![]()
Part one
The state of the art
![]()
Chapter 1
The state of the art and some definitions
It is generally agreed that we do not understand the process of reading, or what happens when a child does or does not learn to read. (Labov, 1970)
⌠the incredibly confused and inconclusive state of reading research ⌠(Smith, 1973, p. 5)
Very little of the great mass published on the subject (of the teaching of reading) is significant in contributing to a theory of literacy. Indeed much of it reflects the absence of a theory, and the absence of any awareness of the need for a theory. (Mackay et al., 1970, p.78)
One reason why the literature on reading is so vast and unintegrated is that topics have been approached from different directions from within different disciplines, including psychology, education and linguistics. Often these approaches have been largely self-contained, making little reference to work within other approaches, and, in fact, putting forward contradictory definitions of reading and literacy. Furthermore, research on reading has been dominated by experimental psychology, which has seen reading as primarily a perceptual process. Until relatively recently, reading has been regarded only peripherally as a process of handling written language or as an activity with particular social functions. Certainly, the vast majority of research has concentrated on the psychological processes of reading in the individual reader, and therefore on the internal relations between perceptual processes, orthographic systems and, to a lesser extent, the readerâs knowledge of his own language. But it has neglected the relation of reading to writing, the place written language plays in different societies, and so on.
As a result, we know quite a lot about reading as a psychological process, although findings tend to remain unintegrated, sometimes contradictory, and often unrelated to linguistic and sociological approaches to reading and writing. But we know relatively little about reading and writing as linguistic processes, and even less about the social functions of reading and writing.
Research on reading has been carried on in a concentrated fashion since the beginning of the twentieth century. Compulsory education, with the minimum requirement of literacy for all, was introduced by the Education Acts of 1870 in England and Wales and of 1872 in Scotland, and systematic research was underway by the 1900s. The discovery that the eyes move in a jerky fashion during reading, fixing on a span of words or letters, then moving on rapidly to another fixation point, was made as early as 1879 by Javal. The term âcongenital word blindnessâ was first used by Morgan (1896) in an article in the British Medical Journal. And as early as 1897, Pillsbury had shown the importance of expectation on the perception of words: he presented words with deliberate typographical errors for very short periods by tachistoscope. Subjects were often certain they saw letters which were not present. The Armed Forces revealed high levels of illiteracy during the First World War and this provided fresh impetus for research. And National Reading Surveys have been carried out since 1940 (Morris, 1972). But despite seventy-five years or more of research, there is still nothing approaching a coherent theory. In a recent major book, Smith (1973) declares simply that reading research is âincredibly confused and inconclusiveâ. A fashionable disclaimer at present is that we have not learned much more about the psychology of reading than is set out in Hueyâs classic book of 1908 (e.g., see Gibson, 1972, in Kavanagh and Maddingley, 1972; Kolers, 1968, cited by Gudschinsky, 1976, p. 9). Often, in fact, researchers appear to have despaired and relapsed into a mystic belief that it is all too complex to describe and involves the whole man. One finds unhelpful statements such as: âPerhaps reading, like mystery, can only be described and evoked.⌠Reading must engage the total organismâ (Jenkinson, 1969, p. 107). It is probably unfair to pick out this particular quote, as any one of so many might have been quoted in its place.
It will be useful to begin with a simple list of some of the potential confusions that will have to be borne in mind as the argument proceeds. Many of these distinctions appear obvious enough once they are pointed out, but they are often not made explicit in the literature.
1.1 Some potential confusions
Definitions of reading and writing
There is, first of all, still no general agreement on what is meant by reading and literacy. (See 1.2 below.) Most collections of articles on reading contain papers with titles such as âWhat is reading?â, The nature of readingâ, or even âReading: is there such a thing?â (see, for example, Melnik and Merritt, 1972b). The basic debate has been between those who hold that reading means essentially the âmechanicsâ of reading, that is, the ability to decode written words into spoken words; and those who maintain that reading essentially involves understanding. There has been no agreement on this since Thorndike raised the issue very clearly in a now classic article published in 1917. In this article he argued that reading is âunderstanding the meaning of printed wordsâ, and he himself took the extreme view that reading is âreasoningâ and that âunderstanding a paragraph is like solving a problem in mathematicsâ. Whether one agrees with Thorndike or not, he did pose the alternatives very clearly; but there is still no consensus on the answer.
In order to begin to decide this question in a principled way, we must know whether writing systems do, in fact, relate written symbols to sounds, and, if so, how. Incredibly, however, one thing which is missing from many discussions of reading in English is a sophisticated understanding of how the English spelling system words. It has been suggested at intervals since the sixteenth century that English spelling and its âirregularitiesâ are a cause of reading failure. And various regularized writing systems have been proposed, either as permanent reforms or as aids in the initial stages of teaching reading. The most recent and best known in the latter category is the Initial Teaching Alphabet (ITA). For various reasons which do not concern us directly here, linguists have only recently shown much interest in English spelling. But there are now several illuminating and important studies of English spelling (including Albrow, 1972; C. Chomsky, 1970;Venezky, 1970), and this work will have to be discussed in detail below (see Chapter 3). It is essential that such an understanding should form a basis of a discussion of reading, because one can never understand how something is learned, without a proper understanding of what is learned. Neither, of course, can one hope to understand the causes of reading failure, until one knows what it is that a child has failed to do.
This lacuna is part of a more general failure, in much of the experimental psychological literature, to regard reading and writing as linguistic processes. Reading has often been seen predominantly as a matter of visual processing, involving characteristic eye-movements, perceptual span, letter shapes, word gestalts, and so on. As a result, it has often been ignored that what people read is linguistically organized and meaningful material.
Another confusion, potentially right at the heart of the matter, is between reading and writing. It is often assumed tacitly that reading and writing are symmetrical mirror-images of each other, and that whatever is said of one is, mutatis mutandis, true of the other. But reading and writing are not symmetrically related (Smith, 1973, pp. 117ff.). For example: one may be able to read a word at sight and understand it, but not be able to spell it correctly or use it appropriately. In general, it is always easier passively to recognize a shape (letter, face, etc.) than to reproduce it actively. Readers do not therefore have to know every word they come across, but writers do; even, or perhaps especially, fluent readers have a much larger sight vocabulary than an active vocabulary that they themselves use. Writing involves motor skills which may be an added problem to beginners who have not mastered them; and so on. Also, reading and writing are not functionally symmetrical: many people read a great deal and rarely write anything; only a few people write much; and very few people, if any, write a lot but read little. In general, if someone can write, this implies that he can read; but to say someone can read, does not imply that he can write. I can, for example, read German printed in Gothic script without any difficulty, but I am quite unable to write Gothic script.
Any writing system has to compromise between the requirements of writers and readers. In terms of the greatest good for the greatest number, the needs of readers ought to take precedence, since there are many more readers than writers, and the majority of people who do much writing are professionals, and ought therefore to be able to tackle a few more problems. This principle is maintained in ITA, which is explicitly biased in favour of readers, so that teachers are not even required to be able to write it, except for notices in the classroom (Pitman and St John, 1969, pp. 138â9). Similarly, the gap between a childâs ability to read a word and write it is recognized by the Breakthrough to Literacy materials (Mackay et al., 1970), which provide children with printed words on cards which can be slotted into a sentence maker. The child can therefore create new sentences without being held up by the inability to spell individual words, or, more simply, by the lack of manual dexterity in using a pencil.
Other potential identifications are not clear either. For example, is it the case that listening and reading comprehension are essentially similar and differ only in the input medium, aural versus visual? (The answer to this question would affect oneâs definition of reading.) Are written and spoken language essentially similar, in lexis and syntax, or do they differ to an extent which could cause problems for learners? There is no real consensus, in fact, about the relations between reading, writing, speaking, listening and comprehension.
Eclectic theories
One crucial source of confusion is between theories of reading (what goes on in the head) and instructional techniques (what goes on in classrooms): although any teacher knows that what he teaches is only indirectly related to what pupils learn! A theory of reading may, of course, suggest a teaching strategy. But partial theories proliferate, and so do techniques, not to say gimmicks. In any case, it is usually impossible to show that a particular technique causes children to learn. For example, the phonic method (technique) clearly does work for most children. But this does not prove that children read phonically (theory), that is, by decoding from letters to sounds. They may, for example, learn by gaining confidence that the system is phonic and regular, proceed on this assumption, and end up by learning and reading in some other way, say in whole words and phrases. This is, after all, how fluent adults end up reading. (Cf. Smith, 1973, p. 6.) In fact, any method appears to work with most children. The consensus view at present appears to be that exclusive reliance on any single technique is a mistake. After waves of enthusiasm and disillusionment for different approaches to the teaching of reading, the consensus now seems to be that there is no single best way to teach reading, and one comes across regular pleas for an eclectic approach. (E.g., see Rauch, 1968;Goodacre, 1971 ; Stauffer, 1971.) But the use of terms such as âeclectic approachâ or âmixed methodâ amounts to a tacit admission that there is no coherent theory that works, so that practitioners should not put all their eggs in one basket and may as well proceed in an ad hoc manner.
This same eclecticism necessarily pervades the literature on reading failure, where the overwhelming impression is that each individual case must be taken on its merits. Every logically conceivable type of explanation has been proposed (see 7.1). There are only a restricted number of possibilities: that there is something wrong with the learner, either medical (e.g., âword blindnessâ) or psychological (e.g., emotional disturbance), or with the learnerâs language (a language âdeficitâ view); that there is something wrong with the learnerâs family background (a social pathology or social deprivation view); that there is something wrong with the teacher (e.g., the learner is not being sufficiently motivated); that there is something wrong with the method (hence the phonic-versus-whole word debate); that there is something wrong with the materials (e.g., they do not match the learnerâs interests, experience, etc.); or that there is something wrong with the medium, that is with the spelling system (hence ITA and similar schemes). There is, logically, nowhere else the explanation could lie: it has to be in the pupil, the teacher, the materials, the method, the medium or the pupilâs background. Most likely there is some truth in all of these explanations, and that different pupils have difficulties deriving from a mixture of causes. That is, we have no coherent theory.
There is a simple but important overall worry. Despite all the research over the past eighty years or so, the vast body of findings still does not satisfactorily explain cases of reading failure. And the suspicion remains that the psychological and psycholinguistic factors, on which the bulk of research has been done, are quite easily swamped by much more powerful social and cultural effects such as the learnerâs motivation, the value which the community places on literacy or on education as a whole, or simply the skill of individual teachers.
Children and adults
Related to the confusion between how readers actually read (theory) and how they ought to be taught to read (technique) is a confusion between children and adults, or between fluent readers and learners. Many studies have not distinguished sharply enough between what fluent readers do, and what children learning to read do. But these may be quite different. As one writer put it, the best way of learning to dive may not be to start from the ten-metre board. The distinction is, of course, quite clear in many places: for example, it is the basic rationale behind ITA. Nevertheless, it may be that children are confused with adults in quite fundamental ways. It is now currently fashionable to assert that children of four or five years have awesome mastery of complex aspects of their native language; and this is true. The knowledge which an average five-year-old has of the phonology, morphology and syntax of English is so complex that linguists have not yet succeeded in describing and accounting for it in any satisfactory way. But there are differences between the phonological and morphological knowledge of five-year-olds and the knowledge of educated adults, and those differences may be crucial to someone learning the English writing system in particular.
For example, to understand fully how English spelling works, readers have to appreciate the identity between the roots of semantically related pairs of words which differ in sound, such as medicine/medical, paradigm/paradigmatic, bomb/bombardier. But many of these pairs are likely to be known only by educated speakers who are already fluent readers (see 3.6). Also, there is considerable debate about how and when young children acquire the ability to segment the speech continuum into phonemes (minimal sound units, see 2.6). And it has been suggested that syllables are a more salient linguistic feature for children. This would crucially affect our knowledge of the best methods of teaching reading, and also our theories about the relation between spoken and written language. Therefore, statements such as âat the age of entering school, children have acquired a mastery of the com...
