The following is not a full response to ‘Teaching phonics in the National Literacy Strategy’ (TPNLS), but offers views on three related points which are particularly important in the current climate:
1. blending (last paragraph on p. 11)
This paragraph clearly describes the NLS version of ‘blending’ as something which takes place after the whole word has been pronounced and segmented into phonemes; letters are used to represent the phonemes thus produced, not to prompt the production of phonemes. This is a spelling routine. By contrast, the genuine synthetic phonics version of blending is part of a reading routine, where the pronunciation of the whole word is the product of sounding out and blending, not a precursor to it. The NLS version of blending simply takes children back to a word heard a few seconds earlier. In real reading, however, the target word is not heard in advance but has to be retrieved from among thousands in the child’s oral vocabulary by the production of the phonemes indicated by the letters and the blending of those phonemes. By comparison, the NLS version of blending for reading, though not totally pointless, is very weak, providing little real exercise for children’s blending muscles.
If ‘going from the segmented word back to the blended word’ is just an introductory step (as may be implied by this paragraph of TPNLS), where, in the NLS materials, is the work on real blending – synthesising grapheme-prompted phonemes to work out a pronunciation which is not already known?
2. ‘Synthetic’ versus ‘analytic’ phonics (pp. 16-18)
As TPNLS points out, the situation is complicated by the fact that ‘there are subtly different interpretations of the terms “analytic” and “synthetic” in the literature’ (p. 17). It then selects one of these interpretations (‘synthetic’ means ‘direct/explicit’ rather than ‘indirect/implicit/constructivist’) to argue that the NLS approach is synthetic. The problem is that although the NLS may indeed meet criteria of explicitness in some respects, it is not particularly explicit in its teaching of genuine synthesising. It does, however, meet a key criterion of ‘analytic phonics’ as expressed in the Feitelson and National Reading Panel (NRP) extracts on TPNLS p. 17: whole words are ‘taught’ (Feitelson) or ‘identified’ (NRP) first and then analysed. If even blending (synthesising) starts with the whole word, as TPNLS itself says is the case (see 1. above), the case for calling the NLS ‘synthetic’ is weak.
Bielby, cited on p. 16, calls the NLS ‘synthetic’, but he, like TPNLS, equates ‘synthetic’ with ‘explicit’ without considering how explicitly synthesising is taught. His work shows that he himself believes in analysis after the acquisition of ‘a sight vocabulary of whole words’. By contrast, Chew regards the NLS as non-synthetic because of its lack of emphasis on synthesis as a word-identification strategy for previously unidentified words, which is the true function of synthesis in reading. Contrary to what is suggested in TPNLS (pp. 16-17), however, Chew did not say unequivocally in RRF Newsletter 45 that the NLS was ‘analytic’, but just that ‘most of the work on phonemes.... leans towards the analytic end of the spectrum, frequently with the assumption that the word is first recognised as an unanalysed whole’.
To counter the allegation that the NLS ‘leans towards the analytic end of the spectrum’, can it be shown that the NLS materials give equal attention to activities involving synthesis as a means of word-identification in reading and analysis of words already identified?
3. the nls searchlights model (pp. 13-16)
The OFSTED report ‘The National Literacy Strategy: the first four years 1998-2002’ criticised the searchlights model on the grounds that it ‘gives insufficient emphasis in the early stages to the teaching of phonics’ (p.3). OFSTED also stated that this model
‘has not been effective enough in terms of illustrating where the intensity of the “searchlights” should fall at the different stages of learning to read.... [B]eginning readers need to learn how to decode effortlessly, using their knowledge of letter-sound correspondences and the skills of blending sounds together. The importance of these crucial skills and knowledge has not been communicated clearly enough to teachers’ (p. 17).
The first need of beginners is indeed to learn to decode. This should mean that letter-prompted sounding out and blending is done as a means to identify words, not after words have been identified. Decoding/word-identification should be ‘context-free’ (Perfetti in Journal of Research in Reading, 18/2, 1995, p. 108). This view is based on research and common sense, not on ‘ideological preference’ as TPNLS suggests (p. 16). Accurate decoding often suffices to ensure comprehension, especially with simple beginners’ texts. If decoding does not produce comprehension, grammar and context can then be considered. TPNLS, however, accepts the use of ‘syntactic and contextual inferences’ (p. 14) for word-identification purposes in text-reading. Admittedly it adds that children should ‘“loop” back to the phonic and orthographic features of new words’ but this, as in NLS-type blending (see 1. above), is phonics after word-identification, not phonics for ‘context-free’ word-identification and not even phonics used ‘simultaneously’ with the other searchlights as recommended on p. 1 of Progression in phonics (PiPs). No wonder phonics is under-used in text-reading. The message to teachers on pp. 1-2 of PiPs can easily be interpreted as follows:
Let the children continue to read ‘familiar and predictable’ texts (e.g. from ‘Book Bands’), where they ‘can often rely heavily on contextual and grammatical knowledge’, but get them now to pay more attention (rather than ‘relatively little attention’) ‘to the sounds and spellings of words’. Let them still be dependent (though not ‘over-dependent’) ‘on remembering or guessing their way through the text’, exactly as illustrated on NLS videos, but make sure that their ‘limited phonic strategies’ become slightly less limited. Indeed, they don’t really need ‘to rely more on their ability to decode individual words’ until ‘later’, when ‘the familiarity of texts diminishes’.
TPNLS states that ‘the ability to decode words remains the first and only direct means of getting meaning from the page’ (p. 4). But does this square with the NLS’s great stress on using phonics after words have been identified or, at best, simultaneously with the other searchlights?