Why We Built MySpeekie®
 

MySpeekie® is the newest application of Speech Sound Mapping. The same principles that help children connect speech and print for reading and spelling can also help non-speaking children communicate.

Many of the children we work with are toddlers, autistic, non-speaking, or have significant speech and language difficulties. We found that communication was often the biggest barrier of all.

Traditional AAC systems can be life-changing, but many rely on children navigating multiple screens, selecting symbols, recognising words, remembering where vocabulary is stored, or learning systems that can be difficult for very young children to understand.

I found that when little children have to learn about print before they can get their words voiced, it can be a long time of feeling frustrated when they're not understood.

So I flipped it.

MySpeekie®, developed with Innovate UK support, is the world's first one-screen AAC for non-speaking children who cannot yet read or spell.

Instead of searching through grids and multiple pages, children type speech sounds using Phonemies®. The words appear on screen and are voiced immediately. This means they can communicate anything they want from the start.

A child can generate messages such as "snack please" or "orange squash in a yellow cup please" without needing to learn to read and spell first.

What surprised us was what happened next.
 

As children used speech sounds to generate words, they began noticing how speech and print connect. They were not only communicating, they were beginning to discover the structure of written language without explicit, systematic phonics instruction, simply because they were generating the code.

Children generated the words they wanted voiced, but could also see the mapping because MyWordz® uses a word mapping algorithm that handles bidirectional word mapping. When we showed the code so that children could type the speech sounds in the correct order, we were also building strong phonemic awareness. They did not need to focus on the pictures of the speech sounds, the graphemes, until they were ready.
 

When children are taught to read, the process typically happens the other way round. They are first shown graphemes, if being taught phonics, or whole words, as still often happens with "sight words", and then have to identify the speech sounds and blend them, or remember the word. There is generally a structured progression, with children only being taught certain grapheme-phoneme correspondences when they reach a particular point in a programme. For example, children are not usually taught split digraphs in week one.

With MySpeekie®, children start with the speech. They can generate words containing split digraphs and any other grapheme-phoneme correspondences because they are not restricted by a Scope and Sequence.

If these observations are confirmed through future research, they may have implications that extend far beyond AAC. They raise important questions about how orthographic knowledge develops when children are shown the code and are able to explore bidirectional word mapping from the start. By making the speech-to-print code visible when children are toddlers, we may be able to accelerate self-teaching, particularly for children who struggle with traditional print-to-speech approaches, and potentially prevent literacy difficulties before they occur.

We have also found that this appears to positively impact phoneme articulation. Parents frequently report that their child's articulation and interest in words differ markedly from siblings who did not use the system. Families for whom English is not the first language have also reported that their children develop clearer English pronunciation.

This system may not only be empowering for non-speaking children by allowing them to communicate exactly what they want to say through MySpeekie®. It may also help us understand how to send the majority of children to school already reading, even before they can hold a pencil, simply by integrating speech-to-print mapping into everyday interactions at home and in nursery. Children can use the technology to record speech, develop phonemic awareness and explore the code, even when the adults around them are unaware of the importance of these underlying skills. A stand-alone training course launches shortly. 

We have also found that generating text from speech sounds helps adults better understand English orthography. They become more aware of how the code works and why at least 1 in 5 children currently struggle to learn to read and spell.

To understand why this may be significant, we first need to understand why so many children struggle to learn to read and spell in English.