The following summaries of parliamentary Select Committee hearings are based on the transcripts of the uncorrected oral evidence from 15 November 2004, 8 December 2004 and 7 February 2005. Points made by each witness are grouped together and are not necessarily given in the order in which they were made. Where comments and queries seem appropriate, they are inserted in square brackets.
Please note that neither witnesses nor Members of Parliament have had the opportunity to correct the record contained in the transcripts, which are therefore ‘not yet an approved formal record of these proceedings’.
15 November 2004
The witnesses on 15 November were Dr Morag Stuart, Reader in Psychology at the London Institute of Education, and Debbie Hepplewhite. Morag Stuart spoke as an academic and a researcher, and Debbie Hepplewhite spoke as a classroom teacher and former editor of the RRF Newsletter.
Points made by Morag Stuart
Before the advent of the National Literacy Strategy (NLS) in 1998, there was a widespread view that reading did not need to be taught as children would learn it as naturally as they had earlier learnt to walk and talk. Frank Smith’s ‘mystic pulling powers’ were very influential. [The influence of such apparently bizarre ideas seemed to intrigue Select Committee members – they returned to Frank Smith several times, commenting that the teaching of reading had been changed ‘on a whim’ and that teachers had been trained in the ‘whole language philosophy without any evidence that it worked’.] In one research study carried out by Morag Stuart, ‘the depth of opposition that was around when the NLS came in’ had been shown by the fact that she almost lost one school from the study because the head teacher said that she was ‘ideologically opposed to taking part in a study which showed that phonics teaching worked’. The NLS at least made it clear that reading does need to be taught.
Research evidence does not show that some children learn differently and benefit from a variety of teaching methods rather than from a firm focus on phonics. This was one of the conclusions of the USA National Reading Panel. As regards the long tail of under-achievement, phonics is particularly beneficial for children from poorer homes and children with English as an additional language. Decoding is very important, particularly in the early stages. ‘You cannot comprehend if you cannot decode.’
We should listen to psychologists, as they are doing the best research: they ‘are all singing from the same song sheet. Psychologists are all saying that children need to understand the alphabetic code and they need to be taught phonics’. But teacher-training institutions show little interest in psychological research: Morag Stuart herself works in one, teaching on Masters’ courses for qualified teachers, but has ‘never been invited to give so much as a single lecture on the initial teacher training course which runs in my own institution’. She will be invited in future, however, as the course leader has changed.
[Referring to the ‘searchlights’ model:] The model of reading presented to teachers in the NLS, ‘which is this black hole of four things operating and disappearing into a text, is completely and utterly misleading, and bears no relation to any research on reading that I know of’. [For further details on this, see the paper by Morag Stuart for the March 2003 DfES phonics seminar. This can be found at:
[In response to a question about whether children who seem to learn by osmosis are held back by systematic phonics teaching:] They will not be held back, because the teaching proceeds so quickly – ‘it is not going to damage any child to do 12 weeks of succeeding at something that they can do’. We do not have to teach children every possible phonic correspondence, because most children soon start to infer correspondences for themselves. The children who form the ‘long tail of under-achievement’ probably ‘need continuing support throughout school and not quick bursts of this, that or the other to catch up’ [a reference to the wave 2 and 3 catch-up programmes now in force].
Points made by Debbie Hepplewhite
The NLS has indeed put the teaching of reading back on the agenda, but it does not promote the very best teaching methods: phonics teaching is often done incidentally in the context of story-book reading, and this is less effective than targeting letter-sound teaching in such a way that children become automatic at using this strategy for word-reading. Simple fast-paced phonics is quite compatible with an emphasis on comprehension and reading for pleasure. Many people think that children need to use the context and their comprehension of a passage to identify the words, but decoding the words actually brings a greater understanding of the text. Parents are often the sensible ones, recognising that guessing from pictures or from first letters (as promoted in the Early Literacy Support programme) is not reading.
Teachers should not be told what programmes to use, but should be given simple information about different programmes and the results they produce, so that they can make their own informed choices. OFSTED has flagged up the fact that schools with very similar intakes can produce very different results; ‘teachers need to know therefore what other schools are doing which is working so much better’. At present, teachers are not given evidence about NLS programmes which is based on pilot tests with experimental and control groups. Baroness Ashton asked RRF representatives some time ago for a list of schools teaching reading very successfully – ‘we gave her a list and we never heard from her again, so we do not know whether she investigated them’.
In other countries, children are older when they start learning to read, but the age at which they start is probably less important than using the best possible method when they do start. [A similar point was also made by Rhona Johnston and Ruth Miskin on 7 February, when the topic of the age of starting school was raised again by Select Committee members.]
[In response to a request for information about the Clackmannanshire study:] ‘They were taught through systematic, fast-paced phonics where children were not taught a sight vocabulary first, they were taught a type of phonics which we call all-through-the-word phonics’ [as distinct from the NLS approach, which has more emphasis on teaching a sight vocabulary and focuses in a more discrete way on initial sounds, then final sounds, then medial sounds]. Earlier Clackmannanshire results, and also results from St Michael’s, Stoke Gifford, have been mentioned in RRF Newsletters.
8 December 2004
The hearing started with Minister Stephen Twigg being questioned on international comparisons of reading attainment. It was interesting that some tough questions were asked about England’s non-participation in the last PISA study: the Select Committee chairman asked whether anyone would be ‘reprimanded or sacked’, and also asked for reassurance that ‘this is not a fix’. Once the focus switched to the teaching of reading, the main witness was Dr Kevan Collins, Director of the Primary National Strategy.
Points made by Kevan Collins
[In response to questions about the research behind the NLS and the ‘searchlights’ model:] A seminal piece of research drawn on by the NLS team was Marilyn Jager Adams’s review of research [Beginning to read: Thinking and Learning about print, 1990 – but see * on next page]. The searchlights model was devised by three or four people drawing on the work of Rumelhart, Marie Clay, Priestley [Pressley?] and the comprehension theorists. ‘It is a visual representation of the view that good readers attend to an array of information, and the priority of information when children are young is developing the phonic knowledge. But equally children are active learners and they will, and should, use other knowledge that is available’. [The RRF would argue, however, that the searchlights model does not make it clear that phonics is the priority at the beginning. Rather, it suggests that phonics is only ever one part of what Dr Collins calls ‘an array of information’ needed for word-identification. Dr Collins said that he was a ‘great fan’ of Morag Stuart’s, but did not mention the criticisms she had made of the searchlights model in her paper for the 2003 DfES seminar.]
The NLS approach is to teach sight-words as well as phonics, and also to encourage children to use the other searchlights: ‘They bring their knowledge of phonics to get the first consonant. The dominant consonant is the first thing and they get to bits of the word. They use other information – the context, maybe the picture, the evolving story. They use their syntactic knowledge, the kind of grammar and pattern of English, and they use their graphic knowledge.’ [Again, see * on next page, and see pages 14 and 15 of Ruth Miskin’s article in this Newsletter.] The NLS does not believe in limiting texts encountered by children to material which they can decode phonically. [It should be noted that not all synthetic phonics programmes use decodable books, but a factor common to Clackmannanshire’s Fast Phonics First, Jolly Phonics and Read Write Inc. is that they all have a period at the beginning when the children themselves are not expected to do any book-reading at all.]
‘The basic decoding, the cvc is part of reception teaching’, but it is not known what percentage of children can manage this by the end of reception. [Finding these figures out for properly matched systematic synthetic phonics schools and pure NLS schools at the end of the current school year might be a very good place for a comparative study to start.]
[In response to a statement by Mr Nick Gibb, M.P., that seven-year-olds could often be seen guessing words, pretending to read, and floundering without a picture:] ‘I would be appalled if I saw a seven-year-old just guessing words.’ [But children are guessing, and they are doing it because guessing is in effect written into some NLS ‘scripts’ – in the Early Literacy Support video, for example, the teacher Andrew holds the book in such a way at times that only the picture is visible and suggests that the children use the picture to work out what the words are.]
Point made by the Select Committee Chairman
The committee had heard evidence of a ‘kind of ideological purism’ that ‘phonics is the only way... the one faith, the true faith, the only faith’, but the NLS seemed to represent the ‘more pragmatic view... that a child should be given phonics but a range of other entries into learning to read’. [Synthetic phonics advocates can understand how this ‘ideological purism’ attitude arises, but would point out that they teach phonics so fast and intensively that it is only a matter of weeks before most children know enough phonics to read age-appropriate books. Research does not support the view that children should be given a range of other entries into learning to read – see, again, * below. Research supports the view that children’s entry into reading should be through learning to decode phonically – two of the ‘searchlights’ (context and grammatical knowledge) should be used for comprehension after the words have been decoded, and the other (‘word recognition and graphic knowledge’) is in fact best acquired via the decoding route.]
*Many of us would never have thought of Marilyn Jager Adams’s work as having supported the NLS searchlights model type of approach. As it happens, her opposition to teaching children to use several different cueing systems has recently been made clear in an article published by the Canadian ‘Organisation for Quality Education’ in December 2004 – coincidentally this appeared very close in time to the 8 December hearing. This article is entitled ‘Two Solitudes’, and has the subheading ‘The three-cueing system is popular with teachers, but researchers are barely aware of it’. Adams would not, it seems, be happy about claims that the searchlights model (in effect a four-cueing system rather than a three-cueing one) is based on her work.
7 February 2005
The witnesses were Rhona Johnston, Professor of Psychology at the University of Hull and Clackmannanshire researcher, Sue Lloyd, co-author of Jolly Phonics, and Ruth Miskin, author of Best Practice Phonics and Read Write Inc.
Points made by Rhona Johnston
‘In the early stages it [synthetic phonics] suits the children’s developmental level. We do not give children conflicting cues. We do not say, “Guess from text what it [the word] is”. They are told, “When you come across an unfamiliar word, sound and blend it and work out what it is”’.
[Responding to a question about the kind of research needed to show different results being produced by different methods:] That was the type of research done in Clackmannanshire, and it showed synthetic phonics children doing better than control groups, one of which was taught in a way resembling the NLS’s Progression in Phonics.
‘There is no doubt that it [synthetic phonics] must be the first thing that you do.’ In the Clackmannanshire study, the children in the experimental groups were given synthetic phonics teaching in the first 16 weeks, and then the same teaching was given to the control groups in the second 16 weeks. At the end of the second year of school, girls were better readers if they had done synthetic phonics early rather than later, and both boys and girls were better spellers if they had done synthetic phonics early. [The point about starting in the right way was also made explicitly by Ruth Miskin: ‘When you decide to teach a child to read, you have to get it right first time.’]
‘Education is not very evidence-based’ – people are bowled over by charismatic figures such as Frank Smith and Kenneth Goodman, and their approaches are rolled out into schools even though no research has shown them to be effective. ‘There has in fact been a huge amount of research on phonics teaching ... The research has actually been there a long time, but the research was not telling people what they wanted to see and it was ignored really by educationists.’ Holland, Austria, Germany and Spain all use synthetic phonics, but even in these countries whole-language methods have had some influence.
[In response to a question from the Chairman about whether a piece of research was needed which would compare NLS-taught children with a closely-matched group of synthetic phonics-taught children:] ‘Yes, I think that absolutely needs to be done to establish what the facts are. I should stress that my research has been paid for entirely by the Scottish government... We were invited by Clackmannanshire to do the study, but that money was only given out if they did pre-tests and post-tests of an experimental and a control group using standardised tests. This is what I think should happen in England.’ [This was the last witness comment made in this session, and it is very much to be hoped that the suggestion will be followed up.]
Points made by Sue Lloyd
A big rise in reading standards accompanied her school’s move from a look-and-say approach to a synthetic phonics approach in the 1970s – the teachers were the same and the children were from the same areas, but the school’s average quotient on county standardised reading tests went from 102 to 110+. The local education authority was told, but showed no interest. Synthetic phonics starts by teaching children a few letter-sounds and teaching them to blend those sounds into words. The NLS gets the children to read books very early on and encourages children to use strategies that synthetic phonics teachers don’t approve of, such as, ‘Look at the picture to help you read the word’ and ‘Try to guess what it is, try to predict what it might be’.
[In response to questions about how the NLS could have raised standards:] Some phonics is better than no phonics, and the NLS has introduced some phonics, which has led to improvements. But when the results produced by the pilot National Literacy Project were published by the National Foundation for Educational Research (NFER), they showed that schools using this approach had not even brought the children up to an average standard. Getting children to read books before they are ready is putting the cart before the horse. It is better to read aloud to them and to talk about the stories. [See also Ruth Miskin’s article in this Newsletter.]
[In response to an MP’s suggestion that we all use prediction when reading:] Prediction for comprehension purposes is acceptable, but prediction for word-identification purposes is not. [An attempt to convey this point was made in the NLS paper ‘Teaching Phonics in the National Literacy Strategy’ for the March 2003 DfES seminar, as Victoria Robinson points out on page 21 of this Newsletter, but the misunderstanding underlying the MP’s question is still common.]
An example of the success of synthetic phonics can be seen in the 2004 Key Stage 2 results of a large primary school in a relatively poor area, where 94% of children achieved Level 4 or above in English (national figure 77%) and 65% achieved Level 5 (national figure 26%). Boys and girls did equally well, and no children with English as an additional language were on the Special Educational Needs register.
Schools may use published synthetic phonics programmes, but they often feel obliged also to follow the NLS ‘because all their literacy consultants are telling them that the National Literacy Strategy is the one that is most important’. ‘The DfES has taken the high ground, and said, “Well, we know better”. Yet there is no evidence that their programmes are better.’ The NFER report on the National Literacy Project showed children not even reaching the national average quotient of 100. The Early Literacy Support programme, aimed at children with difficulties, has not been tested at all and is not supported by research. All this is very different from the scientific work carried out by people like Rhona Johnston.
[In response to the Chairman’s question, ‘Are you saying that of all the departments of education which do so much research, no one has properly evaluated the value of phonics against other methods of teaching children to read?’:] Yes. Programmes have been put out into the whole of the country without being properly tested.
We don’t really know whether a pure NLS approach is producing good results even in leafy suburbs. Kevan Collins claimed that NLS schools in Tower Hamlets [not a leafy suburb] were getting better results than synthetic phonics schools, but in fact these Tower Hamlets schools turned out to be using synthetic phonics programmes. [The point is that as long as schools are using synthetic phonics programmes at all, even unsystematically, it is impossible to tell exactly how much children’s performance is the result of NLS teaching and how much is the result of synthetic phonics teaching.]
Points made by Ruth Miskin
Children need to be able to decode effortlessly. ‘If you have to work very hard at every single word that you come across, asking yourself, “Shall I use a picture cue? Shall I use a context cue? Shall I use a picture cue with a letter cue?...”, the child cannot make that decision while they are reading.’ The NLS represents a compromise, with attempts to placate lobbies representing different sorts of phonics [analytic, synthetic, onset-rime], ‘real books’, etc. A problem is that some children may be still unable to read but may be spending 100 minutes a week trying to follow discussions about plot, characterisation, settings, author’s craft etc.
Phonics doesn’t have to be deadly dull. It can be fun, and the children enjoy being successful. Phonics is much more complicated in English than in other languages, but synthetic phonics starts with the transparent part of the alphabetic code and only then introduces the more complex part. The ‘kn’ spelling for the /n/ sound in the word ‘know’ is one of the complexities, but it is still phonics. Alongside this cumulative teaching of the code, however, teachers are immersing children in literature, but by reading aloud to them rather than by expecting them to read books to themselves. Parents often do things with their children that do not fit in fully with a synthetic phonics approach, but what parents do should never be denigrated. ‘Parents sometimes say, however, “Will you show me what you do at school, so that when we do help we are talking with the same voice”, and then I show them.’
[In answer to a question about whether Kevan Collins had been right, in the seminar of 26 January, to say that a lot of Tower Hamlets schools did well in teaching reading using the NLS scheme:] Enquiries have shown that every school in Tower Hamlets has adopted a synthetic phonics approach – either Ruth Miskin’s or Sue Lloyd’s. Schools say that they have not followed all the NLS programmes which have been brought out, ‘so Kevan was wrong’.
In deprived areas, and particularly if parents don’t speak much English, no one may hear a child reading at home. If the child simply recites a book that he has memorised (e.g. one child boasted, ‘I can do it with my eyes shut, I’m that good’), parents may not realise that this is what is happening. Inner-city areas are often still the ones where teachers believe in getting children to guess from pictures and context – the teachers say, ‘We’ve been taught on a multiple-cueing system’, meaning picture-cue, context, grammar... Publishers are actually producing books which encourage this.
Representatives of the National Literacy Trust and the Early Childhood Forum also gave evidence on 7 February. Points made by them will be only very briefly summarised, without individual witnesses being identified.
There was general support for phonics, but within limits. There was also support, in effect, for ‘emergent literacy’: e.g. it was suggested that babies looking at symbols on cots and children looking at the print on cereal packets were ‘reading’. Children don’t suddenly become ‘literate’ when they go to school. All witnesses supported the idea of good nursery provision.
Reservations were expressed about pushing the NLS down into the Foundation Stage. One of the less good effects of the NLS has been a reduction in the amount that teachers read aloud to children. [Synthetic phonics advocates would probably agree: they believe that reading aloud to children is a better way to inculcate an interest in books than is the practice of getting children to try and read to themselves before they are ready.] The Early Childhood Forum had evidently said, in a written submission, that phonics should not be introduced in the early years – another representative disagreed with this, but said that what she was against was ‘these rather formal “b-a-t” things, which send shivers down the spines of most of us’. All children learn differently, so we need a variety of approaches; we also need a cradle-to-grave, inter-generational approach to literacy.
Kathy Sylva’s work has shown that nursery education gives children a very good start, but ‘we need to go on and see how they are reading at 12 and 14. My gut feeling, and that is not good enough of course,... is that the children who have that kind of experience are the ones who love to learn’. [The RRF would agree that gut feeling is not good enough. Media coverage of the Clackmannanshire study suggested that children ‘love to learn’ synthetic phonics.]
Towards the end, some very searching questions were asked about whether children who could decode mechanically were more or less motivated to read than children who could not decode mechanically. The witness thought that children decoding mechanically would be less motivated, but knew of no scientific evidence supporting this. It also became clear that two of the witnesses had not read any of the Clackmannanshire research, and the other two had read some of it but only some time ago. [Note that it would have been impossible for anyone to have read the most recent report, as this did not come out until four days after this Select Committee hearing. Nevertheless earlier Clackmannanshire findings were very significant, and had been available on the DfES website since August 2003, when they were published as one of the papers presented at the March 2003 DfES seminar.]
There was also a Select Committee seminar on 26 January, but as this was private it is not possible to report on it.