Assembled spelling: There is, however, as there is with reading, an indirect route available. This can be used as a back-up system, a check-up system or when we wish to write new (or nonsense) words for which there can be no entry as yet in any lexicon. And we know we can find plausible candidate spellings for absolutely new, or nonsense, words, like ‘malversation’ or ‘spinkleduff’. We are back at the idea of some sort of grapheme-phoneme / phoneme-grapheme conversion link as we proposed for reading by sound. Language can be assembled into phonemes, which can be converted into graphemes (or candidate letter patterns) and thence to writing as before. This is indirect, or assembled, spelling, though. Such spelling by sound is, as phonically mediated reading is, a roundabout way to the goal. It is also, as assembled reading is, a commonly experienced secondary event, experienced as a result of inevitable spreading activation through the language systems. As a result, as our phonemic code representations are secondarily activated it can feel as if we are spelling by sound even though we are not.

Spelling by sound is optional; fluent spellers do it visually. Getting this right matters in the same way, and perhaps even more, than it does with reading. Students have a (sometimes unfortunate) tendency to listen to their tutors. If tutors advise that spelling is to be attacked through sound - as a phonic activity - there may be long-lasting results. Spelling (and reading perhaps) may forever be mediated indirectly, round the conversion links. Literacy may become, to the unfortunate student, exclusively a matter of turning sounds into letters and letters into sounds – ‘a matter of sounding out’. The indirect, assembled routes may be the only ones ever consulted. Using these indirect pathways clearly makes the establishment of good links (in a word, learning) more difficult. Such unnecessary complication, the insistence on such unnecessarily expensive procedure and the involvement of so many extra primary representations for each item, will produce the characteristically slow and patchy success that has been seen in so many literacy classes in the past. We will have trained the brain to do something of which we know little, in one particular way, and a way it would not have chosen if left to its own devices. Odd behaviours (often just like ‘dyslexic’ behaviours) will appear. We shall have turned a straightforward and direct psychological task into what was so often the definitive ABE experience; an apparently obscure, strangely complex undertaking only fitfully illuminated by either real success or authentic understanding and almost never by genuine confidence or real autonomy.