Bullock (1975)

Notes on the text
Preliminary pages Foreword, Membership, Contents, Introduction

Part 1 Attitudes and standards
Chapter 1 Attitudes to the teaching of English
Chapter 2 Standards of reading
Chapter 3 Monitoring

Part 2 Language in the early years
Chapter 4 Language and learning
Chapter 5 Language in the early years

Part 3 Reading
Chapter 6 The reading process
Chapter 7 Reading in the early years
Chapter 8 Reading: the later stages
Chapter 9 Literature

Part 4 Language in the middle and secondary years
Chapter 10 Oral language
Chapter 11 Written language
Chapter 12 Language across the curriculum

Part 5 Organisation
Chapter 13 The primary and middle years
Chapter 14 Continuity between schools
Chapter 15 The secondary school
Chapter 16 LEA advisory services

Part 6 Reading and language difficulties
Chapter 17 Screening, diagnosis and recording
Chapter 18 Children with reading difficulties
Chapter 19 Adult literacy
Chapter 20 Children from families of overseas origin

Part 7 Resources
Chapter 21 Books
Chapter 22 Technological aids and broadcasting

Part 8 Teacher education and training
Chapter 23 Initial training
Chapter 24 In-service education

Part 9 The survey
Chapter 25: I Introduction
Chapter 25: II Primary commentary
Chapter 25: III Secondary commentary
Chapter 25: IV The questionnaire forms (not online)
Chapter 25: V Technical notes (not online)

Part 10 Sumary of conclusions and recommendations
Chapter 26 Conclusions and recommendations

Appendix A Witnesses and sources of evidence
Appendix B Visits made
Glossary
Index

Report of the Committee of Enquiry appointed by the Secretary of State for Education and Science under the Chairmanship of Sir Alan Bullock FBA

London: Her Majesty's Stationery Office 1975
© Crown copyright material is reproduced with the permission of the Controller of HMSO and the Queen's Printer for Scotland.

Chapter 6 The reading process
[pages 77 - 96]

6.1 Controversy about the teaching of reading has a long history, and throughout it there has been the assumption, or at least the hope, that a panacea can be found that will make everything right. This was reflected in much of the correspondence we received. There was an expectation that we would identify the one method in whose adoption lay the complete solution. Let us, therefore, express our conclusion at the outset in plain terms: there is no one method, medium, approach, device, or philosophy that holds the key to the process of learning to read. We believe that the knowledge does exist to improve the teaching of reading, but that it does not lie in the triumphant discovery, or rediscovery, of a particular formula. Simple endorsements of one or another nostrum are no service to the teaching of reading. A glance at the past reveals the truth of this. The main arguments about how reading should be taught have been repeated over and over again as the decades pass, but the problems remain.

6.2 A study of the way these arguments have been advanced, contested, revamped, discredited and rediscovered is a useful corrective to the idea that any one of them has a monopoly of truth (1). In the last four centuries there has been a succession of them, making claims for word methods, sentence methods, experience methods, phonic methods, and so on. It is interesting to note that they were usually introduced with the description 'new' or 'natural' or 'logical'. Today's discovery was often yesterday's discard, unrecognised as such, or rehabilitated by some new presentation. This does not mean that there has been no advance, that nothing really new has emerged across the years. There have, of course, been many innovations of one kind or another, notably in materials. But the major arguments are substantially the same as they have always been, and to endorse one at the expense of the others is no more helpful today than it has proved in the past.

6.3 Among authorities on reading there is, in fact, considerable agreement, and in recent years they have done much to reduce the polarisation of opinion. There is no doubt, however, that this does still exist, and it characterised much of the evidence we received. One issue that has received more than its share of this kind of attention is that of approaches to the teaching of reading in the early stages. It is argued on the one hand that the essence of the process is 'breaking the code', converting print into sounds and then into words; it is argued on the other that this must take second place to securing and expanding the child's interest, keeping his curiosity alive, and giving reading a meaning. Immediately, a false conflict is created which leads to a number of unnecessary tensions. Some would put so much emphasis on the 'mechanics' of reading that certain children would be handicapped rather than helped. Others advocate so keenly the virtues of mature reading from the beginning that they are in danger of leaving it too much to trust that the skills will be acquired on the way. The children would thus be left ignorant of vital information about the nature of the written code. This emphasis fails to acknowledge that the majority of children also require precise, well-organised instruction if they are to become successful readers. In our view a large part of the controversy arises from the expression of unnecessarily extreme opinions, often more extreme than the real beliefs or practices of those who advance them. In addition, the contentious statements are often based on inadequate information. For example, we received many letters whose writers seemed convinced that the majority of infant teachers had abandoned the teaching of phonics; they argued that a return to the practice would raise standards dramatically. But the results of our survey showed that their supposition was far from correct. The teachers of six year olds in our sample were asked which approaches they were currently using. The results were as follows:

We believe that an improvement in the teaching of reading will not come from the acceptance of simplistic statements about phonics or any other single aspect of reading, but from a comprehensive study of all the factors at work and the influence that can be exerted upon them. In the course of this sequence of chapters, therefore, we shall outline what we believe to be necessary for the effective teaching of reading - from the earliest stages to the advanced skills required of the educated reader. We believe, however, that a fundamentally important question has to be answered before there can be any discussion of how the teaching of reading can be improved. What is Reading? Much of the misunderstanding surrounding the debate about reading results from the lack of a proper examination of what the process involves. Before considering the ways in which children can best learn the skill one must be clear about what is expected of them, in both the short and the long term. This knowledge should then inform decisions about the organisation of the teaching within the school, the kinds of initial and in-service training needed, and the resources required at each level. Thus a detailed understanding of the reading process is of critical importance in terms of its practical implications. It is for this reason that the account which follows includes a good deal of technical detail. We regard this as essential to our task, for we do not believe that a Report making recommendations about reading can examine the issues fairly without defining what is involved for a child when he is learning to read. We must also emphasise here that our discussion of reading is not confined to this section of the Report. Parts Six and Seven have a particular relevance to this one, but since references to reading occur throughout the Report we hope that all the chapters will be read in close association with one another.

6.4 It may be useful to begin by looking at some of the ways in which reading can be defined:

'One can read in so far as he can respond to the language skills represented by graphic shapes as fully as he has learned to respond to the same language signals of his code represented by patterns of auditory shapes.'

'The purpose of reading is the reconstruction of meaning. Meaning is not in print, but it is meaning that the author begins with when he writes. Somehow the reader strives to reconstruct this meaning as he reads.'

'A good reader understands not only the meaning of a passage, but its related meaning as well, which includes all the reader knows that enriches or illumines the literal meaning. Such knowledge may have been acquired through direct experience, through wide reading or through listening to others.'

6.5 These three definitions may be represented as follows:

A response to graphic signals in terms of the words they represent

A response to graphic signals in terms of the words they represent

plus: A response to text in terms of the meanings the author intended to set down

A response to graphic signals in terms of the words they represent

A response to text in terms of the meanings the author intended to set down

plus: A response to the author's meanings in terms of all the relevant previous experience and present judgements of the reader

Finally, there is the view that looks beyond the reading process as such to the range of activities demanded of the adult reader, with all that they imply in social and economic terms. These implications are taken up in Chapter 8. There are, in effect, two basic approaches to the definition of reading. One is to start with the complexities of print. The other is to start with the potential reading demands of the modern world and to define as reading whatever is logically involved in meeting those demands. Taken together they give a more complete understanding than either one could afford alone.

6.6 The reader responds to print at a number of levels. At one level he recognises the shapes of separate letters, groups of letters, and whole words, and he associates appropriate sounds with those letters or collections of letters. The responses at this level are fundamental to reading, and the ability needed to make them may be regarded as 'Primary Skills'. The reader must have a reasonable mastery of this process of seeing a letter or group of letters as a discrete whole before he can respond at another level, i.e. to sequences. The ability to handle sequences - of letters, words, and larger units of meaning - is essential to fluent reading. The various skills involved have been described as 'Intermediate Skills' because they operate at a level above that of the primary skills but below the level of 'Comprehension' in the extended sense of Gray's definition. In examining each of these levels we need to consider three features. The first is the graphic element, the printed word or page; the second is the language element, the sounds, words and meanings to which the print relates; and the third is the pattern of relationships that may be established between the other two in the mind of the reader. An analysis of these will give an indication of the many points of possible difficulty at which a child may falter in learning to read. More positively, it can help the teacher to modify or extend the child's existing skills to lead him to a higher level of general reading competence.

6.7 At the level of the primary skills the child has to learn to perceive separate units - individual letters or groups of letters, and individual whole words. To do this he must learn to respond to two fundamental attributes of letters: shape and orientation. It may seem rather obvious to the adult that a child has to learn to respond to letter shape. But letter outline may convey very little to a child unless it has been invested with some kind of special significance. He may get this by watching someone trace or draw letters, by doing so himself, or by exploring the shape of a three dimensional letter in wood or plastic. Without such experience his interest may be confined to the colour of the letter, its size in relation to the background, or some fanciful pattern that its appearance suggests to him, much as one sometimes sees fanciful images in clouds or ink blots. To see letter shapes as adults see them is by no means a natural and automatic process. On the contrary, each child may have his own idiosyncratic ways of looking at letters, and to see them as the adult sees them means he has to develop a generalised 'learning set'. Since so many children respond to letter shape very readily it is easy to forget that many others may never have enjoyed opportunities which are necessary to accomplish this. Indeed there will be some who have actually acquired 'learning sets' which obstruct them when it comes to responding appropriately to letters. On the other hand, it is perhaps because parents have so often prepared the way fairly well that so many teachers think this aspect of reading requires little attention. This sometimes leads to their assuming mistakenly that there is something inherently wrong with the child if he happens to have difficulty in learning to recognise letters.

6.8 Learning that orientation is a critical aspect of letters may also present problems. For the first few years of his life a child learns to ignore orientation as a means of recognising objects. The doll or the toy train is still a doll or a train whether it is the right way up, upside down, facing left, facing right, or lying on its side. This is a critical part of what has been called 'conservation of identity', the fact that things retain their identity over time and in spite of changes of position or temporary disappearance. At the same time, even the youngest children have little difficulty in orientating themselves correctly to objects when they want to. They can open doors the right way, turn book pages forwards or backwards, turn cups up the right way, and so on. However, when they come to letters they have a problem. They now have to learn that b is not d, p, or q; or that f is not t, and n is not u. If we include such similarities in shape as h : y and m : w, then it becomes clear that over half the letters of the alphabet are ambiguous in terms of the child's previous learning. It is not that children have particular difficulty with orientation as such. They can see as well as any adult that b is p upside down, just as well as they can see that the doll or train is upside down. Where they have trouble is in learning to recognise these reflected and rotated forms as entirely different letters. They would regard it as very odd if a doll had to be called Betty instead of Susan according to which way it faced, or the train was called a motor car when it reversed.

6.9 Of course, children are extraordinarily flexible in what they can learn to cope with when they are strongly motivated. Letters, however, are often much less fun than dolls and trains. The difficulty with letter orientation is that there is no strong incentive to acquire this particular 'learning set'. Moreover, as we have said, it seems to run completely counter to an existing 'set' that has already become very firmly established, i.e. 'ignore orientation in identifying objects'. The situation and materials must therefore be particularly well designed if the child is to 'unlearn' and then relearn in the right way. Reversal tendencies are in fact quite persistent, even in normal readers. It is scarcely surprising that they represent an important proportion of the problems experienced by the children who have difficulty in the early stages of reading.

6.10 When children have learned to respond to a combination of shape and orientation they still have to learn each of the 26 letters of the alphabet. To these can be added the 17 shapes of those capital letters which are very different from the lower case forms, i.e.

This gives a total of 43 letter shapes. In addition to these there are also such typographical variations as:

Encounter with such variations is inevitable, because of the wide range of printed materials to which children are exposed both before and after starting school. Children can, of course, learn all these individual variations. It is simply that they increase the total quantity to be learned and add to the burdens of the slow learning child an extra dimension of difficulty that he could well do without. This difficulty is probably even more marked when the child comes to write, since he may be confused in deciding which of the various forms to set down.

6.11 A more important problem relating to letter shape arises at the level of word perception, as distinct from letter perception. It is often argued that children should be taught to recognise whole words rather than respond to individual letters. Unfortunately, variations in letter shape multiply at the word level, as may be seen in the examples given below. Adults are so familiar with these variant word forms that it is hard for them to appreciate how different they are. The following set of unusual equivalent symbols will give an idea of what they may look like to a child:

Moreover, the whole-word forms of certain different words, e.g. 'hot' and 'hat', are no more different than the variations of the same word e.g. 'hat' and 'hat'.

In spite of this, fluent readers can readily cope with variations such as the following:

iT Is MOsT uNliKELy tHaT tHe
ReADer wiLL haVE preVIOusly
sEEn tHEse WholE WORD SHapES
How DOes hE rECoGNise THem?

6.12 For a further comment on the nature of whole-word perception we must note that detailed vision is possible only for objects that fall within a maximum of 3° of visual angle. What this means for the young reader has been revealed by studies of eye movements. It was found, (5) for example, that among a group of seven year olds the average number of fixations was 2.4 per word. Though they could recognise some words at a single glance most words required an examination of individual letters or groups of letters. If the whole of a word does not fall within the area of clear vision this finding is scarcely surprising. Children do, of course, learn to identify words correctly in running text without scrutinising meticulously every single detail of them, and how this may be achieved is discussed later. Nevertheless, the accurate perception of individual letters and groups of letters is clearly an important factor in learning to read.

6.13 There is no doubt that if children are introduced to letters in inappropriate ways these can have a harmful effect on their subsequent learning. Unfortunately, this has sometimes been used as an argument that letter recognition should not be learned at all. We do not accept this argument. There is, in our view, any number of perfectly reasonable ways in which a teacher or parent may help a child to learn to recognise letters. These include such familiar practices as the following:

• drawing attention to the shapes of letters in an alphabet book;
• letter-matching activities (provided that there is some clear clue to help the child place letters the right way up, e.g. a coloured base line, or a jigsaw shape);
• tracing or colouring letter outlines;
• writing letters in a sequence which helps the child to establish letter differences (e.g. b=downstroke, then clockwise movement; d= anticlockwise movement, then upstroke and downstroke);
• collecting variant forms of the same letters in a scrap-book (e.g. T, t, t);
• mnemonics (e.g. 'S is for S-s-s-snake', or 'O is for orange - you can tell by the shape'.)

6.14 We do not suggest that children of any age should be subjected to a rigorous and systematic training programme of exercises based on this kind of activity. A limited number of such experiences will be quite sufficient to help most children learn to attend to the relevant characteristics of the letters. The important factors are precision in the design of the learning task and careful supervision. Letters will be learned more easily if the materials used are varied in such a way that only the invariant properties of the letters remain constant. Thus, the same letter shape may be presented in different colours, sizes, and materials, against different backgrounds, and in different forms (e.g. T, t, t). Many children learn their letters so quickly that a very limited experience of such activity is sufficient. For those who do not, these variations can help to sustain interest in a fairly limited learning task. Moreover, it will help the child to transfer his learning to the other contexts in which he will meet the letters he has learned, e.g. in word games or in books. Letters which are easily confused should be learned separately. Examples are b, which differs only in orientation from p, q, and d; and h, which has a similar shape to b. The longer an error is allowed to persist the harder it is to eradicate, and it is therefore much better if the initial confusion can be avoided. This may be achieved by the 'over-learning' of any one of the letters which are easily confused, and only then giving attention successively to each of the others.

6.15 Contrary to popular belief the majority of children are perfectly capable, well before they start school, of making the perceptual discriminations necessary for learning letter shapes. They learn to make extremely complex auditory discriminations in language, and show their ability with similarly complex visual discriminations in playing with many of their games and toys. There seems to us no obvious reason why they should be denied opportunities to become familiar with the letters of the alphabet before they start school. The question of reading readiness and the parents' contribution is discussed in greater detail in the next chapter.

6.16 We come next to the relationship between letters and sounds. Single letters, or groups of letters, represent sounds called 'phonemes', which enable the reader to make distinctions between different words. There are approximately 44 phonemes in English. For children with normal hearing the ability to read depends, in the first place, on the ability to distinguish most of these phonemes in normal speech. Unfortunately, they are not quite such clear-cut units of sound as they may appear:

'A tape recorder can be used to confirm a number of remarkable findings of speech analysis. As an example, a tape recorder will demonstrate that 'dim' and 'doom' have no /d/ sound in common. If the two words are recorded, it is impossible to cut the tape in order to separate the /im/ or /oom/ from the /d/. Either one is left with a distinct /di/ or /doo/ sound or else the /d/ sound disappears altogether. One is left with two quite different kinds of whistle. There is no /d/ except as part of these quite different consonant-vowel combinations.' (6)

6.18 As there are only 26 letters but 44 phonemes certain letters have to be used more than once if each phoneme is to be separately represented. These additional sounds are often represented by two-letter combinations called digraphs, e.g. ch, th, ur, aw, ou, or by larger groups of letters such as ough for the last vowel sound in borough. Learning to respond to spelling patterns such as these should present no serious problem to the majority of children. Instead of learning the shapes of additional letters they simply have to learn to treat particular combinations of known letters as single units. Of much greater importance in this matter of establishing relationships between letters and sounds is the fact that there is no simple correspondence between the 26 letters and the 44 phonemes. If one were intent on constructing an alphabetic writing system from scratch the obvious course would be to aim at a one to one correspondence between phonemes and graphemes, the grapheme being any letter or combination of letters which represents a single phoneme. Some idea of the ways in which written English falls short of this alphabetic ideal may be seen in the following examples:

i. one home comes women of or to do

ii. aisle height eye I phial ice high island buy guide sty rhyme

6.19 In one study the researchers (8) examined the 6,092 two-syllable words among the 9,000 words in the comprehension vocabularies of a group of six to nine year old children. They recorded 211 different spellings for the phonemes in these words, and these required 166 rules to govern their use. Over 10 per cent of the words still had to be left aside as 'exceptions'. Sixty of these rules applied to consonants, which are usually thought to be 'regular'. This means that even if a young child memorised these rules while learning to read he would still encounter hundreds of words not governed by them. Although there is certainly a great deal of value in learning to deal with the regularities that do occur, the problem for the beginner is that words do not come tagged to indicate the spelling family to which they belong.

6.20 The word printed below gives some impression of the kind of problem that confronts a child when he has to combine graphemes and phonemes in a phonic attack on an unfamiliar word. This word is in the vocabulary of most adults in this country, but it has been spelled here in an unusual way (though the spelling conventions that have been used are common enough in other contexts).

Obviously, the word cannot be identified at a glance, and the following list gives some examples of sound values that are quite commonly represented by each of the letters:

c: centre; candle; cello
a: man; call; father; cable; many; wand; errand
l: calm; colt
m: ham
b: cymbal; lambing
o: from; whom; tomorrow; come; home; women; form
s: lost; lose
t: listen; station; lost

*The word is 'chemist'.

6.22 Other solutions devised to help the reader cope with the difficulties of irregular spelling are the use of diacritical marks or of colour coding, simplified spelling for the early stages of reading (e.g. the initial teaching alphabet), and vocabulary control of one kind or another. All these are considered in the next chapter. One conclusion is inescapable at this stage: the teacher needs a sound understanding of the problems created for the learner by this evident irregularity in the phoneme-grapheme relationship of the English writing system.

6.23 Broadly speaking, there are two ways in which a child can learn the correspondences between phonemes and graphemes. One is by attending directly to the sounds and letters and the way they relate to one another. The other is by attending to whole words and their pronunciation, and over a period of time learning to make intuitive generalisations about phoneme-grapheme relationships. Even if there is no explicit teaching of phonics, many children will still work out phonic relationships for themselves. The above examples will have shown that although the child must steadily acquire a considerable amount of phonic knowledge he certainly does not use it in any simple way. We must therefore reject as highly suspect any phonic drill which causes children to pay more attention to single phoneme-grapheme relationships than to sequences. Competence in phonics is essential both for attacking unfamiliar words and for fluent reading. The question, then, is not whether or not to teach phonics; of this there can be no doubt. The question is how and when to do it.

6.24 We have already noted that letters pronounced in isolation tend to be very different from the phonemes they purport to represent. To teach a child that 'kuh-a-tuh' says 'cat' is to teach him something that is simply incorrect. It is very doubtful in fact whether in a strictly controlled experiment this way of attacking new words would prove to be much more effective than the old alphabetic method ('see-ay-tee' says 'cat'). Both owe their initial success to what psychologists call 'mediated learning', but at the expense of having the child acquire responses which must later be unlearned. A more common practice these days is for teachers to get children to synthesise rather larger units. The groupings are chosen in such a way as to reduce distortion and allow a genuine blending of sound, e.g. 'bl-ack' says 'black'. It has been argued that the whole syllable is the more appropriate unit, since this produces the minimum of distortion. But even at the level of the syllable the child is still faced with difficulties. If he has been taught to treat 'basket' as 'bas-ket', he will find 'wal-ked' of little use of him when trying to pronounce the superficially similar 'walked'. The rules for syllabification are no less complex than those for English spelling; by the time they are learned the child is past the stage of learning to read when they might have helped. More important, however, is the fact that they often conflict with the morphemic system; and morphemes,* not syllables, are the units of language. There is a case, therefore, for emphasising morphemes from a quite early stage. Thus, when a child encounters the word hear there is everything to be said for showing him that this is the base word in hears and hearing. In this way the emphasis is placed on the relationship between spellings and meanings and not just spellings and sounds. This point is taken up again in paragraph 6.38 after we have considered the importance of intermediate skills in developing word recognition.

*A morpheme is the smallest unit of meaning in language. Both 'hen' and 's' in 'hens' are morphemes. A syllable, on the other hand, relates essentially to pronunciation.

*Vocabulary control is the control of the rate at which new words are introduced and how often they are repeated.

6.27 We described the intermediate skills as the ability to handle sequences of letters, words, and larger units of meaning. To acquire it the reader has to become familiar with the probability with which sequences occur. In other words, when he is reading a sequence he has to be able to anticipate what is most likely to follow it. By this means he reduces the number of possibilities to be considered when he encounters an unfamiliar word. Equally, he reduces the time taken to identify familiar words in fluent reading. He is also able to isolate the specific meaning of the word which changes its meaning according to context. It is important for the teacher to have some understanding of the process if he is to develop appropriate teaching methods. We shall therefore examine each of the levels at which the intermediate skills function.

6.28 A word consists of letters sequenced in a particular way. The left to right direction is, of course, an arbitrary convention, and there are languages in which it is right to left, or even vertical. Whatever the language, there are no strong reasons for thinking any one convention to be more 'natural' than any other. The important conclusion from this is that the direction in which letters or words are written and read does not come naturally; it is another 'learning set' that has to be acquired. As with the orientation of a single letter, this conflicts to some extent with 'learning sets' the child has already acquired. He may, for example, have learned to order the carriages in his train in a number of different ways, and to run it in any direction, but it still remains the same train. Little wonder, therefore, that many children are bemused by 'on' and 'no', 'was' and 'saw', 'won' and 'now', 'stop' and 'pots' etc; and these are only the more obvious examples. It is less generally appreciated that children often begin to attack a word correctly from left to right, reverse some of the medial letters, and end up, perhaps, with the correct final letters. Sometimes the reversal makes a real word, as when a child says 'clot' for 'colt' or 'trail' for 'trial', but he will often produce a nonsense word, assuming, perhaps, that this is a word he has not yet absorbed into his spoken vocabulary, e.g. 'engery' for 'energy'. The fact that adults often make mistakes of this kind shows how persistent these faulty word attack habits can be, and how important it is to ensure that they do not get firmly established in the first place. The first essential is to recognise that children cannot take in unfamiliar whole words in a single fixation. This provides a starting point for considering possible ways in which these wrong habits can be avoided.

6.29 There are many letter sequences that occur rarely, if at all, in written English, e.g. dx, kng, wpvt. At the other extreme there are many that are very common, e.g. bl, str, atio. In principle, therefore, for any particular letter sequence one could calculate its frequency of occurrence in the reading experience of any child or adult. The more frequently a letter sequence occurs the more likely it is that the reader will come to expect the remaining letters to follow whenever he sees the first letter of that sequence. If, for example, he sees, the letter 'p' in a particular fixation he may anticipate such letters as a, e, h, i, l, etc at varying levels of probability. The expectancy that he will see such letters as b, c, d, f, g, etc will be virtually zero. As he takes in more and more letters, however, the range of possibilities gradually reduces. Thus 'pr ...' produces a narrower range of expectancies than 'p' alone, and 'pri ...' a still narrower one. Even if confronted by part of a nonsense word such as 'redulanti ...', the majority of readers would be likely to anticipate only such possibilities as 'ng', 'on', 'ous', 'c', or 'ne' to complete the 'word'. There is a good deal of evidence to suggest that speed in identifying letters is closely related to the degree of expectancy that the given letter will occur. If only one letter is possible, recognition will be virtually instantaneous. If the number is as many as five the perceptual task is slightly more complicated and takes more time. If any one of the 26 letters could be anticipated the problem becomes obviously very much greater and takes even longer. If, however, groups of letters, rather than single letters, can be anticipated the speed of response is much increased.

6.30 The swift identification of single letters, or groups of letters, is obviously of critical importance in fluent reading. However, teaching children to recognise isolated letters, or groups of letters, is only part of the problem. The other part is to help them develop the habit of anticipating likely letter sequences. Children must therefore learn to do two things at the same time, i.e. identify one letter, or group of letters, and anticipate the next. It follows that any excessive emphasis on one at the expense of the other may well hamper the development of this rather delicate complex of skills. The effects of inadequate teaching in either skill can be seen in the reading behaviour of certain children. There is the older child who still reads haltingly and falls back very quickly to an examination of letters. On the other hand there is the child who reads quickly and makes all manner of 'careless' mistakes.

6.31 All the points already made about letter sequence and probability apply equally to phoneme sequences. Thus, not only has a listener to recognise phonemes as they are uttered, but also to anticipate very efficiently those that are most likely to follow. Only by this means has he any chance of keeping up with a speaker's rapid flow of words. Even an incomplete nonsense word spoken aloud will cause anticipations of a limited number of possible endings, as did the written nonsense word 'redulanti ...' In tackling an unfamiliar word, therefore, the reader can call upon this knowledge of probable phonemic sequences to match the word on the page - as was no doubt evident to those who correctly deciphered the nonsense spelling 'calmbost' as 'chemist' in paragraph 6.20. Most words do not present so many alternative possibilities. In many cases, an unfamiliar word may be decoded by a child as easily as an adult might decipher an unconventional spelling such as 'trand', where the sounds 'tr-nd' suggest the possibility of 'trend', or 'trained', in terms of known words. What we would emphasise here is the fundamental importance of prediction in attacking unfamiliar words. Words are recognised as a result of matching a small number of possibilities against the printed model rather than by mechanically working through all the possible sound values of the separate elements.

6.32 We have been discussing the implications for word attack of letter and phoneme sequences within a word. When it comes to reading groups of words the child has to acquire a 'learning set' to read successive words from left to right. This convention is simple enough to establish, but to achieve it there must be some teaching, or deliberate structuring of the learning experience. It should be learned early, or some initial confusion may provide yet another adverse reading experience. Though the child would acquire the skill in due course this early confusion could still contribute to a negative attitude to reading. It must be remembered, however, that even when it is established there is still a tendency for the eyes to move back and forth across a line of text. This is because additional factors are at work. A word might be incorrectly identified at first glance and this could become obvious from the words that follow. Moreover, it is not always possible for the reader to sort out the syntactic structure of a sentence at first reading. The structural cues provided by print are by no means as powerful as those of the intonation patterns of speech, and punctuation can sometimes confuse rather than make plain. Even when the structure is clear the meaning might not be, and this is another reason why it is often necessary to do a certain amount of backtracking. There are, then, a number of different causes of regressive eye movement. Once the habit has been established of reading from left-to-right, backtracking is simply a sign that the reader is having problems with the text at some conceptual level. It is a case, then, of deciding whether or not he needs help at this level rather than of concentrating on the symptom.

6.33 As in the case of letters, words vary in the degree to which they can be expected to follow certain other words. Since the number of different words in use vastly exceeds the number of different letters, it follows that in the case of the former there is much less chance of being able to anticipate a particular sequence. Among the high frequency words, however, there is a slight tendency for some words to collocate, e.g. 'on the', 'in the'. There are about a dozen high frequency words which together account for approximately a quarter of any piece of continuous written material. If the child learns these thoroughly and reads them frequently in running text a useful contribution can be made to his fluency. However, there should not be an undue emphasis on recognising them in isolation, since this could encourage the wrong kind of 'learning set'. Apart from these high frequency words there are others that tend to collocate in everyday speech, e.g. 'bus stop', 'telephone call', but they are not encountered so frequently as to make it worthwhile to give special attention to their visual forms.

6.34 Word recognition is also made easier by the ability to anticipate syntactic sequences. A number of studies show that a printed text is easier to read the more closely its structures are related to those used by the reader in normal speech. This means that for the young child certain kinds of reading material must present a problem. Such sequences as 'look, look, see the elephant' do not come naturally off the tongue of the average five year old in everyday speech. Even the language of story books for young readers sometimes deviates appreciably from the speech patterns of those for whom they are intended. Since it also differs from the spoken language used in the various kinds of literature for older children and adults one wonders what is the justification for it. It certainly fails to provide a useful graphic representation of an acceptable linguistic style. Research has shown that pre-school children use a surprisingly wide range of sentence structures in their spoken language. Reading material which presents children with this unreal language therefore lacks predictability and prevents them from making use of the sequential probability in linguistic structure. The result is that they have to depend too much on a laboured phonic approach to unfamiliar words. This important issue is taken up in more detail in the next chapter.

6.35 The anticipation of sequences is also called into play at the level of meaning. It has been estimated (9) that the most common 500 words in English share between them some 14,050 meanings. The ambiguity of letters considered in isolation is almost trivial when compared to the ambiguity of isolated words. Only by using the surrounding sequences can the reader identify which of the many possible meanings an author intends in a given passage. The most effective teaching of reading, therefore, is that which gives the pupil the various skills he needs to make fullest possible use of context cues in searching for meaning. The habit of responding sensitively to context in order to detect significant nuances of meaning is not one that can be acquired quite simply in the early stages of reading. It develops over a lifetime; and an important condition for its development is that in every reading task there should be an incentive to read with this kind of alertness. Anyone who has discussed with children the meaning of something they have been reading will appreciate how little use many of them make of the rich contextual cues that are available. This is scarcely surprising, as they are rarely taught how to do it. The teaching of reading virtually ceases once the child can read aloud with reasonable accuracy at a reasonable speed. Yet to discontinue instruction at this point is rather like halting the training of a pianist once he can play the scales and a few elementary tunes.

6.36 The point needs no labouring that the intermediate skills are important in word attack, in fluent reading, and in comprehension. What is less often realised is the very great potency of these skills when they operate in combination. Research has shown that the reading vocabulary of most children begins to widen considerably between the ages of seven and nine. This rapid acceleration almost certainly depends to a large extent on the effective development of intermediate skills, for it is these which enable the reader to cope with new words and new meanings in context. Dictionary skills are, of course, extremely important, but this enrichment of the reading vocabulary has as its primary source the ability we have been describing. Failure to develop this competence may partly explain the difficulties experienced by retarded readers in making progress beyond a reading age of eight or nine. Much 'remedial' work has emphasised word building at the expense of reading for meaning, and this may be at least partly to blame.

6.37 If a child rapidly develops a reading vocabulary this does not in itself mean that he is developing as an effective reader. He may become skilled in recognising in print words he has previously heard, but he may still be unable to respond adequately to larger units of meaning in any given text. We are not able to recommend exclusively any one approach to helping children to respond to context cues. However, it may be useful to illustrate the possibilities with one instance. In recent years there has been a growing interest in the use of cloze procedure for developing the intermediate skills. Cloze procedure is the use of a piece of writing in which certain words have been deleted, and the pupil has to make the maximum possible use of the context cues available in predicting the missing words. There is no single 'right' answer in each case, as some of the words may be as suitable as the author's own words, or for that matter even more suitable. Merely filling in missing words as a routine exercise appears to have no measurable effect, but animated discussion on a piece of writing of real interest is a different matter entirely. When a cloze test is being prepared the deletions may be random, or they may be confined to certain parts of speech or words which are 'cued' in different ways. We would prefer to see approaches such as these as part of the teacher's repertoire rather than as part of a structured programme. They can be used with a whole class, a group, or an individual pupil as and when the need arises.

6.38 The great importance of anticipation has implications for the choice of words to be used in early reading material. If these are of the 'Can Dan Fan Nan?' variety then the pattern of phonic expectancies built up at this critical period will be very much at odds with the spelling patterns of English. (The sentence quoted is an actual example from a book which follows the 'linguistic' method of teaching reading). Several researchers have been highly critical of reading schemes in which the form of vocabulary control was to select similar words with regular spellings, a principle advocated by earlier linguists such as Bloomfield and Fries. One such critic (10) reviewed a number of studies which compared the early 'linguistic' reading schemes with basal* readers, and he concluded that the former 'tend to produce inferior oral reading in both rate and accuracy'. This judgement received some indirect support from a study (11) which showed that the learning of words with minimal contrasts, e.g. 'rat', 'fat', 'hat', made it more difficult for the child to learn more complex words at a later stage. It was better to give the child more varied and complex words from the beginning, rather than restrict him to simple words with regular spellings. Other studies have shown that words which differ markedly from one another are more easily learned than those between which there is little contrast. One important advantage of not restricting vocabulary on a phonic or linguistic basis is that words can then be selected for their familiarity to the child, their interest, and their richness of meaning, all of which make them more easily remembered. If, in addition, his attention is drawn to morphemes, as we suggested in paragraph 6.24, the child who has met the words hear, hears, and hearing will have no difficulty with heard in a sentence such as 'Peter had not heard his mother calling'. His anticipations based on the context, combined with his response to the meaning of hear, will give him a very good chance of getting quickly to the meaning and hence to the pronunciation of the word heard. In conclusion, we must again lay emphasis on the need for the child to learn phoneme-grapheme relationships within the context of actual reading. To be able to do this he must develop an adequate sight vocabulary at an early stage, a subject discussed in the next chapter.

*The term 'basal readers' is used in the USA to refer to books which are graded in terms of difficulty and which form a reading scheme. The scheme will also include graded workbooks, teaching materials of various kinds, and one or more teachers' manuals.

6.40 At the level of literal comprehension (para 8.15) the reader identifies material explicitly set down in the text that happens to relate to his purpose. To do this he has to be able to select significant detail and identify main ideas, which may be contained in descriptive or explanatory sequences, comparisons, and summarising statements. When he has achieved a grasp of the literal content the reader is then in a position to analyse, paraphrase, synthesise, and summarise it in whatever way suits his reading purpose. In varying degrees of difficulty this capacity for reorganisation is required of the child throughout his school work. The quality of comprehension at this literal level, however, is often very low. In project work, for example, many pupils have a strong tendency merely to string together sentences culled from various sources or, at best, to make only minor modifications in an attempt to paraphrase. This is clearly of very little value to them. If they are to reorganise and relate ideas, rather than confine their attention to specific phrases and sentences, they need to be taught particular skills. These include the ability to make well-structured notes, to integrate notes from various sources, and to use flow diagram techniques or other kinds of model.

6.41 When the reader goes beyond what is explicitly stated he is engaged in inferential comprehension (para 8.16). Here he interprets the significance of ideas or thoughts which might conceivably have been included or made explicit, but were not. This includes the interpretation of figurative language and the prediction of outcomes. Thus the reader is reading not only between the lines, but beyond the lines. Generally speaking this kind of reading receives scant attention in school except in the treatment of literature. The pupil has little incentive to respond sensitively to inferences in his reading in other curriculum areas. Another aspect of comprehension which receives little attention is evaluation (para 8.17), where the reader applies 'truth' tests to the material. He may, for example, need to evaluate the internal logic of a passage, and its authenticity, adequacy, and appropriateness. In spite of the very obvious shortcomings of much of the printed material placed before children, there is little systematic teaching designed to show them how to approach it in this critical fashion. The affective or aesthetic equivalent of evaluation is appreciation, which is usually regarded, in the secondary school at least, as being exclusively the province of the English specialist. It is discussed at length in the chapter devoted to the place of literature. Even in functional reading, however, there is everything to be said for developing the pupil's ability to respond to an author's use of language - to imagery, style, and structure. This should be seen as an integral feature in the development of other aspects of comprehension.

6.42 An important attribute of the competent reader is the ability to apply flexible reading strategies, according to his purpose and the nature of the material. Unfortunately, if most of their reading is of the single speed kind, children will be habituated to becoming single speed readers. The danger is then that the only technique they will ever use is inflexible, one pace, line by line reading. Flexibility should be acquired at school and should be exercised throughout the curriculum. The only way in which it can be effectively developed is for the pupil to encounter a full range of reading tasks which make demands on the relevant skills.

6.43 Dealing efficiently with information must now be recognised as one of the major problems in modern society. It has been estimated that in the United States, for example, one third of the national product is currently used in producing information. Despite the growth of other media, the vast bulk of all information is recorded in printed form. So far in this chapter we have concentrated on the ability to read the various kinds of printed material in which this information is recorded. But dealing with information in the mass presents a broader set of problems of which the reading process itself is simply one element. It becomes increasingly necessary for a person not only to be able to cope with print efficiently, but to organise his own use of it. This means that he must be able to identify his own information needs, a much less simple matter than it sounds. He must then know the sources which will answer to them, judging the value of these from a wide range of material and selecting the limited amount which will serve him best. The first implication of this is that children should have extensive experience in defining their own purposes. They need to become skilled in working out exactly what questions they should seek to answer by reading. The second is that they should be given the opportunity to explore many different kinds of printed media, and learn how to obtain what they need. Pupils should be led to confidence in the use of bibliographical tools and in tapping sources of information in the community at large, and as the sources of information continue to change and multiply the teacher must be prepared to learn alongside his pupil.

6.44 Many individuals develop the various comprehension skills to a high level with very little guidance, but the majority need a great deal of positive help. Research (see paragraph 2.2) strongly suggests that as many as one third of the population may be incompetent in the kinds of reading comprehension to which we have referred. We also suspect that many people merely reorganise what they are reading in ways that confirm existing ideas and prejudices. If reading comprehension is to be significantly improved, then, an important principle has to be established. Every subject teacher in the secondary school must assume responsibility for developing all those kinds of skill that are needed by his pupils to read intelligently the material he presents to them. The ability to do this must therefore be seen as an essential element in the professional competence of the subject specialist, a principle elaborated in Chapter 8. It is a further reinforcement of our argument that reading should receive increasing attention from all teachers at each successive stage of education. We suggest later in the Report that each primary school should have a teacher responsible for advising his colleagues in language and reading, and this is one important area in which that teacher's contribution would be helpful. In the secondary school it would be an aspect of the policy of language across the curriculum, to which Chapter 12 is devoted.

References

1. Disick H In Spite of the Alphabet Chatto and Windus: 1965.

2. Fries CC Linguistics and Reading Holt, Rinehart and Winston: 1962.

3. Goodman KS Behind the Eye: What Happens in Reading in Reading: Process and Program Urbana, Illinois: National Council of Teachers of English: 1970.

4. Gray WS The Teaching of Reading and Writing: An International Survey Alfred: Paris-UNESCO: 1956.

5. Taylor EA The Spans: Perception, Apprehension and Recognition American Journal of Ophthalmology, Vol. 44: 1957.

6. Smith F Understanding Reading Holt, Rinehart and Winston: 1971.

7. Daniels JC and Diack H Progress in Reading 1956.

8. Berdiansky B, Cronnel B, and Koehler J Spelling - Sound Relations and Primary Form - Class Descriptions for Speech - Comprehension Vocabularies of 6-9 Year Olds South West Regional Laboratory for Educational Research and Development, Technical Report, No. 15 (1969). Cited in Smith F Understanding Reading Holt, Rinehart and Winston: 1971.

9. Fries CC Linguistics and Reading Holt, Rinehart and Winston: 1962.

10. Spache GP and EB Reading in the Elementary School Allyn and Bacon: Boston: 1973.

11. Levin H and Watson J The Learning of Variable Grapheme to Phoneme Correspondence: Variations in the Initial Consonant Position Cornell University Cooperative Research Project No. 639: 1963.