Half of the country's adults remain poor readers, but Diane McGuinness argues that the governments latest literary strategy is in danger of ignoring some vital lessons.

After six years of research on reading, after running a reading clinic and trying out new discoveries in the classroom, my students and I were convinced we were close to the answers to teaching all children to read. My former student Carmen McGuinness began refining our discoveries with the goal of developing the ultimate method to teach all children to read. When the first 87 poor readers were taught using her new method, nearly all of them learned to read with only 12 hours of individual instruction. The worst readers made the greatest gains. We thought local teachers would be interested. They were not.

In response to this indifference I decided to write a book. If parents and teachers could understand how to teach reading in a simple, logical way, they would be in control of their child's destiny. Since the book appeared in America in 1997, not only have thousands of teachers jumped on board, so have local education authorities. More and more educators in the US have become weary of the reading wars between advocates of "phonics'' and "real books'' methods, and want something better. What could be better than close to 100 per cent success?

The latest reading wars began when the first independent national literacy survey revealed that California, the bastion of the real books method of teaching (in which children are encouraged to learn to read "naturally" by repeated exposure to stories but with little instruction about the correspondences between letters and sounds) was bottom in the nation with a 60 per cent functional illiteracy rate.

³ÉÈËVRÊÓƵ

The California children did badly despite the fact that a study revealed that their teachers did not really believe in the real books method. Out of ten major tenets of the method, teachers agreed with only two. For instance, they did not believe that a writing system is just an extension of "natural language'' and that children can teach themselves to read "naturally''. They did not believe that direct instruction in letter-sound correspondences was "bad'' for children and would turn them off reading. The problem was that the teachers did not know what to do about it. They did not have the appropriate training to follow their beliefs.

When I came to England to do further research I expected that sanity would reign. After all, the Brits do get "things literary'' right. I was all the more a believer because my children were brought up in England and had had a wonderful education.

³ÉÈËVRÊÓƵ

No sooner had I stepped on to British soil than I began to hear disparate accounts of children's reading success. Parent after parent began recounting tales of woe. One parent lived at the "posh'' end of a town largely overtaken by housing estates and council houses. She sent her son to the local school. When he failed to make much progress in reading, she went to see his teacher. She was told that "these children could not be expected to learn very much". Her job was to keep them amused on simple tasks for the allotted time. Instead, the mother was expected to do extra work at home to educate her child. By age nine, her child was a virtual non-reader. She thought that he must be dyslexic. Fortunately, she found a good tutor who taught her child to read and she enrolled him in a private school.

Talking to school officials, curriculum specialists and university lecturers in education, I heard a different story. This story confirmed what I originally believed. I was told that everybody learns to read in England, unless they are "dyslexic'', but that the educators have no real solutions for these "dyslexic'' children, though very few children are this bad. Furthermore, educators in the United Kingdom are apparently much wiser than the Americans about teaching reading. Americans, I was told, have historically gone overboard for new methods, whereas British teachers critically appraise each new idea, and add it to the teacher's bag of tricks. This has resulted in the British classroom being truly "eclectic'' and able to tailor a method to suit each child's particular learning style. I busily took notes. Meanwhile, I noticed that on bookshelves, scarcely a book was to be seen that did not advocate the "real books'' method of learning to read. When I asked more penetrating questions about how teachers knew which method fitted which child for what reason, I got no straight answers.

Then I read the groundbreaking studies published by the National Foundation for Educational Research on the fate of the trainee teacher in Britain both at university and in the first year in the classroom. It did not inspire confidence in the educational system. In fact, things looked even worse than they were in America. According to the NFER results, teachers on university and college-run teacher training courses learned little about how to teach reading, and what they did learn could not be applied. Trainee teachers and first-year classroom teachers consistently reported that they had no confidence in how to teach reading.

I came home confused. What was really going on? Studies using standardised tests showed British children were in trouble. When they were tested by classroom teachers, their results were better than on the standardised tests. But the teachers' tests were highly subjective. Government (Ofsted) inspectors did "estimates'' based on having children read to them. According to the inspectors' estimates, the results were not too bad, but on closer inspection it seems they heard only the children the classroom teacher picked out.

Then came the Organisation for Economic Cooperation and Development's 1997 international adult literacy study. This was the first rigorous study of adult reading skills. Not only were the UK and Ireland not better than other English-speaking countries, they were not even as good. The UK and Ireland had the highest proportion of poor readers and the lowest proportion of proficient teachers, compared with the United States, Canada, Australia and New Zealand. Fifty-two per cent of adults in the UK are poor readers or non-readers, versus 42 per cent in Canada. Sweden has set the benchmark for what can be achieved, with a 4 per cent functional illiteracy rate for 16-25-year-olds. The figure is 18 per cent in the UK for the same age group.

³ÉÈËVRÊÓƵ

The National Year of Reading is one of the government's answers to this astonishing news. The National Literacy Strategy has been launched at a cost of Pounds 58 million, with a suitcase full of curriculum materials, audio and video-tapes. This is certainly a commendable effort, but it is a pity the Government, its advisers and the Department for Education and Employment did not take more time to review the research literature and run a few pilot studies first. The NLS is "eclecticism with a vengeance'', a disastrous marriage of phonics, onset-rime, look-and-say and real books. Not even the phonics component, which is the most in-depth, is accurate enough to teach all children to read.

All writing systems are designed to represent sound units below the level of the word. Languages with a simple syllable system (Chinese), or repetitive consonant-vowel units (India, Southeast Asia, Japan) are written at a sound unit that is easy to hear. When the syllable structure of the language is extremely complex, as in English with its 15 basic syllable types, only an alphabet will work. Alphabets are written for the smallest sound unit of all, individual consonants and vowels, known collectively as "phonemes". (In English there are 43 different phonemes.) Alphabets make extreme demands on auditory analysis. Most people are not aware of phonemes in words. Recently, I asked an audience of adults to tell me how many sounds they heard in the word "stretch.'' (The right answer is five). Although the majority were correct, the rest were evenly split between guesses ranging from three to six "sounds''. If adults have difficulty hearing phonemes in words, then it is obvious that children need to be taught phonemes directly so they can use an alphabetic writing system. Otherwise they will not know the phonemes are there.

What is not understood by the developers of the NLS is that nothing that contradicts the logic of an alphabet should be taught.

³ÉÈËVRÊÓƵ

Alphabets are letter codes that represent phonemes. The letters do not stand for whole words ("sight word'' or look-and-say), with or without context (real books). Teaching phonics, "word families'' or "onset-rime'' (the 'at' in "the cat sat on the mat") is a sure-fire way to confuse a child about which level of sound the alphabet code is written for.

Teaching word parts, analogies and word families creates the "part-word assembler''. This is the child who searches for little, familiar word parts and assembles them into a nonsense word, hoping it will be close enough to guess what it is. Jane sees the word "watermelon", which has these parts in its: wa wat wate at ate ter term erm me mel el lo on. Which of these 13 parts would you use? Jane chose "weatermeon".

The NLS advocates lots of onset-rime exercises, yet the research is abundantly clear that teaching an onset-rime strategy has no impact on a child's ability to read or spell. By contrast, the ability to segment and blend phonemes is critically important.

Britain and America have had decades during which education professors, who train the teachers, have failed. Now an attempt is being made to do something about this failure. This time educationists must get it right.

³ÉÈËVRÊÓƵ

Diane McGuinness is professor of psychology at the University of South Florida. Her book: Why Children Can't Read is published on June 25 by Penguin, Pounds 8.99.