There is mounting evidence that the "whole language" method of teaching reading has been little short of a disaster, critics say, 20 years after its widespread adoption in United States schools.
New research released at the Seattle meeting of the American Association for the Advancement of Science (AAAS), from a trial involving 375 elementary school children from poor homes, shows dramatic disparities between whole language and the older phonics teaching it replaced.
Whole language or Real Books tuition developed from the notion that traditional phonics, with children taught to break the alphabetic code of consonants and vowels and translate written words into speech, was laborious and dull.
Its stated aim was to interest children in books first and leave accurate grammar, spelling and pronunciation to develop in the context of love for the written word. But both in Britain, and increasingly in the US, a debate is raging over whether it simply skirted the learning process.
California recently became the first US state to call for a return to phonics, citing poor reading standards. And critics now even blame whole language, the modern orthodoxy at most American schools and teaching colleges, for a rising incidence of dyslexia.
As many as one in five American children has difficulty learning to read, Dr Barbara Foorman, of the University of Houston, told the AAAS. In some urban schools, reading failure is the norm. But her report suggested that particularly for children from poor socio-economic backgrounds - who rely on schools for their education more than the offspring of wealthy, well-educated families - performance is tied to the amount of phonics tuition.
Dr Foorman and her researchers documented a trial in which nearly 400 six and seven-year-olds beginning to read were divided into three groups, to be taught whole language, phonics, and a mixture called "embedded phonics". Measured by their ability to read aloud from a list of 50 words, a third of those in the whole language classes showed almost no improvement over half a year, learning to read fewer than three new words.
For the phonics-based classes, by contrast, the figure was only 6 per cent. In tests, they thoroughly outperformed the others. "Children who were directly instructed in the alphabetic principle improved in word reading skill at a significantly faster rate," she concluded.
There is a grass-roots reaction against whole language under way in America, led by parents who are having to fill the gaps left by teachers, said Professor Mark Seidenberg, who led the AAAS panel on "Learning to Read".
Reports of dyslexia increased in some states by as much as 25 per cent, and "unless we assume that there is something in the water, something is doing that", he said. Concern over declining test scores suggests that the national experiment in whole language teaching is a failure, he says. But its supporters insist that what children learn is not statistically measurable, he said.