Barnardiston Hall Preparatory School in Suffolk accepts pupils with a wide range of learning difficulties. Their head of pre-prep, Ann Marie Cheason, has a very sophisticated understanding of the principles and practices that make synthetic phonics work. She uses Jolly Phonics for whole-class instruction, but inevitably there is a substantial number of children who need a lot more intensive teaching. This is where Sound Foundations comes in. Last school year, all their 7-year-olds achieved Level 2B or above on their Key Stage 1 English SATs. Considering that many of their pupils have very severe dyslexia, this is no mean feat.
Barnardiston first started using Sound Foundations in 2002. At first, we trained five classroom assistants to work with their SEN pupils. This didn’t work well: in a bustling school like Barnardiston, you need to have focused instructors who can ensure that pupils get the short, daily lessons that they need. The following year their SEN department employed two teaching assistants, Iona Hayes and Neil Jeffries, to give a substantial percentage of their time to using Sound Foundations. The results were immediate and dramatic. Although we gave Iona and Neil no training, and they have no qualifications of any sort, they quickly understood the logic of our system. Working closely with Ms Cheason and her department, they have helped to make Barnardiston into a school where there are no excuses for reading failure.
Sound Foundations consists of separate decoding and spelling programmes. We decided against an integrated reading and spelling programme because some SEN pupils make faster progress on decoding, and others find spelling easier. Also, it is easier to teach spelling in groups: only one child can read at a time, but any number can write.
The decoding strand, Dancing Bears, is designed to be used for 10 minutes per day with each pupil. Considering that the average infant school class of 25 pupils will have 6 pupils who need extra help, this amounts to only an extra hour of instruction per day. Most schools waste far more time than this with ineffective ‘shared reading’ schemes. The spelling strand, Apples and Pears, can be used with groups of up to 10 pupils, so long as they are well matched for ability.
Sound Foundations works from the principle that decoding and spelling are lower-order cognitive skills which are most efficiently taught through straightforward behaviourist training. Decoding and spelling are the keys that enable children to become literate, in the highest sense of the word. Whole-language advocates have done immeasurable harm in English-speaking countries because they fail to understand that higher-order skills have very little to do with mastering the English spelling code.
From this principle, it follows that the most efficient means of teaching decoding and spelling are to induce the pupil to make the maximum number of correct responses in the shortest period of time. Everything in Sound Foundations is oriented to this goal. All materials contain massive amounts of overlearning – far more than in any other remedial literacy programme. All this is presented in a variety of formats, and reinforced at strategic intervals to ensure that virtually all pupils develop automatic, unmediated responses.
Since children with reading problems almost always have problems with one or more short-term memory functions, daily lessons are essential. Otherwise, so much is forgotten from one lesson to the next that very little gets into the long-term memory. As the tasks in learning to read and spell are sequential, SEN pupils need to develop automaticity of response at each level before they can progress. For instance, Sound Foundations uses flashcards to teach basic letter sounds and unambiguous digraphs. These cards are practised at least once a day until responses are instant. Many children who come to the Promethean Trust know their letter sounds, but their response is so slow that they can’t blend – when they are reading ‘cat’, they have forgotten the /c/ by the time they get to the /t/.
And because daily lessons are so important, we have done everything we can to make Sound Foundations simple to use. Our materials can even be used by parents who are themselves only just literate. In schools, almost any classroom assistant can use them effectively, even with no training at all.
The only thing that is counter-intuitive about Sound Foundations is that when children get a word wrong, they are not encouraged to work it out on their own. This is almost invariably counter-productive, because once an incorrect decoding is lodged in a pupil’s head, the process of trying to correct it is trying for both the teacher and the pupil. Our aim is to encourage pupils to decode, rather than to guess – and we do not want them to get the idea that decoding is too difficult. So when children are unable to read a word or to self-correct an error, we simply tell them the word, and the pupil repeats it. The same item is presented again after a very short interval. The instinct to encourage children to work things out for themselves is very strong, but it must be resisted. For a start, it encourages ‘guessing’. It slows the lesson down. And most importantly, it focuses children’s attention on what they can’t do.
Beyond dispute, Sound Foundations is a synthetic-phonics programme, and it is compatible with Jolly Phonics and most other synthetic-phonics whole-class programmes. Children are capable of thriving on a varied diet, so long as the diet doesn’t contain unhealthy ingredients, such as word-guessing. The broad objective of Dancing Bears is, of course, to teach children mastery of the English spelling code so that response to print is instant and automatic. This must operate independently of lexical matching, or the child’s ability to read unfamiliar words will be seriously impaired. The broad objective of Apples and Pears is to teach the pupil to spell a broad range of high-to-medium frequency words without a mediated response, and to teach phonemic and morphological knowledge so that new spellings are more easily remembered. The upper levels of Apples and Pears also teach the use of the apostrophe, and the meanings of many ‘academic’ words – that is, multi-syllable words with Greek and Latin etymology.
Ultimately, the goal of all instruction is to produce unmediated responses. Pupils who have to remember a mnemonic device such as ‘big elephants can always understand small elephants’ (to spell ‘because’) will necessarily be distracted from the content of their writing. Likewise, formal rules – both for spelling and reading – are at best temporary crutches. Clymer’s classic study* should serve as a warning to anyone who attempts to use a rule-bound approach to phonics; rules have a nasty habit of letting you down when you need them the most. Consequently, Sound Foundations teaches relatively few rules for spelling, and almost none for reading. Instead, the primary medium of instruction is modelling.
A consequence of this approach is that easily-confused items are not taught together. Relatively unambiguous digraphs such as /ee/, /ar/ and /sh/ are taught first. With ambiguous ones such as /ea/, the short sound is not taught until the long sound has been thoroughly mastered, and the long /a/ sound is delayed until both of the former are firm. The same principle applies to spelling: the different ways of spelling a sound are kept as far apart as possible. Poor spellers are, almost by definition, deficient in visual memory: the common practice of asking them to make such discriminations by writing the word to see if it ‘looks right’ often fails. Not many poor spellers ever do this unless they are told to, so it is clearly a low-utility strategy.
But perhaps the most revolutionary point about Sound Foundations is that it is a living programme. That is to say, it is under constant revision. There is no teaching programme which is so good that it cannot be improved. In fact, we believe that reading pedagogy is barely out of the horseless carriage stage of development. Dancing Bears is only four years old, and it is already in its fourth edition. The fifth edition will almost certainly be in print by the time this article appears, and we are already thinking about the sixth. Of course, we could not do this if we did not teach children ourselves. All the best synthetic-phonics programmes are written by teachers who share our enthusiasm for putting their experience in print.
*Editor’s note: A good source of information on Clymer’s study is Beginning to Read: Thinking and Learning about Print by Marilyn Jager Adams. The study was originally published in The Reading Teacher in 1963. Clymer analysed 45 phonic generalizations that were commonly taught to primary-school children to see how frequently they applied to words which these children were expected to be able to read, for example in ‘basals’ (reading-scheme books). He found that many of the generalisations were not very useful. In 1999, also in The Reading Teacher, Francine Johnston showed that some of the rules became more useful if they were reworded, and suggested that although English orthography is too complex to be reduced to a few rules, there are some generalisations that are worth teaching.
The teaching of a few rules, rather than many, has probably been characteristic of British phonics programmes, and this may be a difference between them and American programmes. Tom Burkard, however, is surely right in saying that a rule-bound approach to phonics is not ideal, especially for children who are struggling.