Learning To Read

Basic Research and Its Implications

Laurence Rieben, Charles A. Perfetti

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781135436452
• Edition: 1
• Topic: Education
• Subtopic: Education General

How does a young child begin to make sense out of squiggles on a page? Is learning to read a process of extending already acquired language abilities to print? What comprises this extension? How children learn to read, and especially how children are taught to read, are problems of sustained scientific interest and enduring pedagogical controversy. This volume presents conceptual and theoretical analyses of learning to read, research on the very beginning processes of learning to read, as well as research on phonological abilities and on children who have problems learning to read. In so doing, it reflects the important discovery that learning to read requires mastering the system by which print encodes the language. The editors hope that some of the work offered in this text will influence future research questions and will make a difference in the way instructional issues are formulated.

I

Principles and Theories

1

Phonology and Beginning Reading : A Tutorial

Isabelle Y. Liberman

Donald Shankweiler

University of Connecticut, and Haskins Laboratories

Introduction

Our research and that of many others in the field for the past 15 years has persuaded us that most problems in learning to read and write stem from deficits in language-related skills, not from deficiencies in the visual or auditory systems. Moreover, we have come to believe we can pinpoint more specifically where in language the difficulties might most reasonably be found. Early in our research (Liberman, 1971, 1973) we were motivated to begin by looking at the phonology: Because an alphabetic orthography represents the phonology, that seemed the right place to start. The results of research have, we think, justified that assumption.

In the beginning, we concentrated most of our attention on deficiencies in awareness of the phonology, that is, on metalinguistic abilities. We review those findings briefly here. Later, we came to consider that metalinguistic deficits may be only the symptoms of a more general underlying deficiency in the phonological domain. Research in verbal short-term memory provided some of the earliest evidence in this connection. Recently, corroborative evidence has come from other sources—for example, experiments in speech perception, speech production, and naming, and the reading of congenitally deaf subjects. We highlight the salient evidence in all these areas, as space permits. We then look into the implications for educational practice.

But first, we wish to comment on the terminology that we all use, often rather loosely. In our view, we cannot discuss the problems of phonology and reading disability reasonably unless we are clear about exactly how we are using certain terms.

Some Clarification of Terminology

“Letter-to-Sound Correspondences”: A Misnomer

It is common practice for investigators to remark on the relatively poor match between the orthography and the “sounds” of English, and to blame reading difficulties on the consequent difficulty of learning “letter-to-sound correspondences.” Leaving aside the question of whether there is indeed a poor match (an interesting question, but a topic for another paper), we suggest that it might clarify our thinking about what is required in the reading task if we used the term phonemes in this connection rather than sounds. The correspondences under consideration, after all, are not really correspondences with “sounds,” but with the phonology of the language, a very different matter, and a distinction of great practical importance. The distinction has practical importance for at least two reasons. First, it underscores the point that the problem is more than one of simply getting the child to learn to associate a visual shape and a sound—even mentally retarded children can be taught quite easily to do that (House, Hanley, & Magid, 1980). Instead, the problem is getting the child to understand how the visual shapes relate to the phonology of the word. There are compelling reasons why that is hard to do, as we see later. The second reason the distinction has practical importance is that it shifts the reading problem away from shape and sound or the sensory modalities and returns it to the phonology and the language modality, where the difficulty so often is found.

The Phonology: Abstract Categories, Not Sounds

But what exactly is the phonology? The phonology is the system of representation that all members of the human race use for producing and storing an indefinitely large number of words by means of a few dozen abstract, meaningless elements, the phonemes. These phonemic elements are, of course, normally conveyed by sound but are not themselves sounds. Instead, they are abstract categories of language.

Perhaps the best way to clarify what the phonology is all about is to imagine what language would be like if there were no phonology. In that case, each word in the language would have to be represented by a separate signal—for example, a sound—that differed holistically from the signals for all other words. The obvious consequence would be that the number of words could be no larger than the number of holistically different signals a person can efficiently produce and perceive. Of course, we don’t know precisely what that number is, but surely it must be small, especially in the case of speech, by comparison with the hundreds of thousands of words that a language commonly comprises.

What a phonology does for us, then, is to provide the basis for constructing a large and ever expandable set of words out of only two or three dozen abstract signal elements. These abstract signal elements, the phonemes, are not themselves sounds. They are represented by sounds, but only after complex transformations.

Words: Basically Phonological Structures

So, in dealing with words, whether written or spoken, we are dealing basically with phonological structures. Whatever else words may be, they are always phonological structures. If we have perceived or produced a word, whether in speech or in reading, we have in fact engaged a phonological structure. If we have misperceived or misproduced a word, we have in fact engaged the wrong phonological structure.

But why, then, should reading words be difficult in an alphabetic orthography, considering that the alphabetic transcription represents, even if only approximately, the phonological structure that the reader must grasp?

Phonological Processing in Listening and Reading

Listening to Speech

To understand the problem one faces when required to read a word, we must first consider how the word is perceived when it is spoken. As we have said, the word is formed by a phonological structure, so when the word is perceived, it is that structure that is accessed. But the speaker of the word did not produce the phonological units one at a time, each in its turn—that is to say, the speaker did not spell the word. Instead, in producing the syllable “ba,” for example, the speaker assigned the consonant we know as “B” to the lips, and the vowel we know as “A” to a shaping of the tongue, and then produced the two elements at pretty much the same time. Note that in normal speech, then, we do not say “buh-ah,” or “B,” “A,” but “ba.”

The advantageous result of such coarticulation of speech sounds is that speech can proceed at a satisfactory pace—at a pace indeed at which it can be understood (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967). Can you imagine trying to understand speech if it were spelled out to you letter by painful letter? So coarticulation is certainly advantageous for the perception of speech. But a further result of coarticulation, and a much less advantageous one for the would-be reader, is that there is, inevitably, no neat correspondence between the underlying phonological structure and the sound that comes to the ears. Thus, though the word “bag,” for example, has three phonological units, and, correspondingly, three letters in print, it has only one pulse of sound: The three elements of the underlying phonological structure—the three phonemes—have been thoroughly overlapped and merged into that one sound—“bag.”

The first question that needs to be answered, then, is how do listeners recover the discrete units of the phonological structure from that seamless sound and thereby manage to make contact with the exact word as it must be stored in their heads? The short, and, for our purposes here, sufficient answer is that in listening to speech, the phonological segments are recovered from the sound by processes that are deeply built into that aspect of our biology that makes us capable of language. Therefore, in listening we are no more aware of the processes by which we arrive at the word than we are aware of how we perform other deeply biological processes, as for example the way we compute the location of a sound from the difference in time at which it arrives at the two ears.

A Difference Between Listening and Reading

Reading is different from listening in that it is, in some significant ways, a secondary, less natural use of language—part discovery, part invention. It follows, then, that even though its processes must make contact with those of the natural and primary language system, special skills will be required if the proper contact is to be made. The point of that contact is the word, which is, of course, represented in the print by a transcription of the phonological structure. But that transcription will make sense to beginning readers only if they understand that the transcription has the same number and sequence of units as the spoken word.

Beginners can understand, and properly take advantage of, the fact that the printed word bag has three letters, only if they are aware that the spoken word “bag,” with which they are already quite familiar, is divisible into three segments. They will probably not know that spontaneously, because, as we have said, there is only one segment of sound, not three, and because the processes of speech perception that recover the phonological structure are automatic and quite unconscious. Moreover, it may be somewhat difficult to teach them what they need to know because, given the overlap of phonological information, the merging by coarticulation that characterizes the spoken word, there is no way to produce many of the consonant segments in isolation. The teacher can try, as is commonly done in the so-called phonics method, to “sound out” our word “bag,” but the learner must still make a formidable leap from that to the spoken word. Fortunately, the phonics method need not, of course, be limited to the “sounding out” approach; it can be expanded to include many other techniques to achieve the intended goal (see Liberman, Shankweiler, Camp, Blachman, & Werfelman, 1980).

In any case, the phonics approach is, in our view, vastly preferable to a currently popular procedure, dubbed by its creators the “psycholinguistic guessing game” (Goodman, 1976). In this widely used procedure and its offshoot, the so-called “whole language” method, the students are encouraged to bother as little as possible with the phonological structure as it is represented by the graphemes. Instead, they are encouraged to use their “natural language processes” (which apparently do not include phonological processes!) to reach the meaning as quickly as possible. This goal is to be achieved by guessing from a few known words in the message and the context. We have more to say about that later.

But even with the much preferable phonics method of instruction,it can be hard to get some children to become aware of the underlying phonological structure of the spoken word and, accordingly, to see why it makes sense to represent the monosyllable “bag” with three letters.

Awareness of Phonological Structure

Development of Phonological Awareness

Given all that we have said, we might well expect young children to have difficulty in becoming aware of the phonological structure and, in fact, they do. Some 15 years ago we began to examine developmental trends in phonological awareness by testing the ability of young children to segment words into their constituent elements (Liberman, Shankweiler, Fischer, & Carter, 1974). We found that normal preschool children performed rather poorly. We learned further that of the two types of sublexical units—the syllables and the phonemes—the phoneme, which happens to be the unit of our alphabetic writing system, presented the greater difficulty by far.

It was clear from these results that awareness of phoneme segments, the basic units of the alphabetic orthography, is initially harder to achieve than awareness of syllable segments and develops later, if at all. More relevant to our present purposes, it was also apparent that a large number of children have not attained either level of understanding of linguistic structure, syllable or phoneme, even at the end of a full year in school. They are the ones we need to worry about, we believe, because they are the ones who are deficient in the linguistic awareness that may provide entry into the alphabetic system. There is much evidence now that awareness of linguistic structure— an awareness that so many young children lack—may be important for the acquisition of reading and spelling.

Awareness of Phonological Structure and Literacy

Much evidence is now available to suggest that metalinguistic awareness of the phonological constituents of words is germane to the acquisition of alphabetic literacy. This evidence comes from studies that have shown that this awareness is predictive of reading success in young children. In English, there are, to name a few, studies by Blachman (1983), Bradley and Bryant (1983), Fox and Routh (1980), Goldstein (1976), Helfgott (1976), and Trei-man and Baron (1981). Their findings have been supported by studies in Swedish by Lundberg and associates (1980) and Magnusson and Naucler (1987), in Spanish by de Manrique and Gramigna (1984), in French by Bertelson’s laboratory in Belgium (Alegria, Pignot, & Morais, 1982), and recently in Italian by Cossu and associates (Cossu, Shankweiler, Liberman, Tola, & Katz, 1988).

Effect of Training in Awareness

Not only is there abundant research showing that phonological awareness is predictive of future literacy, there is now growing evidence that training in awareness may actually help the novice to learn to read and write. It is of special interest, then, to find recent training studies showing that phonological awareness can be trained even in preschool (Content, Morais, Alegria, & Bertelson, 1982; Olofsson & Lundberg, 1983; Vellutino, 1985). Early evidence for the effect of such training comes from the landmark study of Bradley and Bryant (1983). In the first of a pair of experiments, they found high correlations between preschoolers’ phonological awareness as measured by rhyming tasks and the children’s reading and spelling scores several years later. In the second experiment, they found that the children with initially low levels of phonological awareness who were subsequently trained in phonological categorization were superior to a semantically trained group on standardized tests of reading and spelling. Those trained with grapheme correspondences in addition to the phonological training were even more successful.

We see then that phonological awareness can be trained in very young children, and that this training can have beneficial effects on the children’s future progress in learning to read and spell. There is now a broad data base relating reading disabilities to deficiencies in the awareness of the phonology, including studies of people of a wide range of ages, of many language communities, and a full range of cultural and economic backgrounds from inner city and rural poor to suburban affluent.

Most of this research has concentrated on deficiencies in metalinguistic awareness. Lately, as we intimated earlier, evidence of various kinds has led us to consider seriously the possibility that the problems of the disabled reader may not be limited to metalinguistic deficits but may reflect a more general deficiency in the phonological domain.

Other Deficiencies in the Phonological Domain

Evidence from Short-Term Memory and Language Comprehension

Because short-term memory depends on the ability to gain access to phonological structure and to use it to hold linguistic information (Conrad, 1964; Liberman, Mattingly, & Turvey, 1972), we might expect people who have these phonological deficiencies to show various limitations on verbal short-term memory tasks. This expectation is amply borne out. The research literature contains many reports that young children who are poor readers are deficient at holding verbal materials in short-term memory. Typically they retain fewer items from a set of fixed size than age-matched good readers (see Liberman & Shankweiler, 1985; Wagner & Torgesen, 1987)...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Contents
• Introduction
• Acknowledgments
• I Principles and Theories
• 1 Phonology and Beginning Reading A Tutorial
• 2 Changing Models of Reading and Reading Acquisition
• 3 Representations and Awareness in the Acquisition of Reading Competence
• II Starting to Learn to Read
• 4 The First Stages of Word Recognition
• 5 Learning to Read and Spell Words
• 6 Experimental Analysis of the Child's Discovery of the Alphabetic Principle
• 7 Individual Differences and Lexical Representations How Five 6-Year-Old Children Search for and Copy Words
• III Phonological Abilities
• 8 Phonological Awareness and Literacy Acquisition
• 9 Phonological Abilities Effective Predictors of Future Reading Ability
• 10 Segmental Analysis and Reading Acquisition
• 11 The Role of Intrasyllabic Units in Learning to Read
• IV Reading Skill and Reading Problems
• 12 Learning (and Not Learning) To Read A Developmental Framework
• 13 Word-Identification Strategies in a Picture Context Comparisons between “Good” and “Poor” Readers
• 14 The Effects of Instructional Bias on Word Identification
• Author Index
• Subject Index