Last month Education Secretary Ruth Kelly announced that she had accepted the recommendations of a review by Jim Rose (a former director of Ofsted) which found that children learn to read best through synthetic phonics. Synthetic phonics are therefore to be the only system used to teach the basics of reading, unlike the range of methods employed at present. Henrietta Dombey explains the issues.
‘It’s all perfectly straightforward, isn’t it? It starts with ker a ter for “cat”.’ That’s what I was told by the Chairman of the Select Committee on Education at the House of Commons many years ago, during yet another investigation into the teaching of reading in England’s primary schools. It’s a view that’s shared by many people who have never studied the subject, or even closely watched a young child trying to make sense of a text.
Yet a number of researchers, usually with a background in psychology, often seem to be saying something similar (Adams, 1990). Certainly many research studies have shown positive results for children trained in a phonic approach. Johnston and Watson’s Clackmannanshire study, much publicised last summer, is just the latest in a long line (Johnston and Watson, 2005).
If you are a regular reader of Books for Keeps you probably take a rather different view. To you learning to read starts with a small child sitting on your knee, gazing at the pictures of a book, as you talk your way through the text. Making sense is what it’s about, and shared pleasure in both pictures and words, and the interplay between them. There’s plenty of evidence to show that by and large children read to extensively at home, learn to read with greater ease and effectiveness than those who are not (Adams, 1990).
These approaches should be seen as complementary, not in competition. We now have evidence from a number of studies in the UK and the US that make this clear (e.g. Taylor and Pearson, 2002). The most successful schools and teachers focus both on phonics and on the process of making sense of text. Best practice brings these two key components together, in teaching that gives children a sense of the pleasures reading can bring, supports them in making personal sense of the texts they encounter and also shows them how to lift the words off the page.
So phonics is an essential component of early reading teaching, but what kind of phonics? What is the difference between synthetic and analytic phonics? Which is best?
Firstly it should be remembered that in this context synthetic doesn’t mean artificial. It’s about synthesis – putting together small elements to make a larger whole. Synthetic phonics starts by training children to match phonemes, the smallest element of speech sound, with letters, the smallest elements of written language. Some approaches focus first on the letters, training children to match these to sounds, others first on the phonemes, training children to match these to letters. In both cases the letters and sounds are then put together to form words. Once children can readily ‘sound out’, ‘blend’ or ‘build’ words in this way, they are allowed a (limited) access to books chosen for the decodability of their vocabulary. So starting with ‘ker’ ‘a’ ter’ is a synthetic phonic approach.
Meanwhile analytic phonics starts with whole words, which are then analysed into smaller units. To what degree this is done varies. Proponents of this approach recommend teachers to start by teaching whole words, grouped for similarity of structure, then instruct children systematically to analyse familiar one syllable words into their ‘onsets’ and ‘rimes’. (The onset is any letter or letters in front of the vowel, and the rime the rest of the word, from the vowel onwards.) So ‘cat’ is split into ‘c’ and ‘at’.
In some variants there is little or no direct teaching: children are shown groups of words such as ‘cat’, ‘mat’ and ‘hat’, and expected to infer their relationship. However, the term ‘analytic phonics’ has also been used by its opponents as a conveniently dismissive label for any approach with a phonic element that is not systematically synthetic.
What is the reasoning behind these approaches? And what are their drawbacks?
Synthetic phonics takes matching the letters and phonemes to be the logical first step. Only when children can identify a range of words with rapid accuracy, can their attention be turned to the meaning the words enable them to make. For an important assumption underlies synthetic phonics – that young children’s brains do not allow them to focus, at any one time, on more than one aspect of such a complex process as reading.
But there are drawbacks. Perhaps the greatest problem is that many English spellings simply do not fit with the idea of one to one correspondence between letter and phoneme. It’s not really possible to ‘sound out’ such common words as ‘one’, ‘was’ or ‘all’. The vowels are the real problem: English is a vowel-rich language, with 20 vowels or diphthongs, as against the 12 of Spanish. But we have only the same 5 letters (or 6 if you include ‘y’) to represent them. The letter ‘a’ stands for quite different vowel sounds in the words: ‘cat’, ‘car’, ‘came’, ‘call’ and ‘career’.
Then there are the children. Young children have great difficulty in recognising that words are composed of sequences of separable phonemes. It may seem obvious to you and me, but recognising that the spoken word ‘cat’ is made up of ‘ker’ + ‘a’ + ‘ter’ is not something that comes naturally. Games of ‘I spy’ with four-year-olds can be frustrating. It’s even harder for them to identify the final and medial phonemes of a word than it is to say what the word begins with.
And finally there’s the boredom factor. Focusing on the code alone does not grip most children. So the pill is sugared with puppets and games and praise from the teacher. But, except in the most skilled and enthusiastic hands, synthetic phonics on its own remains tedious for very many children and teachers. And the argument that children cannot concentrate on more than one aspect of a complex process has been called into question by recent studies of children’s brains, and is certainly belied by their powerful capacity for language learning (Coles, 2003).
Analytic phonics operates on the idea that children find it easier to recognise the onset and rime of a one syllable word than they do to spot all its phonemes. It chimes with what we know of how children become aware of phonemes – by moving from whole to part, from whole words to syllables, to onset and rime, before developing an awareness of phonemes. It is also built on a recognition that rimes in written words (the ‘all’ in ‘ball’) provide more reliable information about how the word should be pronounced than do the individual letters. For over thirty years we’ve known that the 37 commonest rimes in English, when combined with a range of onsets, produce nearly 500 words (Wylie and Durell, 1970). The whole to part approach also means that children can be introduced to some whole texts and ‘have a go’ at reading them without waiting until they have mastered a whole range of sound/symbol relationships.
But of course there are drawbacks here too. In the less explicit version of this approach, children may not be clear about which part of the word they are being asked to attend to and thus may make slow progress. Arguably, analytic phonics requires more training for teachers, since few will have experienced it. And children still have to relate the consonant letters to the phonemes they represent in the onset position.
So what’s the answer?
There have not been enough studies to make this as clear-cut as it should be. But there have been rather more in the US than in the UK, where no clear superiority has been shown for either approach when explicit teaching is carried out by skilled and committed teachers who are well resourced and supported (National Reading Panel, 2000).
But why should it be one or the other? Just as the most effective schools and teachers focus both on phonics and meaning-making, couldn’t the best way to teach phonics be a principled combination of both approaches? There is already some evidence that this might be the case (Juel and Minden-Cupp, 2001). Attention to onsets and rimes of appropriate levels of difficulty (‘c/at’ well before ‘str/eet’) can usefully be inserted into a synthetic phonics programme, with the most straightforward patterns underpinning the early stages. The more complex ones come in later on, introducing children to spelling patterns such as ‘hinge’ and ‘caught’ that seem beyond most synthetic programmes, but are nonetheless regular.
Such an approach need not be exhaustive. While it seems to be clear that the vast majority of children need to be taught phonics explicitly, they don’t seem to need to be taught every one of the 166 or so rules (and 45 exceptions) that underlie the spelling patterns of English texts at a nine-year-old’s level (Berdiansky, Cronnell and Koehler, 1966). Once children have acquired a certain amount of knowledge, they can start to teach themselves (Juel, 1994). We do not yet know what is the optimal amount of phonics that should be taught explicitly, and it probably differs between children. But we do know that children with a reliable working basis can puzzle out words that exemplify rules they haven’t yet been taught.
So where does this leave us?
To become effective and enthusiastic readers, our children need both synthetic and analytic phonics, together with teaching in making sense of written language. But they also need highly skilled and committed teachers, with plenty of material and moral support, and also a healthy respect for the children they are teaching and the families and communities they come from. Recent studies have shown that the implementation makes a bigger difference than the sequence and composition of activities in a programme (Taylor et al, 2000).
All this means that we shouldn’t be surprised that the children in Clackmannanshire were not nearly so far ahead on tests of meaning as they were on word identification (Johnston and Watson, 2005). Nor should we be surprised to learn that during the seven-year period of the study, the headteachers received highly acclaimed leadership training, and all primary schools were engaged in a generously funded early intervention project, involving much professional development and support for the teachers (Ellis, 2005). Synthetic phonics on its own didn’t provide the answer, and won’t for schools in England either.
Successful ‘dual focus’ teaching of reading is not simple. Children are complex intentional beings and what matters most of all is that their teachers have the skill to recognise the enormously varied knowledge, interests and abilities they bring to the classroom, and build on these through engaging the children’s interest and intellect in a balanced programme of activities.
Adams, M.J. (1990) Beginning to Read: Thinking and learning about print , Cambridge MA: MIT Press.
Berdiansky, B., Cronnell, B. and Koehler, J. (1966) Spelling-sound relationships and primary form class descriptions for speech comprehension vocabularies of 6 to 9 year olds. Technical Report No. 15. Austin, Texas: South West Regional Laboratory for Educational Research and Development.
Coles, G. (2003) ‘Brain activity, genetics and learning to read’ in N. Hall, J. Larson and J. Marsh (Eds.) Handbook of Early Childhood Literacy . London: Sage
Ellis, S (2005) ‘Edict on phonics under attack in Scotland’. Edinburgh: Times Educational Supplement 2.9.05
Johnston, R. and Watson, J. (2005) The effects of synthetic phonics on reading and spelling attainment. A longitudinal study. Edinburgh: Scottish Executive Education Department. www.scotland.gov.uk/library5/education/sptrs-00.asp
Juel, C. (1994) Learning to Read and Write in one Elementary School . New York: Springer-Verlag.
Juel, C. and Minden-Cupp, C. (2001) Learning to read words: linguistic units and instructional strategies. Reading Research Quarterly 35.4, 458-493
National Reading Panel (2000) Teaching Children to Read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction . Washington DC: National Institute for Child Health and Development.
Taylor, B.M. and Pearson, P.D. (2002) (Eds.) Teaching Reading: Effective schools, accomplished teachers . Mahwah NJ: Lawrence Erlbaum Associates.
Taylor, B.M., Pearson, P.D., Clark, K. and Walpole, S. (2000) Effective schools and accomplished teachers: lessons about primary grade reading instruction in low income schools. Elementary School Journal , Vol. 101, pp. 121-165.
Wylie, R.E. and Durell, D.D. (1970) Elementary English 47 pp. 787-791
Henrietta Dombey is Professor Emeritus of Literacy in Primary Education, University of Brighton.