This important paper, from Professors Anne Castles (Macquarie), Kathleen Rastle (Royal Holloway) and Kate Nation (Oxford) seeks to end the verbal war that has raged over the teaching of reading in English for over two centuries. It covers all stages of learning to read, from the beginning to the attainment of adult fluency, and it would not be possible to discuss the whole of it without writing something of comparable length and depth. This note will concentrate on the early stages of reading, and what immediately follows, and I recommend careful study of the remainder.
I stumbled into the reading wars in 1981, with a critical review of Frank Smith's "Reading" in the Times Educational Supplement. It provoked a furious letter denouncing me for failing to see the obvious truth of Smith's dictum "We learn to read by reading." The exchange was widely noticed. It led to over a decade of hostility with most of the primary inspectors in the local authority I worked for, and with most - but not all - of those responsible for teacher training in the UK. Smith had an ally in the Arizona academic Kenneth Goodman, who described reading as "a psycholinguistic guessing game", in which we did not pay attention to individual letters and words, but constructed meaning as we went along by lightly sampling the print to confirm or reject our guesses. Smith had an entertaining journalistic style, and his brilliant conference presentations had people in stitches and led to thunderous applause. One had to concentrate hard to remember the absence of evidence for nearly everything he said.
The tide began to turn in the late 1980s with the research of Stanovitch, Schatz and Baldwin, and Roger Beard, who demonstrated the fundamental errors in Smith's and Goodman's theories, and work in schools, including a survey by Sue Lloyd that associated successful teaching of reading with phonics in Norfolk, and St Clare's, Birmingham, which won an award for successful reading teaching based on phonics. The later stages of the revival of phonics are well-known and further documented in this paper, which accepts the validity of the Clackmannanshire study in Scotland that became the basis of the UK government policy of making phonics - blending and word-building - the basis of early reading teaching in schools. This is one of the review's most important features, as it implies rejection of the criticisms of this study's methodology by a series of academics, which I've discussed here.
The Clackmannanshire study was limited to a relatively short, but crucial period of education, the first sixteen weeks of primary school. The blending and word-building, referred to as "synthetic" phonics, took just 20 minutes of each school day. The key research introduced by this paper in support of the approach is a series of studies of pre-school children by Byrne and colleagues that showed that simply exposing children to words did not enable them to grasp the alphabetic principle - "that symbols represent sounds". This had to be taught, and once it was understood, later learning was "robust". This explains the most remarkable result of the Clackmannanshire research, which was that the work of these sixteen weeks still had a significant effect on children's word identification skills when followed up seven years later.
These studies show that phonics plays a key role in enabling children to adjust their thinking as they add to their knowledge of spoken language the skills of communicating through symbols, confirming the analysis of the transition first advanced by L S Vygotsky in Chapter 6 of Thought and Language. It does not suggest that phonics are the whole story. Advanced devices for tracking eye movements have shown that fluent readers track text, and every letter in it, very closely, but fluent readers do not continue to sound out words one letter at a time. A typical adult will read comfortably at more than 200 words per minute, and it is simply not possible to sound out three or more words per second.
The authors' scientific method solves an additional controversy, in citing studies comparing a purely phonic approach, with one in which some words of high frequency, but unusual correspondences between sounds and letters or groups or letters are taught as "sight words", which children are expected to learn without blending. These studies found no advantage to either approach.
However, there is more to this than phonics, or indeed, than science. English spelling was changed radically by a huge influx of French words after the Norman conquest, and investigation of computerised dictionaries over the past twenty-five years has shown that around a third of English words are either identical to the French words, or very nearly so. In addition, William and his scribes were averse to any Anglo-Saxon influences, and developed the letter group "th" to represent the sound previously represented by single Anglo-Saxon letters, most often "thorn" (þ). This information can be presented in plain, child-friendly English - there is no need for Greek terminology - and gives children an accurate idea of how English spelling works. A key point is that, as we have so many more sounds than letters, letters sometimes need to work together to represent the full range of sounds. The idea of working together and helping each other is more easily understood by children than "grapheme-phoneme correspondence", and it is worth pointing out that the use of such terminology in the National Curriculum is not statutory. A clear and comprehensive description of the development of English spelling is contained in David Crystal's Spell It Out. My scheme, Slimmed Down Spelling, demonstrates a simple way of using the idea in practice.
The paper makes an important contribution to the study of reading beyond the initial stages. The Clackmannanshire study clearly does not cover this, and it is unfortunate that its results have been extrapolated in ways that give some people the idea that synthetic phonics are not the basis of learning to read, but the whole story. The authors say that, "to become confident, successful readers, children need to learn to recognize words and compute their meanings rapidly without having to engage in translation back to sounds," a point which raises two problems. First, how do they move from sounding out each word in order to begin to do this and, second, do we ever move completely to reading without reference to the sounds of words?
An important contribution to understanding the first part of this process is the late Katharine Perera's Phd, The Development of Prosodic Features in Children's Oral Reading, Manchester 1989. Using an ingenious three-line stave to track the grouping of words and intonation, Professor Perera identified phrasing, rather than separate reading of words (sometimes called "barking at print") as emerging once children could read 60 or more words per minute accurately, which in her sample normally occurred around the age of six and a half. This link between fast, accurate word identification and fluency is one of the most valuable discoveries about reading in recent years, and the methodology of this thesis should be developed further. It ties in, I believe, with Vygotsky's identification, at around the same age, of "inner speech", the language in our minds whether or not we are communicating with others. Inner speech is maximally compact, containing all of the shortcuts we may wish to take, and it is almost certainly linked to the development of silent reading.
Whether, and how far, this involves identifying words independently of sound, or simply reproducing the sound in our minds, is an issue that may be impossible to resolve with the means currently at our disposal. One hint comes from the study of English as a foreign language, which identifies stress patterns in words and sentences - one principal stress in each sentence, and one on a syllable in each polysyllabic word (photograph, photography, photographic). These stress features are crucial to the construction of meaning in English sentences, and it is hard to imagine how they could be read without reference to them, and hence to sound, whether physically or mentally produced.
There is a great deal of further work to do, not least in addressing the differences in pre-school language experience, including experience of books, that are discussed in Hart and Risley's Meaningful Differences and The Social World of ChildrenLearning to talk, in Wells The Meaning Makers - where the least successful pupil at 11 had had no recorded experience of books or stories before starting school, and the most successful had had 5000 - and Tizard and Hughes Young Children Learning, where parents who had left school without qualifications tended to have a negative view of children's questions.
It is also highly doubtful that the paper will achieve its goal of ending the reading wars. One policy pursued by the primary inspectors in the local authority I worked for, with the explicit aim of reducing the advantages of children of more highly educated parents - referred to in shorthand as "middle-class" - was to base work as far as possible on first-hand experience rather than on books. This approach seriously limited the development of literacy in primary schools, but this was not an issue for inspectors whose goal was primarily the political one of negating the influence of what the Marxist Pierre Bourdieu called "linguistic capital". 1This remains true of many participants in the reading wars. We are dealing , not with the disputes over territory, resources and power that characterise most shooting wars, but with something close to a religious schism. This paper clarifies the issues involved in literacy, and should be an important stimulus to further work. Neither it, nor anything else, will persuade those on the other side of the argument to convert.
Note.
1. Bourdieu and Passeron, La Réproduction. Paris,1970