Phonics Developments in England from 1998 to 2018
The first UK government attempt to give phonics a higher profile in England was in the National Literacy Strategy (NLS), introduced in 1998. Whole-word and whole-language approaches had dominated before that, and their influence did not disappear overnight, so for that reason and others progress from then on was often erratic.
In 1998, it was just a matter of ‘phonics’, as the term ‘synthetic phonics’ was not yet widely used. The NLS did raise the profile of phonics, but there was considerable emphasis on onset and rime, one pre-requisite for which, according to Goswami, was that teachers should foster ‘good sight word knowledge’ (p. 184 in Reading Development and the Teaching of Reading, 1999, eds. Oakhill and Beard). Some look-and-say teaching therefore continued, and there was also great emphasis on the ‘searchlights’ model, which presented phonics as just one of several strategies to be used in reading. One activity in the May 1998 NLS training pack provided a defaced text in which some words or parts of words were missing, and the accompanying audiotape informed teachers that the words could be worked out by ‘a contextual or syntactic strategy in the instances of first letter or letters and guess’. Whole-language supporters were reasonably happy with the implication that what could be done by proficient readers in an abnormal situation was a good guide to what should be normal for young children – systematic phonics supporters were not.
Some training took place in the last term of the 1997-8 school year to enable schools to start implementing the NLS from September 1998. In the intervening period the materials were criticised by the McGuinnesses, who had meetings with NLS officials when they were in London for the publication of the UK edition of Diane McGuinness’s book Why Children Can’t Read and the promotion of Carmen and Geoffrey McGuinness’s Reading Reflex. These meetings led to another NLS publication, Progression in Phonics (1999), which attempted to incorporate the McGuinness approach. In the introduction, a section on ‘What the evidence tells us about the teaching of phonics’ stated that ‘Traditional approaches to phonics instruction i.e. teaching the sounds that match letters and letter combinations is inefficient and often confusing because of the many hundreds of correspondences involved. The most effective phonics instruction teaches children to identify phonemes in spoken language first, then to understand how these are represented by letters and letter combinations (graphemes)’. Accordingly, there was a great deal of emphasis on the analysis of spoken words into phonemes but very little on teaching beginners a few letter-shapes and sounds at a time and teaching them to read unfamiliar printed words made up of those correspondences by saying sounds for the letters from left to right and blending the sounds. This was in spite of the fact that Watson and Johnston had reported the previous year that results were particularly good in a classroom where this strategy was taught earlier rather than later or not at all, and that a training study had then confirmed the effectiveness of this approach: https://pdfs.semanticscholar.org/3f20/92a76bb21dae0a102a0aa01d2b5b01a77925.pdf (see pp. 5-6).
By 2003, it was clear that the NLS was not raising standards as had been hoped. Prof. Greg Brooks (later a co-author of the 2006 Torgerson, Brooks and Hall review) wrote the following:
‘To what extent, and in what ways, does the phonics element of the National Literacy Strategy need modifying? This was the overarching question for a consultative process undertaken by the Standards and Effectiveness Unit (SEU) of the Department for Education and Skills in early 2003. The most publicly visible part of the process was a one-day expert conference in London on 17 March 2003.’: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.200.9018&rep=rep1&type=pdf.
Brooks describes himself as ‘the expert facilitator to the process’. In his report, he recognised that the NLS version of blending was different from the Clackmannanshire version: the NLS version was ‘scaffolded’ (his term) as the teacher pronounced the target word, then the children identified the phonemes and re-blended them, thus knowing the target word in advance rather than working it out from the graphemes, whereas a key feature of Clackmannanshire synthetic phonics was that the children did not know the word in advance but worked it out. This feature of synthetic phonics as implemented in Clackmannanshire was clearly explained at the March 2003 seminar by Rhona Johnston in her paper on the 5-year follow-up (http://dera.ioe.ac.uk/4938/1/nls_phonics0303rjohnston.pdf). Brooks recommended that ‘the advocates of the two positions discuss and analyse this difference of opinion in order to design and mount relevant research’, but nevertheless concluded in his report that the phonics in the NLS was synthetic phonics. Government ministers accepted this conclusion, which meant that scaffolded blending continued to dominate in the next government-produced programme, Playing with Sounds (2004).
Playing with Sounds thus made little attempt to incorporate the Johnston and Watson findings on early unscaffolded sounding and blending, and this later had unfortunate ramifications in connection with the Early Reading Development Pilot (see below). It is in the earliest stages of reading that children most need a strategy for working words out from the letters, as virtually all the words they encounter are unfamiliar in their printed form. The approach therefore has an obvious conceptual coherence – a coherence which Brooks himself recognised, at least from 2005 onwards, as is clear from his chapter in the 2017 book Reading the Evidence: Synthetic Phonics and Literacy Learning (ed. Clark). There he refers to a meeting in late 2005 at which Carole Torgerson presented the findings to be published in the 2006 Torgerson et al. review of which he was a co-author. He writes ‘I was convinced then, and still am, that theory suggests that synthetic phonics is more coherent than analytic phonics as a strategy for young learners working out unfamiliar printed words’ (p. 75). That may suggest that his position had changed between 2003 and 2005: in 2003, he did not regard unscaffolded blending as an essential feature of synthetic phonics, but by 2005 he was apparently thinking of it as central, as in the Johnston et al. studies. He goes on to say, however, that ‘theory can only suggest hypotheses – what is the empirical evidence?’ (p.76). It is true that there is not much evidence meeting stringent research criteria apart from that provided by Johnston et al., but in view of that evidence and the conceptual coherence of the approach, it is surprising that few researchers have investigated the impact of teaching unscaffolded sounding and blending from the start.
Another development in 2004 was that the cross-party parliamentary Education and Skills committee held a series of hearings on the teaching of reading. That happened between November 2004 and February 2005. By February 2005, the Johnston and Watson 7-year longitudinal study had been published: http://dera.ioe.ac.uk/14793/1/0023582.pdf. This had impressed the committee, whose April 2005 report contained the following:
‘In view of the evidence from the Clackmannanshire study, as well as from other schools where synthetic phonics programmes have been introduced, we recommend that the Government should undertake an immediate review of the National Literacy Strategy. This should determine whether the current prescriptions and recommendations are the best available methodology for the teaching of reading in primary schools. We therefore strongly urge the DfES [Department for Education and Skills] to commission a large-scale comparative study comparing the National Literacy Strategy with phonics “fast and first” approaches…’.
As Rhona Johnston has pointed out, the phrase ‘fast and first’ is important as ‘A later introduction of sounding and blending means that other methods are used first, approaches which often undermine the synthetic phonics approach’ (http://rrf.org.uk/2018/06/17/examining- …). This point also needs to be borne in mind in relation to the 2006 C. Torgerson et al. review, as the later introduction of sounding and blending was a feature of one of the only three studies included; this was a study by J. Torgesen at al. (1999). The other two studies were Johnston and Watson (2004 – Experiment 2) and Skailand (1971), but this last was problematic as it was unpublished and the words used in teaching were beyond beginner level. In their meta-analysis, C. Torgerson et al. used J. Torgesen et al.’s data from before there was emphasis on unscaffolded sounding and blending, but Rhona Johnston’s re-analysis shows that entering the data from after this point into the analysis gives a very different picture. It had been made clear not only at the 2003 seminar but also in a 2004 article by Johnston and Watson published in Reading and Writing: An Interdisciplinary Journal 17 (pp. 327-357) and in the 2005 seven-year longitudinal study (see above) that the early teaching of what amounted to unscaffolded sounding and blending was a crucial feature of the Clackmannanshire synthetic phonics approach, but the 2006 Torgerson et al. review seems not to have recognised its importance and has thus left people thinking that teaching this way had less impact than it actually had.
One very positive outcome of the Education and Skills committee’s report was the setting up of a review of the teaching of early reading to be conducted by Jim Rose. Andrew Adonis, then Minister of State for Education in the Labour government, was key to this initiative, and also to the subsequent decision to produce the government programme Letters and Sounds. A less positive outcome was that although people on all sides had welcomed the committee’s recommendation of a large-scale comparative study, in the hope that this would settle things once and for all, the study was never done. Instead, the government’s July 2005 response to the select committee’s report stated that Playing with Sounds (2004) had ‘all the key components of a synthetic phonics programme’ and would be the basis of a pilot study which would test, among other things, the feasibility of increasing the pace of teaching. Even taught faster, however, Playing with Sounds did not stress early unscaffolded sounding and blending as the Clackmannanshire programme had done, and was therefore not a fair test of that type of ‘phonics fast and first’. The pilot was the Early Reading Development Pilot (ERDP), which ran in the 2005-6 school year and fed into the first year of the Communication, Language and Literacy Development (CLLD) programme, implemented in 2006-7. The CLLD materials were seen and found wanting by a synthetic phonics expert before they went into schools, but they seem to have been used in that first year anyway, while at the same time the government had noted the concerns about the phonics being taught and had arranged for a genuine synthetic phonics programme, Letters and Sounds, to be written. This was not ready until May 2007, however, so it could have had little or no impact on the teaching children received in the 2006-7 school year.
The cohorts from the ERDP year (2005-6) and the first CLLD year (2006-7) were then used as experimental groups in a study by Machin et al. which followed the children up to the age of 11 (http://cep.lse.ac.uk/pubs/download/dp1425.pdf). The children were therefore taught with materials whose inadequacy had led the government to require the writing of a genuine synthetic phonics programme. As economists rather than reading specialists, Machin et al. may not have understood this and may have thought that the faster Playing with Sounds teaching which their experimental groups received in 2005-7 was a fair test of synthetic phonics teaching fast and first, as had been implied by the government’s response to the select committee’s report (see above). Unfortunately, this erroneous view has been widely accepted even by people who should know enough about the teaching of reading to understand the issues. For example, Torgerson et al. accept the Machin et al. findings in their 2018 tertiary review: https://www.tandfonline.com/doi/abs/10.1080/02671522.2017.1420816.
The 2006 Rose review acknowledged the ‘uncertainties of research’ (para. 47), but Jim Rose and his team also took into account the conceptual coherence of synthetic phonics (‘it teaches children directly what they need to know’ – para. 47)) and what they had seen in classrooms. On this basis, they recommended synthetic phonics as the way forward. That 2006 review had exposed the weaknesses in the NLS searchlights model, and searchlights strategies were warned against in Letters and Sounds (see Notes of Guidance, p. 12), but in spite of this, the whole-language/look-and-say influence lingered on and continued to mean that many schools taught synthetic phonics only alongside other strategies, including guessing at words using clues from pictures, first letters or context. In 2010, Nick Gibb became Schools Minister in the Conservative-Liberal Democrat coalition government, and he initiated another attempt to strengthen synthetic phonics teaching by making match-funding available to schools so that they could get approved synthetic phonics programmes and training.
In that same year, wheels were set in motion for the introduction of the Phonics Screening Check (PSC) to consolidate the synthetic phonics approach in Reception and Year 1. A pre-trial was held in November-December 2010 with Year 2 children (the closest in age at that time of year to the eventual Year 1 target group), and a pilot was then done with Year 1 children in June 2011 – i.e. at the time of year when future PSCs would be done. See the technical report for details: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/368290/phonics_2011_technical_report.pdf.
In the pilot, only 31.8% of children met the expected standard, which was set by teachers at standard-setting expertly run by the Department for Education (formerly the DfES). The national roll-out of the real thing was in June 2012, and teachers must in the meantime have put more emphasis on sounding and blending for the working out of unfamiliar printed words, as the pass-rate went up to 58%. The results from 2013 to 2017 were 69%, 74%, 77%, 81% and 81%. In 2018, the check was done in the week starting 11 June, and we await news of the results.
It is clear, then, that UK governments of all colours have been trying hard since 1998 to raise the profile of phonics, but there have often been two steps forward and one or more back. The PSC does seem to have been more successful than any previous initiative in increasing the emphasis on early sounding and blending, even if most schools have also continued to put some emphasis on other strategies as is clear from the evaluations of the 2012-2014 PSCs carried out by the National Foundation for Educational Research. Opponents of the PSC expect it to have a negative impact on children’s comprehension and enjoyment of reading, while its supporters expect the reverse. So far, England’s results in the 2016 Progress in International Reading Literacy Study and in Key Stage 1 and 2 tests suggest that the supporters may be more right than the opponents, but it is to be hoped that the picture will become clearer in due course.