RRF Conference 2006: D. McGuinness 'Fourth Time at the Ball'

Transcripts and Reports of Talks given at past RRF Conferences.
Susan Godsland

RRF Conference 2006: D. McGuinness 'Fourth Time at the Ball'

Post by Susan Godsland »


Diane McGuinness based much of this talk on her 2004 book Early Reading Instruction. She started by talking about how writing-systems work, quoting from a 900-page book edited by Daniels and Bright and published in 1996, with contributions from leading experts. In the introduction, Peter Daniels defines writing as follows: ‘Writing is a system of more or less permanent marks used to represent an utterance in such a way that it can be recovered more or less exactly without the intervention of the utterer.’ Diane pointed out that this means that ‘a writing system is not something that requires the reader to GUESS the meaning of the text, or to ask the creator of this text to explain it.’ Also, ‘writing is bound up with the phonology of a language – not with whole words…whole word pictographs cannot represent abstract ideas, verbs, grammatical inflection, people’s names, and so forth’. Moreover, the human mind cannot remember more than 2000-2500 abstract symbols. Even Chinese writing requires no more than this, as 99% of words can be written by combining 2400 syllable symbols with classifier signs.

Ancient Sumerian writing ‘contained 800 symbols standing for the syllables of the Sumerian language and 60-70 number signs’. It did not develop gradually from picture writing but was invented as a whole. Sumerian clay tablets found by archaeologists also provide overwhelming evidence that the writing system was explicitly taught to children – they were not left to absorb it ‘by osmosis or casual inspection’. The need for direct teaching of a written code was recognised by scholars in all advanced civilizations of the ancient world, but it has been lost sight of by many of those responsible for children’s education in the modern world.

The phonological unit used in for a writing system depends on the structure of the spoken form of the language. When a language has a small number of syllables (as with Chinese and Sumerian), written symbols represent syllables. Most of the world’s languages are made up largely of CVCVCV speech patterns (as in the word ‘potato’), and CV Units (‘diphones’) form the basis of the writing systems. The ‘GOLDEN RULE’ for any writing system is, ‘Choose that largest unit of sound (easiest to hear in the speech stream) that fits the language, but doesn’t overload memory’.

Diane McGuinness discussed the difference between European languages written with a ‘transparent alphabetic code’ and those, like English, with an ‘opaque’ code. In transparent codes, ‘each phoneme in the language is coded by one and only one symbol’ so simple to learn that that children master the written code in 10-12 weeks and quickly move on to reading proper text. Nearly all children learning this type of writing system learn to decode to a high level of skill. This means that poor decoding (the definition of “dyslexia” in the UK and US) cannot be attributed to ‘some type of “brain disorder” or faulty gene’. Instead, the cause is the nature of the English alphabet code and how it is taught.

In view of all the above, ‘modern research ought to show that methods which come closest to teaching an alphabet code properly should produce greater gains than methods which do not’. One remarkable study conducted in 1985 by Carr and Evans in Canada showed this by recording ‘time on task’ for each individual child on 50 occasions per child over several months. They then correlated ‘time on task’ with each child’s reading-test score. They found that ‘ONLY three activities were positively and significantly correlated to reading skill: that is, the more time spent on these activities the higher the reading scores were. These are: practice segmenting and blending sounds in words (phonemes), specific phonics activities such as learning letter-sound relationships, and writing words, phrases and sentences, by copying or from memory’. The memorizing of sight words, lessons on vocabulary and grammar and listening to the teacher read showed strong negative correlations to reading scores – in other words, the more time children spent on these activities, the poorer their reading test scores were.

Diane then talked about her ‘prototype’ for teaching reading, which is based on ‘what we know about writing systems, what we know about how our code works, and what the classroom studies have revealed’. The purpose of a prototype is to get beyond confusing programme details. The prototype contains the following elements: teach the specific sound units that are the basis for the code (and no other units), teach ‘the arbitrary, abstract symbols that represent these sounds’, teach that the code is ‘reversible’, connecting encoding (spelling) and decoding (reading), avoid teaching sight words (‘except the few high-frequency words with rare spellings’), teach segmenting and blending in real words using letters, expect spelling to be accurate ‘or, at a minimum, phonetically accurate’, first teach the symbols for the 40+ phonemes of English, and then teach ‘the advanced spelling code (the remaining 136 common spellings).’ Read ‘stories or texts that match what is taught so far’.

‘WE KNOW WHAT WORKS. So why is it so hard to fix something that is so obviously not working when we have the solution?’ Because the English alphabetic code is so complex, it is tricky to teach UNLESS you are working with a programme that makes this manageable’. There are many good programmes (some of the authors were present at the RRF conference!), but we need university departments of education to give their students the right theoretical knowledge so that they can evaluate programmes properly. More than 100 years ago, people such as Isaac Pitman, A.J. Ellis and Nellie Dale had a good grasp of what should be done, but sight-word reading schemes took over. After the publication of Rudolph Flesch’s book, Why Johnny Can’t Read, in 1955, phonics advocates pushed for change. In the 1960s, large-scale studies were carried out by Jeanne Chall in the USA and by Daniels and Diack in the UK. Bond and Dykstra conducted a huge study, but their analysis was faulty – when Diane herself re-analysed their data, she found that ‘there was one clear and unassailable winner – a programme with an uncanny fit to the Prototype’. This was the Lippincott programme, which ‘closely resembles many of the good UK synthetic phonics programmes’.

‘Today, nearly at the end of 2006, we have another golden opportunity. Will we muff it again? I submit that we need to be highly proactive and demand much more accountability than in the past… With the new government’s latest incentive and the publication of the Rose report, the UK is in the position to put things right at last.’ If we do manage this, the whole English-speaking world will have to follow. We need to establish a Prototype ‘specific enough for there to be no confusion about what should and should not be taught, but not so specific that it doesn’t allow for programme variations’. We need to ‘start a campaign NOW to end any kind of whole word teaching in which children are asked to memorize whole words by sight and guess whole words by context clues’. We need to ensure better initial training for teachers, and to adopt good testing procedures. The government should fund proper studies to compare “phonics” programmes fitting the prototype to “phonics” programmes that do not. Diane sees no purpose, however, in comparing phonics to the failed whole-word methods. She ended by praising the rigorous longitudinal study conducted by Rhona Johnston and Joyce Watson as an exemplar of this type of research.