “I’m telling you something”. Millie is very new to her PODD book, so I doubt she knew what this cell said – regardless, I reacted positively by saying “Ooh! You’re telling me something! Awesome. Let’s see what it is”

Finger pointing to an item on a PODD communication book

Millie pointed to “It’s already happened”, again, purely through exploration. I hadn’t modelled “It’s already happened” too much with Millie, but again – I responded positively and navigated to the Categories page.

Millie pointed to “People”, then “Mum” and I responded by saying “It’s something that’s already happened to Mum, ok, this sounds really interesting. Tell me more!” (I then navigated back to Categories). Millie pointed to “Transport” and “Bus”. I expanded by pointing to additional words on the page by saying “Did mum GO on a BUS did she? Cool! What happened next?”. After navigating back to Categories, Millie pointed to “Animals” (a favourite category of hers) and pointed to “pig”.

My response was only positive: “Hey everyone! Guess what Millie just told me with her PODD! She said her Mum went in a bus and saw a pig out the window! How crazy is that!”. Millie’s response was a shining smile on her face.

This conversation was absolute nonsense.

Millie didn’t know the words she was pointing to – and in a way, she was babbling. We were having a back-and-forth conversation, and she was playing with the words in her AAC system and seeing the effect it had on me.

I think the experience of testing out a huge AAC system and gaining a positive response from your communication partner (even if you don’t know what you’re doing) can be so encouraging and supportive for new AAC users.

And it looks like the research backs this.

Is this practice not attributable to typical language development?

And does this strategy not support this new research which states that having a back-and-forth conversation is key?

The research states that a combination of hearing a wide variety of words (i.e. modelling) and having back-and-forth conversations (i.e. scaffolded conversations) each play an important role in learning language – and I have to agree in terms of learning to speak using AAC.

Modeling plays an important role, in that it exposes the beginning user to the correct social, strategic, linguistic and operational use of AAC – but I think pairing this with times dedicated to having scaffolded conversations and having a little bit of faith and fun, is as essential to ensuring the best outcomes for the people we work with.

If you’d like to learn more about scaffolded conversations, modeling, and other strategies to support AAC language development, contact Scope’s Communication and Inclusion Resource Centre on 03 9843 2000 or