John Bald/language and literacy

Ofsted's English research review, and "Whole Phonics". An error.

I am concerned that the view of phonics in this report goes beyond the limits of evidence and endorses "whole phonics". The very existence of the term "grapho-phoneme correspondence" shows that phonics will not tell a child all they need to know in order to read a word.  We teach children that letters represent a sound, and that they blend these to read a word, as in c-a-t. This depends on the letters and sounds being in the same order. When they are not, as in table, the child can't blend. When a child finds a letter indicating a sound .it does not normally indicate, the child has no means of knowing whether it indicates the normal sound, or not, unless they have already learned the word. Eg, was.

The truth is, as Byrne in Castles et al showed, that phonics establishes the alphabetic principle - that we read words from their letters. This principle works in different ways in different languages, eg in French the very large number of silent letters used for grammatical connections.  In English, mostly from French and Saxon influxes, some phonics are derived from other languages, so that phonics tells us most, but not all we need to know. Until this fact is recognised and understood,teachers following official guidance will continue to present phonics in  a way that is both too simple, and too complex. Too simple, because it does not always work. Too complex, because giving lists of possible correspondences and expecting children to memorise and select the appropriate one is inconsistent with what we know of the development of their young minds.

Other aspects of the English review are excellent. This is an error. I've had no reply to attempts to raise the issue directly with Ofsted. The alternative is on this site. 

The Simple View of Reading, and Reading Development.

The simple view of reading says that we read for meaning at the same time as we decode. Can't get much simpler than that, or indeed more obvious. Read Write Inc builds meaning checks into all of its materials from the outset. At a more advanced level, I once gave the Schonnell reeading test to a grammar school pupil, who scored very highly, but did not know what most of the words meant after a reading age of around 8. He was using what I term phonics "second wind", where the phonically regular features of longer words, mostly derived from Latin and Greek, make them easier to work out from the letters than shorter words, such as answer, which retain historical elements in their spelling. 

There is a distinction to be drawn between fully developed reading, and how reading develops, to which the key remains K Perera's U Manchester thesis of 1989, downloadable here. The key issue is the movement from accurate reading of individual words, sometimes known as barking at print, to grouping them into meaningful phrases, which requires accurate reading of around 60wpm - see the thesis for detail.  This movement has already been accomplished by  fluent readers, though with varying degrees of success. It remains one of the key development points in the teaching of reading, and it is a great pity that Professor Perera was promoted to Pro-Vice-Chancellor, and ceased to conduct research.

So, if the simple view is that we need to pay attention to meaning as well as phonics, no-one can really argue, and it may be that this is not obvious to those who have never considered the matter. Its value in reading development remains limited, unless the additional factors set out here, and in the Reading Rope, are also part of th process. 

Susan Gets Stuck: How Picture Cues Can Confuse Children Learning to Read.

This is the text of an article published by The Times Educational Supplement, Feb 14 1992). It remains relevant, and I have added some notes to clarify the context of the time.

Susan, who is nine and finds reading very difficult, chooses to read The Magic Key, from the Oxford Reading Tree, Stage 5. She looks at the picture on the first page, and begins to read the text, which is “The box was by Chip’s bed. Something was glowing inside it.”

She gets stuck on “something”, but scans the illustration, which shows a box with a bright light coming from it. Susan reads “shining” instead of “something”, then cannot make sense of the rest of the sentence as this has led her into a semantic blind alley. Asked how she knew that the word was “shining”, she replied confidently that the picture tells her.

This is only partly true. Susan had used light sampling of the phonological information in word – the error has the same initial letter and the same letter as the word in the text – and her reading makes sense in the context of what she has just read, even though it prevents her from making sense of what comes next.

Stopping at the full stop might have helped, but the use of punctuation in reading receives only a passing mention in the non-statutory guidance for the national curriculum (1989 version), while picture cues appear as the first point of reference at Level 2, both as a source of meaning and as an aid to deciphering new words. This formulation is based on the well-established idea that young children use all available strategies to make sense of text, and that concentration on one approach limits their effectiveness as readers.

It does not, however, give any guidance as to the relative contribution of these strategies to the development of independent reading. This lack of differentiation, combined with the cramming of so many of the word-recognition and decoding aspects of reading into Level 2, has blunted the impact of the national curriculum on the quality of children’s reading. More specifically, the prominence given to picture luces, re-inforced by the use of heavily-illustrated books in the assessment procedures, could easily lead many more children like Susan into problems which are largely the result of following their teacher’s advice.

Part of the problem is that the illustrations have so much to offer. They attract children to books, help sustain their interest, and offer an additional channel of communication that enables the author particularly in the early stages of reading, to conevy greater detail than is possible through words alone. But the disadvantage to this use of illustration to extend meaning is that it makes them much less reliable as cues to individual words in a text.

Put simply, if the picture contains more than the text, how is the child to know just what it is in the picture that provides the connection to the word he or she is trying to read? To the fluent reader, to whom picture and text are equally and instantly comprehensible, the connection is obvious, and it seems equally obvious that the child should be encouraged to use it. However, there has been very little research into the practical benefit of encouraging children to think in this way rather than concentrating on text-based strategies. What research there is suggests that clues derived from pictures do not help children to remember words outside that specific context.

Susan’s error is not uncommon. Weak readers often allow information from a picture to override the print when they do not recognise words on sight, even when it means departing from the text completely. (Note, 2019. Morag MacMartin’s PhD thesis, Factors Affecting Reading Comprehension in Primary Pupils, OU 1992, confirms this point and explains it in detail). Where text and picture are designed to convey the same meaning, this is less likely, but in most real contexts, where the connections are more sophisticated, the strategy of prediction becomes a two-edged weapon for a would-be reader.

The solution is not to deny the role of pictures in the process of learning to read, but to clarify it and set it in the context of other factors that have been clearly identified by research. These included the effectiveness of phonological teaching, to tune children into patterns in sound before expecting them to attach them to letters, (Note, 2019, more accurate to say that letters and sounds are taught together, not that phonological awareness is taught on its own) the importance of fast, accurate word identification in the development of phrasing (Perera, K, Phd, Manchester 1989); and the detailed instruction in sound-symbol correspondence in the Reading Recovery programme (2019 note: The element “Hearing the Sounds in Words” is not enough to secure this connection in the context of the program as a whole, which has taken scant, if any, interest in the research in phonics). Using pictures as cues to words is a pre-reading skill, and should be moved to Level 1. (2019 note: this again relates to the 1989 NC – levels are now abolished.)

B and D confusion. Using speech to clarify the problem.

In the early twentieth century, Annie Sullivan (Helen Keller's teacher) and Grace Fernald discovered the benefit of using movement to reinforce learners' understanding and use of letters for communication. Writing "d" as an extension of "a", for example, made it less easily confused with "b" than traditional printing, which starts both letters at the top, and leaves a decision to be made at the bottom of the first stroke - does the ball go to the left, or the right? With my current fourteen year old, this was not enough. She did not make errors in reading, but would confuse the letters when writing, even after practice.

The solution lay further back, in the production of the sounds "b" and "d" in speech. If we say "dad", we can feel our tongue touch our top palate, just above the teeth, and our lips don't touch. When we say "boy", we press our lips together to release the consonant (keeping our tongue low, though this is incidental to the teaching point). So, saying the word we want to spell gives us a direct and reliable indication of which letter to use. Lips touch - b. Tongue touches roof of the mouth - d. So far, it is working.

Ending the Reading Wars: Reading Acquisition From Novice to Expert; A note on the first part of the paper.

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

Another casualty of guessing game and multicue theories of reading, and a solution.

For the last two weeks, I've been teaching a perfectly intelligent young man who had reached the age of 17 without being able to read at all. All the way through school, he had been told to guess rather than taught how to use the information contained in letters. He had then, of course, been assessed as dyslexic, an assessment that indicates a failure in the schools he attended - the last of them consistently rated outstanding by Ofsted - rather than any disability on his part.

I say that because, as usual, a clear explanation of how English spelling operates had him beginning to read complex and simple words by the end of our first session. Working on the materials for his FE course brought predictable problems from residual guessing - eg reading assignment as assessment - but that had gone by the end of the second session. This particular error in education has North American origins - Frank Smith and Kenneth Goodman - and I thought we'd got rid of it. Apparently, there is still work to do.

I'll post more fully on this case, which was referred to me by my MP, as the work progresses. For the moment, I'm particularly interested in any comments from colleagues in North America and other parts of the English-speaking world in which guessing game and multicue theories of reading persist. They are a cardinal error, but are not to be replaced by the other error of pretending to children that phonics is the only game in town. Phonics are the basis of learning to read, but without explanation of how and why they vary, they cause problems with individual children that are nearly as bad - try to sound out the word "the" one letter at a time, and you'll see what I mean.

PS. For friends who have not seen the explanation of "th", it was introduced by William the Conqueror's scribes as a Latin alphabetic alternative to the Anglo-Saxon þ, that denoted this sound.

Would we, could we, should we? From a Guardian comments discussion with Michael Rosen.

This revolved around how to teach a child to read would.  Michael used context with his children - read on and see if that gets you to the word. My reply was that this worked sometimes, but depended on being able to read the context, which a weak reader was less likely to be able to do. I added that if I did use this approach, I would explain the connection between would, could and should, so as to begin to build a mental network.

But what was the connection? Would and should,  have a Germanic origin containing an l, which is retained in the modern equivalents, wollen and sollen. Could is different - the squire in Chaucer's Prologue coude do several things, with no l in sight. The OED tells us that, around 1525, its spelling was changed to match that of would and should, an arbitrary decision, but which at least gives us a group of three words rather than two, and so makes the whole easier to learn, once it is explained. Very easy to reinforce with all sorts of games for adults and children.

Reading Reconsidered, Doug Lemov, Colleen Driggs, Erica Woolway. Plus some thoughts on phonics.

Towards the end of my time with Ofsted, I led the inspection of an unusual primary school in London, where nearly all of the pupils spoke English as a second language. All of them - except for some who arrived just before they took the tests - met the nationally expected standard in English, maths and science,  and two-thirds of them exceeded it. No fewer than seven of the staff were guest lecturers at the Institute of Education, and our report was used as evidence when the school won the Evening Standard School of the Year award.

We were neither the first nor the last to recognise the excellence of Gateway Primary School, but what was really interesting was how they did it. Careful and intensive initial teaching of reading, using the Jolly Phonics scheme, was followed by systematic use of reading in virtually every lesson in the junior school, using texts designed to extend the children's vocabulary and understanding beyond the everyday language that they met outside the classroom. The outcome, in sharp contrast to the "first-hand experience" that I'd seen preached in Essex, was that the pupils developed the linguistic understanding they needed in order to make sense of the rest of their work, and to succeed in secondary school.  They were, as Nicky Morgan has put it, "secondary ready",  at a time when most of the pupils we were seeing in other schools were not.

Behind this success lay brilliant planning by a senior member of staff, who understood paths of language development, guided her colleagues with great clarity, and subsequently became the headteacher.  How, though, could we extend this excellence so that all children could benefit from it?  The guest lectureships were clearly not enough - Gateway's principles needed to be understood and applied throughout the school system, and there was no easy path.

Doug Lemov, managing director of the US group Uncommon Schools, and author of Teach Like a Champion, has provided such a path in his new book Reading Reconsidered. He began with an unsolved problem. The good discipline and hard work of his organisation's pupils were not leading to the "march of triumph" they had expected when they went to college, and their ability to handle the demands of advanced, independent reading, was the main cause. Mr Lemov and his colleagues, Colleen Driggs and Erica Woolway, have tackled it by recognising the importance of all types of sustained reading to the development of pupils' intellect, and putting into their book a very wide range of source material and techniques that make it the most significant contribution to reading teaching since the Clackmannanshire research on early phonics.

 The technique behind Teach Like a Champion was to analyse excellence and present it clearly to teachers in training, with the aim - largely successful - of putting old heads on young shoulders. Reading Reconsidered does the same for reading.  Many of its techniques - for example, on ways to prompt pupils when they are hesitant, on developing phrasing in reading, and in selecting texts that go beyond their immediate experience, are ones I've found valuable myself, though I had to learn by stumbling into them. This concentrates so much outstanding material in the space of a book, plus an excellent DVD showing the approach in practice, that it should be read by every headteacher and English co-ordinator in the country and incorporated into training for new and practising teachers.

Two pupils I've been teaching over the past few weeks bring me back to phonics, and its limitations as well as its strengths.  They are aged 9 and 14, have parents of mixed Spanish and English nationality, and had been assessed as dyslexic because of severe problems with reading in English with, in one case, an apparent weakness in memory. Spanish is one of the most regular languages in the world, with no surprises in spelling and only one silent letter - h, at the beginning of a word. What you see is what you say, and what you say is what you write. English, as I discovered some years ago by trying to have a pupil sound out a word that could not be sounded out, is different. We need not only to know and blend the most frequent sounds indicated by letters - synthetic phonics - but to understand and use the variations between words such as should and shoulder, and interactions between letters that change the indicated sound, such as cat and centre.  Evidence from brain scans has shown that English speakers develop activity in an identifiable area of the brain - (Blakemore and Firth, The Learning Brain 2005) that is not needed in a regular language.

Fluent readers often adapt to these variations spontaneously. Others, like these two girls, need to have them pointed out and explained, and then to practise. Allowing them to choose their own books , reflecting the interests of children of their age, led them to meet the variations in English as they occur in the normal run of writing. Explaining these and their origins, usually by switching to a word with a similar pattern to the word that had caused the difficulty, allowed words such as neighbour, and tinier, to be unpacked, explained and learned. The parents were, as usual, delighted. "I've never heard my daughter read so fluently," said one, and the other, having seen me move from shine, to shiny, to tiny, to tinier, described the process as "amazing".  Not a split digraph or phoneme in sight, and no more mention of dyslexia.

Grace Fernald's techniques for severe reading difficulties. A letter from Australia.

Karina McLachlain, who worked with me in London, has moved to Australia and has sent this letter describing her work with a young pupil.

Hi Mr Bald,

I had a mum bring her son from Mudgee yesterday for tutoring. This is a 1 hour 45 minute drive from Dubbo. He has problems similar to the pupil we worked with in London. He has difficulty with the irregular words, not so much with phonically regular words.

Although Fernald and yourself only use tracing with non-readers I decided to use the tracing method with him for the tricky words. I asked his mum to bring a list of words that he has trouble reading and spelling on a regular basis. The ones he brought were words like were, where, there, though etc. His mother said that no matter how many times he reads those words or tries to remember them, he can't do it.

I followed the tracing method as best as I could from the descriptions from the book and journal article. Once he traced a word several times, I took the word away and asked him to rewrite it without copying and he was able to do it. I also asked him to close his eyes and try to remember the sight of the word and spell it for me from memory and he was able to do it. After that I wrote the words in a grid like you do and asked his mother to check the words with him each day to see if they had been stuck in memory. If not, I asked her go back to the beginning and follow the tracing method for any word when his recall was poor. This is something that Fernald mentioned doing for some children. I don't anticipate that most words will need to be taught more than once from the beginning, but Fernald mentioned that some children took a while to catch on and then something 'clicked'.

The boy also has fine motor skills problems and very messy writing. Before beginning the tracing method this lesson, in past lessons I was working with the boy on beginning cursive writing. I gave his mother some exercises to get his arm flowing across the page and then started with letters and joins. Then when we had enough letters to make some of the tricky words, we began tracing some of the words on his list. At first I guided his hand, then let him do it himself. After that I rubbed it out and asked him to write it out in cursive script. It worked well. Then I asked him to read it in print, which he did. Then I asked him to close his eyes and spell the word aloud, which he also did.

Everything is going well at the moment and the mother is pleased. As we go on with tricky words, I will be explaining the irregularity of the language in the way that you do and link similarly irregular words together. Once we have covered a sizeable chunk of tricky words, so that they no longer present a big stumbling block to his reading, I thought about getting a programme that specialises in fluency, as his reading is slow, even when the passage is mostly phonetic.

I'm quite happy with the Fernald method and using your refinements as well. This means that I can deal with more reading problems than the ones that Orton Gillingham programmes deal with. O-G do not have a good grasp of how to teach irregularity, so are very good for phonic dyslexia, but not so much for surface dyslexia.

I'm very grateful to you for recommending that I become familiar with Fernald's work. I think that all teachers of reading should be required to learn her methods.

Thanks and kind regards,

Karina

More speed, less understanding.

Guardian has an excellent report of a survey of speed reading techniques, showing a loss of understanding when people try to read more quickly than their brains can process the information. I've never believed in speed reading, or skimming and scanning, except when something is not really worth reading properly - typically, the morass of policy documents produced under New Labour's strategies. What is worth reading is worth reading either at the pace of the human voice, or at the pace of our own inner speech, which is faster, but takes shortcuts we can control.