Contemporary Perspectives on Reading and Spelling

Clare Wood, Vincent Connelly

• Publisher: Routledge
• Year: 2009
• ISBN:
  9781134004232
• Edition: 1
• Topic: Education
• Subtopic: Education General

With contributions from leading international researchers, Contemporary Perspectives on Reading and Spelling offers a critique of current thinking on the research literature into reading, reading comprehension and writing. Each paper in this volume provides an account of empirical research that challenges aspects of accepted models and widely accepted theories about reading and spelling.

This book develops the argument for a need to incorporate less widely cited research into popular accounts of written language development and disability, challenging the idea that the development of a universal theory of written language development is attainable. The arguments within the book are explored in three parts:

• 
  
  

• overarching debates in reading and spelling
• reading and spelling across languages
• written language difficulties and approaches to teaching.

Opening up the existing debates, and incorporating psychological theory and the politics surrounding the teaching and learning of reading and spelling, this edited collection offers some challenging points for reflection about how the discipline of psychology as a whole approaches the study of written language skills.

Highlighting ground-breaking new perspectives, this book forms essential reading for all researchers and practitioners with a focus on the development of reading and spelling skills.

Part 1
Overarching debates in reading and spelling

Chapter 1
Phonological awareness

Beyond phonemes

Clare Wood, Lesly Wade-Woolley and Andrew J. Holliman

Current models of reading development emphasise segmental phonological awareness as the basis for success in reading and spelling development. This chapter explores the possibility that suprasegmental phonology may have a contribution to make in explaining both the origins of segmental phonological awareness in beginning readers, but also polysyllabic word reading in more experienced readers. The chapter reviews the evidence suggesting a link between decoding skills and sensitivity to aspects of speech rhythm. These accounts build to a model, which explains the possible contribution of speech rhythm to reading development.

Phonological deficits and reading difficulties

One of the most frequently cited explanations of reading difficulties is the phonological deficit hypothesis (Stanovich, 1986; Stanovich & Siegel, 1994). According to this theory, developmental reading difficulties are characterised by poor or underspecified phonological representations. The quality of these representations are such that when children learn to read and write they experience delays in acquiring the letter-sound correspondences necessary for reading in an alphabetic language.

Although questions over the extent to which there is evidence of a causal link between phonological deficits and reading difficulties have been raised (Castles & Coltheart, 2004), there is extensive evidence that children with reading difficulties demonstrate deficits on tasks that have a phonological component relative to children of the same age, and in some instances, relative to younger, typically reading children who are at the same level of reading skill as they are (see Snowling, 2000, for review). The question, however, remains ‘Why do children with reading difficulties have phonological deficits?’

The answer to this question could be simply a case of individual differences. As with all learned abilities, some children are less able, less skilled, less fluent than their peers in a particular domain. In this sense, it is therefore possible that children with reading difficulties are just the unlucky ones in terms of phonological competence. However, there is the question of whether there are also deficits in the underlying layer of skills and abilities that developmentally precede or are associated with phonological skills. If we find no relationship between phonological deficits and other more fundamental skills, then this would suggest that phonological deficits are characteristic, and perhaps causal, of reading difficulties. However, phonological deficits may equally be the expression of more fundamental difficulties in cognitive development. If we could find the origins of the phonological deficit in skills that developmentally precede the development of segmental phonological awareness, this could expand possibilities for early identification and intervention.

The idea that if a child experiences difficulties perceiving speech this might impact on the quality of the phonological representations that the child goes on to form was explored by Wood and Terrell (1998). However, as they noted, the literature concerned with children’s speech perception is mixed (see McBride-Chang, 1995). While some studies reported that children with reading difficulties showed deficits in speech perception (Brady et al., 1983; De Weirdt, 1988; Freeman & Beasley, 1978; Godfrey et al., 1981; Hurford, 1991; Reed, 1989; Tallal, 1980; Werker & Tees, 1987), others did not (Snowling et al., 1986; Pennington et al., 1990). This has led Snowling (2000) to conclude that the phonological impairments of children with dyslexia are not the result of perceptual difficulties, but of difficulties in the encoding and retrieval of phonological representations.

However, a closer examination of the papers reviewed by McBride-Chang (1995) revealed that many of the papers reporting to assess speech perception were actually presenting children with tasks of phoneme discrimination. For instance, the speech perception tasks in De Weirdt (1988) and in Werker and Tees (1987) required children to discriminate between phonemes such as /pa/-/ta/ and /ba/-/da/ respectively. It was therefore not surprising that children with reading difficulties, known to experience difficulties with phonemic awareness in particular, were found to be deficient on such tasks. As a result, Wood and Terrell (1998) turned their attention to a specific component of speech perception that seemed to be less of a direct measure of phoneme discrimination: that of spoken word recognition.

Spoken word recognition and speech rhythm

Fluent speech contains no reliable acoustic cues that indicate where one word ends and the next word begins. The question of how infants come to break fluent speech down into component units of meaning was one that Wood and Terrell (1998) felt might hold the key to the quality of a child’s phonological representations. A review of the literature on spoken word recognition revealed one theory that offered a plausible account of how children may come to bootstrap their way into spoken word recognition from infancy. This was the model outlined by Cutler and Mehler (1993). They suggest that children are born with a periodicity bias that predisposes them to ‘tune in’ to the rhythmic characteristics of their first language. This information enables them to make initial attempts at segmentation as rhythmic information appears to provide a rough indicator of word boundaries. For example, English is a stress-timed language. This means that, roughly speaking, equal amounts of time elapse between the occurrence of stressed syllables in natural speech (this contrasts with syllabletimed languages, such as French, in which equal amounts of time elapse between syllables). Content words in English tend to begin with stressed (or ‘strong’) syllables. As a result, sensitivity to stressed syllables in speech would provide a child with a basis for segmenting speech into units of meaning.

However, a notable characteristic of linguistic stress is that phoneme identification appears to be much easier in stressed as opposed to unstressed syllables. For example, Chiat (1983), reporting on the speech errors of a phonologically delayed child, observed that the prosodic features of speech were, for that child at least, central to his ability to articulate the appropriate representation of the word. The 5-year-old child showed that although he had difficulties with segmental phonology, he was sensitive to the prosodic features of speech, and that sensitivity influenced his ability to produce the correct phonological form of the word he was trying to say. Chiat (1983: 292) argued that: ‘segmental features are not independent of prosodic features in lexical representations … Lexical representations are not, then, strings of phonemes on which stress is marked, but prosodic structures on which segmental features are specified’.

Wood and Terrell (1998) hypothesised that children who have poorly specified phonological representations may also show evidence of insensitivity to speech rhythm. To test this they matched a group of children who were delayed in their reading attainment to typically developing readers on either chronological age or reading age. They then tested their reading, spelling and phonological awareness, as well as their ability to accurately perceive time-compressed speech and their sensitivity to metrical stress. To assess metrical stress sensitivity, sentences were recorded and low pass filtered until no phonemes could be identified and only the intonation pattern of the utterance could be discerned. The children were then presented with two entirely different sentences. One of them shared the same metrical pattern as the original, whereas the other differed in rhythm from the filtered sentence by just one syllable. The children were asked to indicate which of the two sentences was ‘the same’ as the filtered sentence. It was found that the children who were delayed in their reading development were significantly worse than chronological age matched controls on both the speech perception and speech rhythm measures. Crucially, unlike the difference in speech perception performance, the difference in speech rhythm was found to exist after individual differences in vocabulary were controlled.

Since the publication of Wood and Terrell (1998) there has been increased interest in the possible contribution of prosodic processing to reading acquisition and phonological decoding. It should be noted that the contribution of prosodic sensitivity to reading comprehension is well established (e.g. see Kuhn & Stahl, 2003). However, Wood and Terrell’s study signalled a possible contribution of prosodic sensitivity to word reading ability (decoding) that had not been found previously.

Some predictions

If speech rhythm and related aspects of prosodic sensitivity are important for reading, then a number of research predictions should be supported by the data obtained by studies in this area.

1 Individuals with reading difficulties will be insensitive to speech rhythm relative to typically developing readers.

There should be clear evidence of a deficit in sensitivity to, or processing of, or representation of, speech rhythm in individuals with reading difficulties. The question of whether such a deficit is specific to individuals with reading difficulties or simply indicative of a developmental delay also needs consideration.

2 Speech rhythm sensitivity should be associated with reading ability in typically developing readers.

If sensitivity to speech rhythm is important for reading development then it should also be associated with reading success as much as it is with reading failure. Studies therefore need to demonstrate an association between measures of reading attainment and speech rhythm sensitivity in typically developing populations.

3 Speech rhythm should be linked to reading ability regardless of the language spoken.

Wood and Terrell (1998) specifically examined sensitivity to metrical stress as their measure of speech rhythm. However, other measures of speech rhythm that are less closely associated with the features of a specific language are possible (e.g. overall intonation patterns), and other languages have other elements that are characteristic of speech rhythm. It is therefore important to consider whether the results found by Wood and Terrell are peculiar to stress-timed languages.

4 There will be associations between speech rhythm sensitivity and segmental phonological awareness.

This prediction directly tests one of the suggestions made by Wood and Terrell (1998); namely, that sensitivity to speech rhythm may be a necessary prerequisite for successful reading development. If this is the case we should expect to find evidence of an association between segmental phonological awareness and prosodic sensitivity. Ideally, there should be longitudinal evidence of an association between speech rhythm and phonological awareness.

5 Speech rhythm sensitivity should be linked to skilled reading.

Finally, we should expect to find that sensitivity to speech rhythm will be associated not just with young children acquiring literacy, but also with skilled reading in an adult sample. If it is, we can say with some confidence that representation of suprasegmental phonology is a key component of reading processes, rather than merely enabling young children to bootstrap their way into segmental phonology.

The sections that follow review the evidence available to date in relation to each of these predictions in turn, before offering a synthesis of this literature in the form of a potential model that accounts for the mechanisms through which reading may be influenced by sensitivity to the prosodic features of speech.

Reading difficulties and speech rhythm

The initial Wood and Terrell (1998) study identified a deficit in children who were experiencing a delay in their reading development: none of the children in the sample had been identified as experiencing developmental dyslexia, although that did not necessarily mean that none of the children had dyslexia, merely that they had not yet had an assessment. Another important feature of the sample was that the children in the reading difficulties group represented a broad age range, which also meant that the children were potentially in different stages of reading development.

Subsequent studies of speech rhythm sensitivity have used more focused samples and have been concerned with children formally identified as experiencing dyslexia. The most notable examples of such work have looked at children’s sensitivity to amplitude envelope onsets or ‘rise time’. Goswami et al. (2002) conside...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• List of figures and tables
• List of contributors
• Introduction: contemporary perspectives on reading and spelling
• PART 1 Overarching debates in reading and spelling
• PART 2 Reading and spelling across languages
• PART 3 Written language difficulties and approaches to teaching