First Things First
The project is based on a yet untested idea, that a speaking page will make it easier for learners to learn to read, and increase their motivation to do so. So step one is to test the basic assumption, by building a prototype app that can be used by learners, teachers, parents, and illiterate adults, to get their feedback and measure its impact.
The effectiveness of the app will depend on its design, the user interface, and the materials available inside it. Phase 1 of the project will involve building the app, and writing the initial texts for beginning readers.
Test, Test, and Measure!
Reading competence is very easy to test and measure, so it should not be difficult to get good data on the usefulness of the app. If it is as helpful as we suspect, there will also be plenty of testimonials available from learners, teachers and parents, as well as a great deal of guidance about how best to develop and spread the use of the app. So the second phase will focus on putting the app into the hands of several hundred learners at different levels of reading ability, as well as parents and teachers, and getting hard data on the progress they make by using it, and the difference it can make both with beginning readers and with disaffected learners who have lost confidence and motivation.
You said what?
Local accents can make words unrecognisable to those with a different accent. For children learning to read, if the sound of a word is not familiar, they have a double translation to do, from text to unknown sound, to known sound. This could make learning much harder. So the app will make it possible for any book to have alternative sound libraries attached to it, so that a reader from London, Glasgow, Alabama or Delhi can hear the recognisable sounds of their local accent when they touch for help. Using the Wikipedia model, we can make it possible for members of the community to add local versions, making the programme locally relevant worldwide.
We want to test the app with one set of materials that are based on what is generally known as Received Pronunciation, BBC English or Oxford English; and soon after with another set based in a strongly differentiated regional accent, in the area where that accent is spoken. Doing so thoroughly will mean reclassifying many words based on the regional pronunciations, for instance so that "bath" rhymes with "hath" in one case, and "hearth" in another. It will mean creating beginner stories or texts that are congruent with the accent, which may mean a different selection of words in the early stories. And it will mean re-creating the syllable-by-syllable readings of many words, as both the sounds and the syllable breaks will vary between accents.
If the use of accent-specific learning materials proves to be helpful to learners, we will make efforts to create many varieties, covering both British regional accents and others (e.g. Alabama, New York, Nairobi, Sydney and Delhi). We will also make it as easy as possible for other people to create these versions, opening the platform to a large degree of crowdsourcing, as has led to the huge success of Wikipedia.
Writers and Publishers
There are many "learn to read" books on the market already, which we expect would benefit from being sound-enabled. We will make it as simple as possible to import such books into the system, adding sound and syllable pronunciation through automation, and probably with the help of AI. Our intention is to go far beyond "learn to read" books and encourage writers great and small to add their books to the system, so that the breadth and depth of available materials grows rapidly.
Once the system is working well in English, we want to clone it in other languages, so that in the developing world, where there is often a dearth of skilled teachers and learning materials, local people can create their own learning materials, and boost literacy in their local languages. This is particularly important in India, with 20 official languages, and in Africa where the number of spoken languages is in the hundreds, and resources are often scarce. Again, the local authors would have the choice of making their materials free or paid, at prices relevant to the local economy.