The NFER's report on the second year of the government's phonics check is worth reading. It shows a generally positive impact of the check on schools' approach to early reading, combined with a feeling among half of the teachers of young children that there is too much emphasis on phonics, which, they feel, should be used alongside "other methods to teach children to decode words". The problem is that they do not say what these other methods are. Only one per cent of teachers surveyed said that phonics needed to be taught in a way that did not pay attention to meaning - look at Ruth Miskin's Read Write Inc and you will find meaning emphasised from the very beginning.
But meaning in an alphabetic language is communicated through the use of letters, predominantly, but not exclusively, to indicate the sounds of words. The only way to "decode" a word is to use the letters that make it up. Once we have identified the word and remembered it, we do not need to "decode" it again - we know what the letters are telling us about this word and identify it from what we have learned. In English, an are of our brain adapts itself specifically to the identification of variations in what the letters tell us, in a way that is not required in languages such as Italian, where the correspondence between sounds and letters is almost entirely consistent. Professors Utah Frith and Sarah Jane Blakemore have called this the "word form area".
The alternative "methods", which are not described anywhere in the report, are guesswork, either from pictures, context, or the first letter of the word. Guessing from pictures, as I demonstrated in Susan Gets Stuck (TES 1992) depends on picking the right clue from the picture. In the example I quoted, a picture of children helping in the garden of a house was accompanied by the text "Biff and Chip liked the new house." "Susan" read this as "Biff and Chip were in the garden" - a mile away from the original meaning. The unreliability of contex clues was demonstrated by Schatz and Baldwin in 1986, and later by an experiment in which undergraduates were asked to predict the next word of a text, and usually got it wrong. The definitive research on this point was Dr Morag MacMartin's thesis, Factors Affecting Reading Comprehension in Primary Pupils, 1992, which showed both that pictures could put entirely the wrong idea into pupils' minds, and that they were also liable to misread subsequent words to try to make sense of an error they had made earlier. Guessing from the first letter is so obviously unreliable as to be virtually incredible as a teaching technique - I've known pupils read "the" as "ten" on that basis, and to read every word beginning with f a "fish". The inferiority of this approach was demonstrated clearly in the early Clackmannanshire studies.
This does not, of course, mean that phonics are everything. Roughly a third of English spelling does not correspond to patterns in modern English speech, for reasons set out by Professor David Crystal in Spell It Out. Children need to learn to deal with this, and the best way of teaching them to do so is to explain why things are as they are. This is the source of my success in work with people with reading and spelling difficulties ranging from complete non-readers, to the highest attaining pupils who have problems with spelling - my latest in that area, from a university town in the north, achieved an A* in English that she wasn't expecting after just three lessons by telephone. She had this grade in everything else except English, and it was spelling that was inhibiting her writing. What I didn't tell her to do was guess. Ever.