I’ve been mucking about with models (this is just for people I’ve been talking to! It probably makes no sense….)
For Marian & Jelly & Bean:
Each coloured oval is an ‘ingredient’; you can combine them differently and get different second-order ‘ingredients’ of reading.
Aural vocabulary store + visual symbolic store allows phoneme-grapheme correspondences.
Then phoneme-grapheme correspondence + known referent allows decoded word.
Then the decoded word understood in context is ‘reading’
A known referent + word-sound from aural vocabulary store = one element in an aural lexicon (ie the vocabulary of a pre-literate child or non-literate person)
This element from aural lexicon + context allows understanding of the word in relation to context.
This meaningful contextualized sounded word + visual symbolic knowledge (via the concept of words in text) allows ‘reading’.
That is, each step in towards the middle adds more of the relevant aspects together: ‘reading’ does not exclude grapheme-phoneme correspondence, but needs to combine it with knowledge and context via an aural lexicon and concept of a ‘text’ in order for reading to take place.
So each of the little triangular bits around the centre is something that is *nearly* reading, but lacks one element, which you get by moving over the line into the section which overlaps with the missing aspect (the ‘Reading’ section).
eg If you read ‘the film star showed off her £25,000 engagement fnurg’, you can sound out the sounds (aural + visual), and guess (from context) that fnurg might mean ring, but without a known referent you can’t know. It could be ‘present’, or even ‘calendar'(!)
I’m still trying to think of better ways of labelling the different bits, though…..