Content uploaded by M. Caravolas
Author content
All content in this area was uploaded by M. Caravolas on Mar 06, 2020
Content may be subject to copyright.
1
The Nature and Causes of Dyslexia in Different Languages
Markéta Caravolas
(University of Liverpool)
8/06/2004
In press in: The Science of Reading: A handbook. M. Snowling and C. Hulme (Eds.).
Blackwells Publishers, UK
2
The Nature and Causes of Dyslexia in Different Languages
Introduction
Developmental dyslexia is now a recognized disorder in many literate societies. On its
website, the International Dyslexia Association boasts affiliated branches on every
continent and in as many as 40 countries. Yet to date, far more scientific research has
been reported about dyslexia in English than in any other language. It is therefore not
surprising that many misconceptions still exist about the very existence, the
manifestation, and the causes of dyslexia in languages other than English. For
example, a common lay belief is that dyslexia is more common among
speakers/readers of English because the ‘chaotic’ English orthography is difficult to
learn; in contrast, dyslexia is thought to be virtually nonexistent in ‘easy phonetic’
orthographies such as Finnish and Spanish. Similarly, a popular belief about Chinese
is that its logographic script requires rote memorization of thousands of picture-like
characters and therefore reading difficulties of the sort observed in English must not
exist among Chinese readers. Fortunately, the situation is changing and dyslexia is
now being studied in a large variety of languages (see Goulandris, 2003). It is
becoming clear that, although many of the lay ideas about dyslexia are inaccurate,
they nevertheless reflect an intuitive understanding of one important fact, namely that
reading and writing skills are influenced by the characteristics of different writing
systems (see chapters by Frost, Hanley, and Treiman & Kessler this volume for
further discussion of this).
This chapter considers how such influences affect the manifestation of
dyslexia in different languages. More specifically, a review is presented of what is
currently known about dyslexia in various non-English languages and writing systems
with respect to two main issues. The first is how various characteristics of written
languages influence the types of reading and writing difficulties experienced by
children with dyslexia. The second is whether the characteristics of different
languages and orthographies influence the nature of the underlying cognitive deficits
that are associated with dyslexia. Clearly, apart from language differences, other
factors also play a role in determining cross-linguistic differences in the manifestation
of dyslexia such as educational practices and social and cultural factors. However,
these factors are beyond the scope of the present chapter and will only be touched on
3
tangentially when relevant. Before examining these issues, a brief summary of the
major features of different writing systems is presented.
Characteristics of different writing systems
The world’s languages are represented by a large variety of writing systems,
which may be broadly categorized into three main types on the basis of the units of
spoken language they represent. Alphabetic orthographies such as English, Dutch,
and Czech, use letters and letter clusters to represent phonemes. Syllabic
orthographies, such as Japanese Kanji and Hindi use syllabographs to represent
syllables. (The category of syllabic writing systems will not be discussed here,
mainly because very few studies of dyslexia in these languages are currently available
in English.) Logographic orthographies, such as Chinese and Japanese Kanji use
characters to represent spoken words at the level of monosyllabic morphemes (see
Mattingly, 1992, and also chapters by Frost, Hanley, and Treiman & Kessler, in this
volume for further discussion ).
Presumably, the gross differences in the size and level of the speech units that
are graphically encoded (i.e. phoneme, syllable, morpheme) are associated with
differences in the processes involved in reading and writing; we will examine this
issue by contrasting studies in alphabetic writing systems with those in the
logographic orthography of Chinese. However, reading and writing processes may
also be affected by more subtle differences between orthographies, even within the
category of alphabetic systems.
Characteristics of alphabetic writing systems.
Alphabetic writing systems are based on the principle that graphemes (letters
and letter strings) correspond to the phonemic units of speech. However, these
systems may differ from each other in a number of ways. Several terms, such as
orthographic depth, transparency, consistency and regularity have been used to
describe and compare them (see Frost, and Treiman & Kessler this volume). From
the perspective of the young learner, an ideal (transparent, consistent and regular)
alphabetic orthography should contain a set of one-to-one grapheme-phoneme (or
spelling-sound) and phoneme-grapheme (or sound-spelling) correspondences. Thus,
there should be only one way to pronounce any given grapheme, and only one way to
spell any given phoneme. A small minority of writing systems such as Finnish,
4
Turkish, and Serbo-Croatian approach this ideal. However, the majority of alphabetic
orthographies also encode information other than the phonemic content of words,
such as: morphological information (e.g., in the words “heal” and “health”, two
different vowel phonemes /i/ and /ε/ are represented by a single vowel letter string
reflecting their morphological relationship); original spellings of words imported
from other languages (e.g. the Anglicized words chateau and connoisseur have
retained their original French spellings); and historical spellings that reflect archaic
word pronunciations (e.g., spellings knee and walk retain graphemes that once
represented articulated sounds, but as a result of phonological change have become
redundant) (see Treiman & Kessler , this volume).
Alphabetic orthographies differ in the extent to which they allow these sources
of variation to occur, and these differences determine the degree of spelling-sound
and sound-spelling consistency and regularity. English is considered to be the most
inconsistent and irregular system because (a) the relationship between graphemes and
phonemes is often opaque (e.g. the letter t in “listen” has no corresponding phoneme),
(b) the grapheme-phoneme and phoneme-grapheme correspondences are inconsistent
(e.g. the grapheme ea has different pronunciations in “head” and “heal”; and although
the words “beef”, “chief”, and “leaf” all contain the same vowel /i/, it is assigned a
different spelling in each word), and (c) many exceptions exist to acceptable
orthographic patterns (e.g. the spelling trek violates the rule that word-final /k/ in
monosyllabic words with short vowels is spelled with the grapheme ck) and in sound-
spelling correspondences (e.g. the grapheme oe in “does” and the grapheme string cea
in “ocean” both correspond to exceptional pronunciations).
Most European orthographies are more regular and consistent than English
(see Seymour this volume), although an asymmetry exists in almost every alphabetic
system such that grapheme-phoneme correspondences are more consistent than
phoneme-grapheme correspondences (making reading easier than spelling). For
example, although written French is less complex than English in grapheme-phoneme
mappings (i.e. reading), it is as inconsistent as English in phoneme-grapheme
mappings (i.e. spelling) (Ziegler, Jacobs, & Stone, 1996). The relatively more
shallow orthography of German has more consistent grapheme-phoneme
correspondences than both English and French, however the phoneme-grapheme
correspondences are considerably less consistent, still making spelling more difficult
than reading (Wimmer & Mayringer, 2002). In comparison to English, French or
5
German, the Czech orthography is more consistent in both directions, although some
inconsistencies do exist in phoneme-grapheme correspondences (Caravolas, in press).
Unfortunately, with a few exceptions, direct comparisons of orthographic consistency
are currently not possible because comparable statistical estimates are still unavailable
for most of the world’s orthographies. Thus, at present, less formal estimates are used
to rank alphabetic orthographies in terms of depth, consistency and regularity. Figure
1 presents a schematic, based on published estimates and descriptions, of the relative
depth of some of the alphabetic languages that are considered in the present chapter
(see e.g. Caravolas, in press; Frost, Katz, & Bentin, 1987; Seymour, Aro, & Erskine,
2003)
Characteristics of the logographic writing system of Chinese.
Logographic writing systems are used in Chinese and Japanese, however the
focus here is primarily on the orthography used to represent various dialects of
Chinese. As described by Hanley (this volume), the smallest functional units of
orthographic representation in Chinese are characters, which correspond to
monosyllabic morphemes. A popular misconception about the Chinese logography is
that its visually complex characters are pictographs or ideographs which, like icons,
represent objects or concepts (e.g. the character for ‘rain’ looks like falling rain), and,
that they encode no phonological information about pronunciation.
In fact, only a small proportion of Chinese characters are strictly pictographic
or ideographic (less than 18% according to Zhou, (1978)). Approximately 80% to
90% of characters are semantic-phonetic compounds consisting of an element called
the semantic radical that gives information about the meaning, and an element called
the phonetic, that gives information about the pronunciation of the word. Hence, it is
possible to extract from most Chinese characters cues to word meaning and
pronunciation. The ability to extract semantic information might be particularly
advantageous because Chinese has a great many homophones, and therefore semantic
radicals provide a means of disambiguating the meanings of homophonic characters.
Thus, the homophonic nature of the spoken language may implicitly motivate children
to attend to the internal structure of characters from early stages of reading
development (Shu & Anderson, 1997).
However, in contrast to alphabetic writing systems, it is more difficult for
readers to make use of semantic and phonetic cues because they are not highly
6
reliable. For example, in Chinese children’s reading materials, approximately30% of
phonetic components provide unambiguous information about the pronunciation of a
character (i.e. the word is pronounced in the same way as its phonetic component),
and 40% of the semantic components provide unambiguous information about its
meaning. The remaining components are semi-transparent or opaque offering only
partial clues or no clue to pronunciation or meaning (Shu, Chen, Anderson, Wu, &
Xuan, 2003). Moreover, almost one half of the phonetic components encountered by
primary school children appear in only one word, thus, they are not productive (Shu et
al., 2003). Importantly, Shu et al. found that in the early grades of the Chinese reading
curriculum, children learn mainly irregular characters (i.e. generally high-frequency,
visually simple characters with unproductive phonetic components), while in later
grades they learn increasingly more regular (productive) characters.
Dyslexia among English speakers: A point of reference
A detailed review of dyslexia research based on English-speaking populations
is provided by Vellutino and Fletcher (present volume; see also Snowling, 2000;
Vellutino, Scanlon, Fletcher, & Snowling, in press) and only a few of the more
important conclusions from this research will be highlighted here. The literacy-skills
profiles of English-speakers with dyslexia typically include inaccurate and slow word
recognition skills (caused by difficulties with word decoding) (e.g. Vellutino et al.,
1996) , and very poor spelling skills (e.g. Bruck & Waters, 1988). Also, a deficit (in
preschool and early school years) is usually observed in letter knowledge, which is a
foundation skill for alphabetic literacy (Snowling, Gallagher, & Frith, 2003). In
addition, English-speaking dyslexics have marked difficulties with grapheme-
phoneme transcoding as assessed by nonword reading (Rack, Snowling, & Olson,
1992), and spelling (Caravolas, Bruck, & Genesee, 2003). A number of underlying
cognitive skills that are important components of reading and spelling in English, are
frequently impaired among individuals with dyslexia. The most consistently observed
impairments are in phoneme awareness (e.g. Liberman & Shankweiler, 1979), verbal
short-term memory (e.g. Hulme, 1981), slow naming speed (e.g. Bowers & Wolf,
1993), phonological learning (e.g. Vellutino, Scanlon, & Spearing, 1995), and word
and nonword repetition (e.g. Brady, Shankweiler, & Mann, 1983). In contrast, visual
7
skills such as visual memory and visual-spatial skills have not generally been found to
be deficient in English-speakers with dyslexia (see Vellutino & Fletcher, this
volume). Also, it is important to point out that not every individual with dyslexia
suffers impairments in every one of the above cognitive skills, and often one skill is
more severely impaired than another.
An influential theory, referred to as the ‘core phonological representations
hypothesis of dyslexia’, proposes that all of the cognitive impairments listed above
reflect an underlying deficit in phonological representation and processing (e.g.
Snowling, 2000; Stanovich, Nathan, & Zolman, 1988; Vellutino & Fletcher, this
volume; Vellutino et al., in press ). According to this view, an impairment in the
representation and processing of phonological information at the level of speech
sounds (phonemes) disrupts the dyslexic child’s ability to acquire the foundations of
alphabetic literacy, namely letter knowledge (the names and sounds of the letters of
the alphabet), phoneme awareness (the ability to consciously manipulate the
phonemes in spoken words) and phonological recoding (knowledge and use of
grapheme-phoneme and phoneme-grapheme correspondences in reading and spelling)
(Byrne, 1998; Snowling & Hulme, 1989).
Dyslexia in non-English languages with alphabetic orthographies
How might orthographic depth, as indexed by consistency and regularity,
influence the manifestation of dyslexia in different languages? It is now quite well
established that the rate at which children acquire basic reading skills varies as a
function of consistency and regularity in grapheme-phoneme correspondences (e.g.
Seymour et al., 2003; Seymour this volume). In every direct comparison reported to
date children learning more consistent writing systems have been found to make
greater gains in the first few years of schooling than their English peers in word
recognition accuracy, fluency, and conventional spelling (Caravolas et al., 2003;
Geva, Wade-Wooley, & Shaney, 1993; Wimmer & Goswami, 1994; Wimmer &
Goswami, 1994).In fact, in a number of regular orthographies, such as Turkish
(Durgunoglu, Nagy, & Hancin-Bhatt, 1993), Finnish, Greek, and German (Seymour et
al., 2003), children are reported to reach ceiling levels in accuracy of word and
nonword reading, by the end of grade 1. It has been proposed that the reliable
correspondence between graphemes and phonemes in shallow orthographies, in
8
combination with phonics reading instruction (which is the standard teaching method
in most European countries) helps children to learn the letter-sound correspondences
of their language. This in turn boosts their ability to ‘assemble’ phonemes (e.g. when
decoding) as well as to analyse words into phonemes (e.g. when spelling); thus,
children acquire the skills that form the basis of alphabetic literacy more quickly in
shallow than in deep orthographies (e.g. Caravolas & Bruck, 1993; Cossu,
Shankweiler, Liberman, Katz, & Tola, 1988; Wimmer, Landerl, Linortner, &
Hummer, 1991).
A further, less clearly established, claim is that in shallow orthographies the
role of phonological skills, and in particular of phoneme awareness, is less important
and limited to only the first year or two of literacy acquisition (de Jong & van der
Leij, 1999, 2002; Landerl & Wimmer, 2000; Wimmer, 1993). That is, the effect of
regular and consistent spelling-sound correspondences is thought to be sufficiently
powerful to secure children’s phonological recoding skills after a few months of
reading experience, regardless of their pre-reading levels of phonological awareness.
Data consistent with these hypotheses have been reported in studies of Turkish (Oney
& Goldman, 1984), Dutch (Bast & Reitsma, 1998), and Hebrew (Bentin & Leshem,
1993), in which correlations between phonological awareness and reading were strong
in the early stages of grade 1 but became weaker toward the end of first or second
grade primarily due to ceiling effects in basic phonological recoding and word
recognition. However, these ceiling effects leave open the possibility that the absence
of a correlation was merely an artifact of inappropriately easy reading tasks. Instead,
the best long-term predictor of reading fluency performance is reported to be rapid
automatized naming of visually presented stimuli (RAN) (de Jong & van der Leij,
1999; Wimmer, Mayringer, & Landerl, 2000).
However, discrepant findings have also been reported. Kozminsky and
Kozminsky (1995) showed that phoneme awareness in pre-readers predicted literacy
achievement at the end of grade 3 among learners of the regular Hebrew orthography,
and concurrent correlations were obtained between these skills among children in
grades 1 to 5 in studies of Finnish (Muller & Brady, 2001), Dutch (Patel, Snowling, &
de Jong, in press) and Czech (Caravolas & Volín, submitted), suggesting that, as in
English, phoneme awareness is a relevant component skill in shallow orthographies
beyond the second grade. Thus, whereas the cross-linguistic findings unequivocally
demonstrate the positive effect of orthographic transparency on the rate of literacy
9
acquisition, the importance and duration of the role phoneme awareness, is currently
unresolved.
Reading skills in children with dyslexia in languages other than English.
Dyslexic children of primary school age learning consistent orthographies
seem to have less serious difficulties than their English-speaking counterparts.
English children with dyslexia, generally have persistent deficits in word reading
accuracy, and even more severe deficits in nonword reading, with error rates often
ranging from 50% to 70% (see Rack et al., 1992). Their counterparts in languages
such as French (Sprenger-Charolles, Siegel, Bechennec, & Serniclaes, 2003), German
(Landerl, Wimmer, & Frith, 1997), Dutch (deJong & van der Leij, in press), and
Greek (Porpodas, 1999) typically attain much higher scores, with error rates in the
order of 6% (Dutch) to 25% (French). Importantly, even these relatively low error
rates are typically significantly higher than those of age-matched controls, and
sometimes also higher than reading ability matched controls. Hence, children with
dyslexia who learn relatively transparent orthographies do experience problems with
word and nonword reading accuracy, but these appear to be less severe than those of
their English-speaking peers.
Turning to word and nonword reading fluency, individuals with dyslexia seem
to have more serious and pervasive difficulties across alphabetic orthographies in that
they typically show a much greater reading speed deficit relative to their reading
accuracy skills (e.g. Landerl et al., 1997; Wimmer, 1993; deJong & van der Leij, in
press);. It is not clear, however, whether children with dyslexia learning more
consistent orthographies read more quickly than their English counterparts. For
example, in a direct comparison of German and English children, Landerl et al. (1997)
found that the German dyslexic and normal readers read one-, two-, and three-syllable
words more quickly than the English children. In contrast, in a recent comparison of
Czech and English children with dyslexia, in which reading fluency was measured in
syllables per second (as opposed to mean word-reading time), Caravolas and Volín (in
preparation) found that, although both dyslexic groups had a reading fluency deficit
relative to younger spelling-ability-matched peers, they did not differ from each other.
However, a recent study by Ziegler, Perry, Ma-Wyatt, Ladner, and Schulte-
Korne (2003) illustrated that absolute magnitude differences in cross-linguistic
comparisons may be misleading due to inherent differences between the statistical
properties of word lists being assessed in different languages. In a direct comparison
10
of German- and English-speaking children with dyslexia, the authors constructed
word lists that were similar in form and meaning in both languages (using cognates,
such as ‘box’ and ‘sport’), and similarly constructed nonword lists that were matched
for number of letters, orthographic regularity, and consistency as estimated by
neighbourhood size (i.e. the number of other words with the same orthographic rime
as the target word). Children were assessed on accuracy and speed of reading.
Although the German children (dyslexic and control) tended to read more quickly and
accurately overall, the English and German dyslexic groups showed similar decreases
in accuracy and speed when reading nonwords relative to words and when reading
longer, compared to shorter, words. Both results are consistent with an impairment in
phonological recoding skills. Moreover, both groups of children with dyslexia
demonstrated neighbourhood effects of similar magnitude as their respective reading
ability controls; that is, they read words and nonwords with many neighbours more
quickly than those with few neighbours, suggesting that they are senstitive to the
statistical properties of orthographic patterns that extend beyond the grapheme in their
language. The language groups did differ in one respect: that is, in contrast to
normally developing readers, the English children with dyslexia did not seem to make
use of their knowledge of larger orthographic units (e.g. orthographic rimes) to read
words– they continued to read in a serial manner regardless of the length and degree
of consistency of the words. This effect was much less pronounced in German, where
all three groups tended to rely more strongly on a serial reading procedure. In sum,
this study demonstrates that, when similar materials are used, children with dyslexia
show similar patterns and magnitudes of reading impairment regardless of the degree
of transparency of the orthography.
Spelling skills in children with dyslexia in languages other than English.
Spelling skills have been studied less than reading skills, but, it is clear that
individuals with dyslexia experience serious and pervasive problems in learning to
spell in a variety of alphabetic orthographies (e.g. Alegria & Mousty, 1994; Bruck &
Waters, 1988; Caravolas et al., 2003; Wimmer, 1996b);. Direct comparisons of
spelling ability between speakers of English and other languages are rare, however, a
recent study of English- and French-Canadian children (Caravolas et al., 2003)
demonstrated that, as in reading, English-speaking children with dyslexia tend to
produce higher absolute rates of errors than their counterparts learning more
transparent orthographies. In this study comparable lists of words and derived
11
nonwords containing regular as well as irregular and inconsistent graphemes were
administered to 3rd grade groups of poor spellers and age-matched good spellers. On
both tests, the English poor spellers produced very high rates of errors (roughly 87%),
and these were twice as high as those produced by the French poor spellers (roughly
43%). Despite the advantage relative to English-speaking children, these results show
that even in writing systems that are more transparent than English, children with
dyslexia tend to produce high error rates, particularly with irregular and inconsistent
spelling patterns.
The latter was convincingly demonstrated in several French studies carried out
by Alegria and Mousty (Alegria & Mousty, 1994, 1996) who found that on words
containing highly consistent and context-independent phoneme-grapheme
correspondences (e.g. spelling the phoneme /u/ which is always represented with the
grapheme ou), 9- to 14-year-old dyslexics spelled as accurately as their normally
developing, reading-matched peers. Their deficits relative to normally developing
children were in learning the inconsistencies and irregularities of conventional
spelling (e.g. learning that the phoneme /g/ is represented by the grapheme g except
when followed by /i/ and /e/, in which case it is represented by the grapheme gu).
Whereas normal readers demonstrated the ability to learn inconsistent, and context-
dependent spelling rules as their reading age increased, dyslexic children showed no
such developmental pattern. They suggested (Alegria & Mousty, 1994) that because
these French dyslexic children could represent consistent graphemes quite accurately,
they may not necessarily have had impairments in phonological awareness (and
analysis) but rather in learning spellings for specific irregular words. They argued
that this deficit arises because French dyslexics probably fail to acquire fully specified
orthographic representations as they do not attend to word spellings fully in reading.
However, the phonological spelling accuracy of the children with dyslexia and their
normally developing peers was not compared in statistical analyses in either study;
nor was phonological awareness assessed. Consequently, the status of the dyslexic
children`s phonological skills remains unknown. Nevertheless, the main findings of
Alegria and Mousty are consistent with several German studies in which dyslexic
children produced many conventional spelling errors but few phonologically
implausible spelling errors (e.g. Landerl, 2001; Wimmer, 1996a) ), and in some cases,
the children with dyslexia were as accurate in phonological spelling as their age-
matched normal reader peers (Landerl & Wimmer, 2000).
12
Results such as these have lead to some debate regarding the basis of spelling
difficulties among children learning orthographies other than English. An assumption
from the phonological deficit hypothesis of dyslexia is that due to their phonological
processing deficit, dyslexic children fail to acquire basic phonological recoding skills,
which are required for ‘assembling’ the spellings of unknown words and nonwords,
and in turn these weak foundation skills lead to poor reading and spelling skills (e.g.
Byrne, Freebody, & Gates, 1992; Caravolas, Hulme, & Snowling, 2001). Thus, poor
spelling is thought to be a consequence of poor phonological processing skills (e.g.
Snowling, 2000). A contrasting view, similar to that proposed in Bowers and Wolf’s
(1993) ‘double deficit’ hypothesis, has been proposed for learners of more regular
orthographies (Wimmer et al., 2000; Wimmer & Mayringer, 2002). Pointing to the
relatively low phonological recoding error rates in spelling and in nonword reading
that were observed in several studies (e.g. Landerl & Wimmer, 2000; Wimmer, 1993,
1996), proponents of this view posit that although children with dyslexia may suffer
phonological processing difficulties at school entry their difficulties resolve by about
the end of the second grade, due to the benefits conferred by predictable grapheme-
phoneme correspondences and phonics literacy instruction. The children`s persistent
difficulties in spelling accuracy (and reading fluency) are attributed to a faulty “timing
mechanism” that is independent from the early phonological impairment, and which
impedes the formation of associations between phonemes and graphemes. This view
has profound theoretical implications as it suggests that characteristics of the writing
system affect the nature of the literacy and cognitive deficits in dyslexia in different
languages.
However, several studies have obtained results that are more compatible with
a simpler, unitary phonological deficit account. For example, dyslexic children
learning several different alphabetic orthographies have been found to experience
persistent deficits not only in conventional spelling but also in phonological spelling
accuracy well beyond the 2nd grade. In the study by Caravolas et al. (2003), a
qualitative analysis of the children`s misspellings revealed that French 3rd grade poor
spellers produced significantly fewer phonologically plausible spellings than good
spellers, although unlike English poor spellers, they were not significantly impaired in
representing the segmental structures of words (i.e. they were less likely to omit
graphemes). Sprenger-Charolles and colleagues (2000) reported that 4th grade,
French children with dyslexia showed a nonword spelling deficit (mean error rate of
13
23%) not only relative to chronological age peers (mean error rate of 4%), but also
relative to younger reading-age matched children (mean error rate of 9%). It might be
argued that the above results of persistent phonological recoding difficulties simply
reflect the relatively low phoneme-grapheme consistency of French (Ziegler et al.,
1996), and would not be observed in more consistent orthographies. However,
Wimmer (1993) and Landerl (2001) found that German dyslexic children in 2nd to 4th
grade, whose overall levels of performance were admittedly high, nevertheless
produced significantly fewer phonologically plausible spellings than age-matched
control children, suggesting that their phonological recoding difficulties had not
resolved in grade 2. Moreover, Caravolas & Volín (Caravolas & Volín, 2001) found
that dyslexic children in grades 3 to 5, learning the highly regular orthography of
Czech, produced significantly fewer phonologically plausible spellings of words (19%
error rate) than their age mates (4% error rate), and did not differ in this respect from
spelling-matched control children who were two years younger (17% error rate). In a
more recent study, Czech dyslexic children also showed a nonword spelling deficit
relative to younger spelling-ability matched peers (Caravolas & Volín, submitted).
Summary: the reading and spelling profiles of children with dyslexia learning shallow
orthographies
The results for spelling are consistent with those obtained for reading: children
with dyslexia suffer milder deficits in transparent orthographies than in English. Like
English children with dyslexia, children with dyslexia learning more transparent
orthographies lag seriously behind normally developing readers in conventional
spelling skills, and experience persistent – if subtle – phonological spelling problems.
Further research is required to elucidate the causes of the discrepant findings
in some of the studies (e.g. Landerl & Wimmer, 2000; Wimmer, 1996) that seem
more consistent with a timing mechanism impairment than with a phonological
deficit. A central question, is whether the reading and spelling difficulties observed in
dyslexic children learning transparent orthographies, as well as English, can be
reduced to a common underlying deficit in phonological skills? In this view the
milder literacy difficulties found among children with dyslexia learning transparent
orthographies simply reflect the lesser demands that are placed on phonological skills
for all children learning such orthographies. The alternative, but less parsimonious,
view is that there may be different cognitive impairments underlying the literacy
difficulties of children who show dyslexic difficulties in transparent orthographies.
14
This issue is explored further in the next section, where the main cognitive
impairments that have been shown to be associated with dyslexic difficulties in
transparent orthographies are presented.
Cognitive profiles of children with dyslexia in languages other than English.
In this section, we consider the findings regarding the roles of phonological
awareness, verbal short-term memory, and RAN. The role of phonological awareness
skills, and of phoneme awareness in particular, has received most attention in studies
of developmental dyslexia in shallow orthographies. Partly, this is because it is
arguably the most robust predictor of literacy achievement, at least among
speakers/readers of English (Stanovich, 1992). However much of the interest in
phoneme awareness has been generated by the claim that phoneme awareness is a less
important and time limited component skill for normally developing learners of
relatively shallow orthographies and that it does not discriminate between good and
poor readers beyond the second grade (de Jong & van der Leij, 1999, 2002; Landerl &
Wimmer, 2000; Wimmer, 1993).
In line with their findings regarding phonological recoding skills in nonword
reading and phonological spelling, Wimmer and colleagues (Landerl & Wimmer,
2000; Wimmer, 1993; Wimmer, 1996a; Wimmer et al., 2000) report that German-
speaking children with dyslexia experience significant difficulties in phoneme
awareness in first grade, but that these problems resolve by about the end of the
second grade and do not influence reading and spelling development thereafter.
Compatible findings have been reported for Dutch 3rd grade poor readers who had
mastered phoneme blending, which did not correlate with word reading speed,
although this is one of the simpler phonological awareness tasks, and accordingly, the
range of scores was restricted by ceiling effects (Wesseling & Reitsma, 2000).
This profile, however, is at odds with some findings in German as well as in
other languages. For example, in a training study with German kindergarten children,
Schneider, Roth, and Ennemoser (2000) demonstrated that pre-literate children who
had poor phonological awareness, and hence were potentially at risk of reading
failure, benefited from phoneme awareness training, in combination with letter
knowledge training, more than from training in either phoneme awareness or letter
knowledge alone, and the benefits of the training persisted at least until the end of
second grade. Similar results were obtained by Bentin and Leshem (1993) in a
training study with at-risk Hebrew-speaking kindergarteners. In this study, children
15
who underwent a 10-week phoneme awareness training programme (with or without
additional training in letter knowledge) were indistinguishable at the end of first grade
from same-aged peers with good initial levels of phoneme awareness on word reading
accuracy and phoneme awareness tasks. Equally importantly, at-risk children who
received general language (comprehension, vocabulary, syntax) training or no
intervention went on to develop reading difficulties. Although neither of these studies
assessed performance beyond the second grade, which according to Wimmer and
colleagues is the critical point at which children with dyslexia catch up with normally
developing peers in phoneme awareness and skills requiring phonological recoding,
they offered no hint of relative improvement over the course of the first (Bentin &
Leshem, 1993) and/or second grade (Schneider et al., 2000) for the children untrained
in phoneme awareness.
In their cross-sectional study with older children with dyslexia (described
earlier), Landerl et al., (1997) found that groups of German and English children with
dyslexia aged 10 to 12 years, attending grades 5 and 6 both performed less well than
chronological age and younger reading ability matched control children on
spoonerisms, a phoneme awareness task requiring the transposition of onsets in word
pairs (e.g. boat-fish Æ foat-bish; blue-red Æ rue-bled), which is thought to be
conceptually appropriate for older children. Thus, these children’s phoneme
awareness difficulties had clearly not resolved by the second grade. Recently
however, Landerl and Wimmer (2000) reanalysed the data from this study, arguing
that the dyslexic children’s difficulties may not have reflected a phoneme awareness
deficit but rather a memory deficit (because the spoonerisms task places high
demands on verbal short-term memory). Using a more lenient scoring scheme that
reduced the effects of short-term memory on performance, they found that indeed the
error rate was reduced dramatically (from 63% to 15%), and the mean of the group
with dyslexia was now on par with the reading ability control group. Nevertheless,
the children with dyslexia continued to perform less well than the age-matched
control children.
Persistent phoneme awareness and phonological recoding difficulties among
children with dyslexia have also been documented in a number of other languages
with transparent orthographies. A spoonerism task was also administered in the
Czech-English cross-linguistic study of Caravolas & Volín (submitted). Like Landerl
et al. (1997) they found that both groups of children with dyslexia were significantly
16
impaired on this task relative to chronological age peers. Moreover, controlling for
verbal memory span did not alter the results. In addition, on an easier phoneme
deletion task, both groups were impaired relative to age-matched peers, and, the
Czech children were also impaired relative to younger spelling-ability-matched peers.
In keeping with previous results on reading and spelling measures, the Czech dyslexic
and younger control children obtained higher scores on both phoneme awareness
tasks than their English peers, suggesting that orthographic transparency affects not
only the rate of literacy development but also of phoneme awareness. Phoneme
awareness deficits were also observed in recent studies with Dutch children (de Jong
and van der Leij, in press; Patel, Snowling, & deJong, in press) when a sufficiently
difficult phoneme deletion task was used. Similar results have been obtained among
children with dyslexia from second to fifth grade in Hebrew (Ben-Dror, Bentin, &
Frost, 1995; Breznitz, 1997) and French (Caravolas et al., 2003; Sprenger-Charolles et
al., 2000).
To date, the results regarding the role of phoneme awareness in accounting for
individual and group differences in literacy development in languages with relatively
transparent writing systems are mixed. While some are consistent with the
orthography-specific account of Wimmer and colleagues, others suggest that phoneme
awareness plays a similar role in the development of alphabetic literacy regardless of
differences in orthographic transparency. Whether the discrepancies are due to
differences in the types and reliability of phoneme awareness measures that have been
used, or to differences in the types of literacy skills, or age groups, or dyslexic
population that have been assessed across studies is difficult to determine. The issue
clearly warrants further study, as it has serious implications for theories of the effects
of orthographic consistency on cognitive processes in reading and spelling, and for
our understanding of the nature of the cognitive deficits in dyslexia in different
languages.
Investigations of the role of verbal short-term memory in the profile of
individuals with dyslexia have produced much more unanimous results across
alphabetic orthographies. On various measures, such as digit span, word span and
nonword span, school-age children with dyslexia almost invariably experience
persistent impairments relative to children of the same age, and sometimes also
relative to younger children with equivalent reading skills. This has been found to be
the case in Dutch (de Jong, 1998), French (Sprenger-Charolles et al., 2000), German
17
(Schneider et al., 2000), and Czech (Caravolas & Volín, submitted). Although verbal
short-term memory deficits have sometimes been proposed to underlie dyslexic
children`s poor performance on phoneme awareness tasks (Landerl & Wimmer,
2000), this has not been borne out in analyses in which the role of both variables has
been considered simultaneously, in English or in other languages. That is, group
differences in phoneme awareness were not eliminated when the effect of verbal
short-term memory was partialled out either for English or Czech children in
Caravolas and Volín`s (submitted) study, and, phoneme awareness contributed unique
variance to reading ability over and above verbal short-term memory among English
(McDougall, Hulme, Ellis, & Monk, 1994)
The role of rapid automatized naming (RAN) has been investigated in a
number of alphabetic orthographies. English-speaking individuals with dyslexia are
typically slow on RAN tasks (Vellutino & Fletcher, this volume), and this is also true
of their peers learning more transparent orthographies (de Jong, & van der Leij, 1999,
2002; Wimmer, 1993, 1996). However there is considerable controversy about the
nature of the cognitive mechanisms that are tapped by naming speed (see Vellutino &
Fletcher, this volume). Although RAN correlates moderately with phonological
awareness skills (e.g. Schatschneider, Carlson, Francis, Foorman, & Fletcher, 2002),
the two abilities are separable and may be partially dissociated in some groups of
dyslexic individuals (e.g. English: Bowers & Wolf, 1993; German: Wimmer et al.,
2000; Wimmer & Mayringer, 2002). Moreover, in several non-English studies, RAN
has been found to be the best long-term discriminator of reading and spelling ability
(e.g. de Jong & van der Leij, 1999; Wimmer, 1993; Wimmer et al., 2000), while
phoneme awareness and verbal short-term memory were not. These findings have led
some researchers to argue that deficits on RAN tasks indicate an impairment in the
orthography-phonology timing mechanism that is independent from the phonological
system per se (e.g. Bowers & Wolf, 1993; Wimmer et al., 2000). However, others
have suggested that RAN tasks fall under the phonological processing skills umbrella
as estimators of the efficiency in accessing phonological representations, and that
deficits on RAN tasks further demonstrate a phonological processing deficit (e.g.
Snowling & Hulme, 1994; Torgesen, Wagner, Rashotte, Burgess, & Hecht, 1997).
The latter claim is supported by English-language findings of Torgesen and
colleagues (1997) that the predictive effects of RAN on later reading ability are
eliminated when children`s initial reading abilities are also taken into account. This
18
suggests that differences in naming speed may be a consequence rather than a cause
of differences in reading ability. Perhaps a more convincing finding, however, comes
from a recent study of Dutch children with dyslexia, in which both RAN and the
speed with which children performed a phoneme deletion task were simultaneously
assessed as predictors of reading ability (Patel et al., in press). The results showed
that although RAN was a unique predictor of reading over and above several other
variables, its effect was no longer significant when phoneme deletion speed was
added to the equation. This result suggests that RAN estimates some aspect of
phonological processing. Nevertheless, further research is required to fully
understand the RAN-literacy relationship.
Finally it is important to note the skills that have been found to be unimpaired
in children with developmental dyslexia learning to read more transparent
orthographies, as well as in English studies. These primarily include the short-term
retention of visual stimuli (e.g. Sprenger-Charolles et al., 2000) and visual processing
speed tasks (e.g. de Jong, 1998; Landerl, 2001). In sum, studies in languages with
alphabetic orthographies suggest that both the literacy and the cognitive profiles of
individuals with dyslexia have many commonalities, and on balance these seem to
outnumber the differences. Importantly, dyslexia presents as a language-based
disorder, in which a critical deficit involves skills related to phonological
representation and processing. The results to date indicate that the manifestation of
this deficit is more pronounced in English. However, whether or not the magnitude
relative to normal readers, or the duration of the phonological deficit varies as a
function of orthographic transparency has yet to be determined by large-scale
longitudinal studies in a variety of languages.
Dyslexia in the logographic writing system of Chinese
Research examining when normally developing readers become sensitive to
the components of Chinese characters is reviewed in detail by Hanley (this volume)
and will not be repeated here. It is critical to understand, however, that numerous
studies have now shown that children do become sensitive to the functions of the
phonetic and semantic radicals in Chinese characters implicitly through reading
experience. This suggests in turn that, as in English and other alphabetic writing
systems, phonological processing skills (such as phonological awareness) as well as
19
semantic skills may be important components in learning to read Chinese. Moreover,
whereas visual processing and memory skills may be of importance in the early stages
of reading (when children may place a greater reliance on rote memorization),
phonological and semantic analysis skills may become increasingly important in the
later years of primary school.
Component skills of reading in Chinese.
The importance of various component skills in reading has been investigated
in several correlational studies. In a four-year longitudinal study, Ho & Bryant (Ho &
Bryant, 1997) examined the predictive relationship between phonological awareness
and visual skills of Hong Kong Chinese pre-readers, and their reading skills up to four
years later. The authors found that children’s pre-reading visual skills were relatively
stable long-term predictors accounting for variance in reading over and above age, IQ,
and maternal education, at all three subsequent time points. Pre-reading phonological
awareness skills did not predict reading at the beginning of the first year in school, but
did by the end of that year and more so in the second year. Thus, consistent with
predictions based on children’s reading strategies, the authors concluded that Chinese
children tend to rely more strongly on visual skills in the early stages of learning to
read, and they seem to rely increasingly on phonological skills as their reading skills
develop. However in this study the contribution of phonological awareness skills was
not assessed over and above that of visual skills, and thus it is not clear whether
phonological skills accounted for unique variance when visual skills were controlled.
A study of 3rd grade Chinese children by Huang and Hanley (1994), suggested that it
does not. In their study, measures of phonological awareness and visual skills were
used along with other measures as predictors of word reading, and the results showed
that although both types of skills correlated significantly with reading ability, only
visual skills accounted for unique variance, after accounting for IQ, vocabulary, and
phonological awareness. However, the same authors (Huang & Hanley, 1997) failed
to replicate this finding in a longitudinal study of Taiwanese first-grade children in
which phonological awareness skills among children at the beginning of the first
grade correlated much more strongly with subsequent reading than did a visual
paired-associate learning task. Together these results suggest that phonological skills
are associated with reading in Chinese, as are certain (but perhaps not all) types of
visual processing skills. Support for the hypothesis that visual skills are more
important in the earliest stages of learning to read while phonological skills increase
20
in significance in later stages has to date been equivocal; the issue awaits further
longitudinal research.
Literacy and cognitive profiles of children with reading difficulties in Chinese.
What is the nature of the literacy and cognitive difficulties of Chinese-
speaking children with dyslexia, and how do they compare with those observed in
English and other alphabetic orthographies? A study addressing the question of how
reading difficulties manifest in Chinese was carried out by Tseng, Zhong, Hung and
Lee (1995) who investigated the extent to which good and poor readers in 3rd and 6th
grades are sensitive to the statistical properties of the Chinese orthography. Reading
ability was determined by children’s grades in language arts (Chinese) class. The
authors were specifically interested in whether children were influenced by
characteristics such as the regularity and consistency of the phonetic components of
characters in word recognition tasks. Each phonetic component has a dedicated
pronunciation in isolation, though its pronunciation may differ when it forms part of a
compound character. Thus, phonetics that always have the same pronunciation in
characters as they do in isolation are fully regular, but many (semi-regular) phonetics
have different pronunciations in some compound characters and others (completely
irregular phonetics) have different pronunciations in every character in which they
appear (see also Hanley, this volume). In the Tseng et al. study, children were asked
to read lists of pseudocharacters in which the phonetic component was either
completely regular, semi-regular (i.e. appeared in real characters sometimes with a
regular pronunciation and sometimes with an irregular pronunciation), or completely
irregular. The measures of interest were reading speed and the extent to which
children regularized the pronunciations, that is, read the pseudocharacter according to
the dedicated pronunciation of the phonetic element.
The analysis of reading speed revealed that all children were slowest in
reading the irregular characters, regardless of grade or reading ability. However, the
groups differed in the quality of their responses. Overall, the poor readers in both
grades produced fewer regularized responses than the normal readers, especially to
the regular items. In addition, the poor readers made more regularized responses on
the irregular items than the normal readers. This suggests that poor readers are less
sensitive than good readers to the internal structure of characters and to the statistical
information about the phonetic component. However, no control task was
administered to assess whether these children’s difficulties were specific to use of the
21
phonetic component or whether they also included the semantic component and/or the
visual features of characters.
Good and poor readers’ ability to use the semantic radical was examined by
Shu and Anderson (1997). In this study 1st, 3rd, and 5th grade children from Beijing
were selected as good and poor readers on the basis of teachers’ evaluations. All
children were given a morphological inference task in which they were presented with
a cue word in phonological form (Pinyin) and were to identify the corresponding
character from a choice of four alternatives on the basis of its radical component. The
degree of semantic transparency (the degree to which the radical was semantically
related to the word) and character familiarity (whether children had learned the
character in a previous term/year, in recent weeks, or had not yet learned it) were
manipulated. The results showed that poor readers were as skilled as good readers at
using the semantic component to identify words that were familiar to them, however
they were less able to use the radical to infer unfamiliar words. Moreover, when faced
with unfamiliar characters, good readers were better at inferring the meaning of words
with transparent radicals than opaque radicals, whereas poor readers found both types
equally difficult. This study suggests that Chinese poor readers in primary grades
may be aware of the components of characters they have already learned, but have
difficulties in processing and using the morphological-semantic information encoded
in characters productively. Unfortunately Shu and Anderson did not include measures
of phonetic inferencing again leaving open the question of how the poor readers’
difficulties in making use of the semantic radical compared with their ability to use
the phonetic component. Moreover, no (nonreading) measures of vocabulary or
morphological knowledge were included and so it is not clear whether the poor
readers’ difficulties were associated with oral language difficulties in the domain of
morphology and semantics, or whether they reflected these children’s low sensitivity
to the morphological features of Chinese orthography.
Chan and Siegel (Chan & Siegel, 2001) investigated not only reading
performance but also performance on tests of short-term memory and phonological
(tone) awareness among normal and poor readers in 1st to 6th grades of Hong Kong
primary schools. Tones play an important phonological role in Chinese as they
determine the intonation of each syllable, and they often distinguish syllables that
would otherwise be homophones. Importantly, tone information is not represented
orthographically and therefore if tone awareness is found to be associated with
22
reading ability, this association cannot simply arise as a consequence of learning to
read. Thus tone awareness presents an interesting measure for investigating the
relationship between phonological processing and reading in Chinese. Chan and
Siegel selected poor readers to be those who scored in the bottom 25% for their grade
on tests of reading; the remaining children in each grade formed the normal-reader
groups. On a test assessing reading performance, children read words that varied in
terms of their visual complexity (defined by the number of strokes in each character)
and frequency. Error analyses were carried out to determine whether children were
influenced by salient visual features, phonetic information, or the semantic radicals of
characters. Interestingly, poor readers were no worse than normal readers in
recognizing visually complex characters relative to simple ones, suggesting that their
difficulties do not arise from poor visual processing skills. The poor readers were
found to be as accurate as normal readers on high frequency words, but they were
significantly less accurate on low frequency words, suggesting that they are sensitive
to the statistical properties of printed words. The error analyses revealed that poor
readers generally behaved like younger normal readers. That is, both groups made
more semantic and visual errors than older, normally achieving children, who in turn
made more phonologically related errors (cf. Woo and Hoosain 1984). Poor readers
showed a delay relative to normal readers on short-term verbal memory, pseudo-
character reading (a test of the ability to use the phonetic component to generate
pronunciations), and tone discrimination. In sum, the poor readers of this study were
not impaired in recognizing visually complex characters, or in reading high-frequency
words – which they potentially learned by rote visual memorization. However, they
seemed to persist in using developmentally earlier, visual word reading strategies
rather than a phonetic processing strategy, suggesting that their reading difficulties
may be related to weak phonological skills.
The role of linguistic skills and working memory in Chinese reading
development were further examined by So and Siegel (1997) in a study of 196 normal
and poor readers in grades 1 to 4 in Hong Kong. The authors investigated whether
four types of oral language skills that are important for reading development in
English -- namely phonological awareness, semantic knowledge, syntactic processing,
and working memory --, are also important in Chinese. Moreover, they asked
whether poor readers of Chinese have specific difficulties in these skills. Children
scoring in the lowest quartile on a word recognition test were identified as poor
23
readers, and were compared to the remaining normal readers. Regression analyses
showed that children’s word reading skills were most strongly accounted for by tone
awareness (one of the tests of phonological awareness) and syntactic processing in
grades 1 to 3, and additionally by semantic skills in 4th grade. Poor readers showed a
significant lag in the development of all linguistic and memory skills, although their
most serious and persistent difficulties were at the levels of phonological and
semantic awareness. Notably, whereas normal readers had developed tone and rime
awareness to high levels by grade 1, the poor readers acquired this skill very slowly,
and in 4th grade were still only half as accurate as the 1st grade normal readers. Thus
this study demonstrated that phonological, syntactic, and semantic skills are important
to reading development in Chinese, and, in line with the study of Shu and Anderson
(Shu & Anderson, 1997), suggest that semantic analysis skills may become important
to reading somewhat later than phonetic analysis skills. Furthermore, it suggests that
Chinese poor readers are particularly impaired in the two skills that may be necessary
for the development of character-reading skills, namely phonological and semantic
awareness.
Although suggestive, the results of the preceding studies of poor readers must
be interpreted with caution in terms of what they reveal about children with dyslexia.
That is, none of these studies reported any information about the cognitive or broader
educational profiles of the poor readers leaving open the possibility that many were
not specifically impaired in reading but had more general learning difficulties.
Moreover, with the exception of some results in the study of Chan and Siegel (2001),
none of the above studies included reading ability matched control groups. Thus the
results provide no indication of whether poor readers’ difficulties are commensurate
with or are more severe than would be expected given their reading ability. These
design flaws were redressed in two studies by Ho and her colleagues.
Ho, Law, and Ng (2000) investigated whether Chinese children with dyslexia
(grades 2 to 5) experience phonological processing difficulties similar to those
experienced by English-speaking children. Fifty-six children diagnosed with dyslexia
participated. They were categorized in two groups: a group with reading and copying
delay, as well as visual spatial difficulties, and a group with specific reading
difficulties (both groups had a reading lag of at least 1 year). Both groups were
similar in age and had IQs in the normal range. The dyslexic children were
individually matched with chronological age- and IQ- controls as well as with reading
24
ability- and IQ-controls. The test battery included regular and irregular character
reading, pseudocharacter reading, onset and rime awareness, and "phonological
memory" (i.e. word and nonword repetition) tasks. Not surprisingly, both groups with
dyslexia were worse than their age peers on regular and irregular character reading,
and they did not differ from the reading ability control groups on either measure. On
pseudocharacter reading, the group with broader difficulties did not differ in ability
from either control group, while those with a specific reading problem performed less
well than the age-matched group but similarly to the reading-matched control group,
suggesting that they had a more severe difficulty in ‘decoding’ characters on the basis
of the phonetic component. However, on the phonological awareness and
phonological memory tasks, both dyslexic groups performed less well than the age
control groups, and no better or worse than their reading-matched counterparts on all
tasks. Ho et al. concluded that a phonological deficit is a feature of dyslexia not only
among readers of alphabetic orthographies but also among readers of Chinese.
More recently, Ho and her associates (Ho, Chan, Lee, Tsang, & Luan, in
press; Ho, Chan, Tsang, & Suk-Han, 2002) investigated whether different subtypes of
dyslexic readers exist among Chinese children. Ho et al. (in press) examined the
cognitive profiles of 147 children with developmental dyslexia on tests of visual
processing, phonological processing including phonological awareness, rapid
automatized naming (RAN), and orthographic processing, on the assumption that each
of these skills measures dissociable abilities, each related in a different way to reading
ability. Specifically, they assumed that: visual processing skills may be important for
rote learning of whole characters; phonological processing skills may be especially
important for deriving the pronunciations of words on the basis of the phonetic
component; the RAN tests estimate the ability to rapidly make connections between
orthographic patterns and their phonological representations, a skill that has been
proposed to underlie the acquisition of orthographic representation (cf. Bowers et al.,
1999) and, the orthographic processing tests, which required various responses to
printed characters, reflected children’s ability to learn about orthographic conventions
(such as the legal ordering and orientation of the components of Chinese characters).
Surprisingly, in contrast to some studies described earlier, only 29% of the
children presented with a phonological awareness deficit. The deficit affecting the
largest proportion of children was in RAN (57%), followed by orthographic
processing (42%). Visual deficits were less common (27%), and they were most
25
weakly associated with reading. This pattern of results was similar to that obtained in
a study with a smaller sample of children (Ho et al., 2002). Moreover, RAN and
orthographic processing made significant unique contributions to literacy performance
whereas phonological processing and visual skills did not. An interesting finding was
that the RAN deficit never occurred in isolation; that is when a child was deficient in
RAN, s/he also had a deficit in at least one other skill, most frequently orthographic
processing. The authors interpreted the results to mean that the primary cause of
dyslexia in Chinese might be a problem in acquiring orthographic knowledge and
representations as a consequence of RAN and orthographic processing deficits.
Phonological awareness deficits, while present in a non-negligible minority of poor
readers, may be of secondary importance due to the fact that Chinese has a
morphosyllabic and not an alphabetic writing system.
These findings seem to contradict a number of earlier claims about the
cognitive processes underlying reading in Chinese and the causes of reading failure.
A positive aspect of the study was that a variety of potentially important component
skills of reading were assessed in a single cohort of children. However, Ho et al. (in
press) did not report reliability coefficients for the measures used, though they did so
for essentially the same battery of tests in their earlier, smaller-scale version of the
above study (Ho et al., 2002). In that study, by far the least reliable measures (well
below r = .80) were those assessing phoneme awareness, while the RAN and
orthographic tests were highly reliable. It is possible, therefore, that more children
with dyslexia were found to have RAN and orthographic processing deficits simply
because the tests were more sensitive to individual variations in those skills than were
the tests assessing phonological awareness. Also, until the cognitive processes
underlying the RAN tasks are better understood, it can not be ruled out that
performance on RAN in fact estimates phonological processing (cf. Patel et al., in
press). It is also difficult to interpret the finding of a prevalent orthographic
processing deficit because all the tasks required judgments about the plausibility of
printed characters; that is, they were types of reading tasks. Consequently, it is not
clear whether these measures estimate a specific cognitive ability to set up
orthographic representations or whether the tasks simply reflect children’s poor
characters reading skills.
Clearly, further research will be necessary in order to describe and understand
the causes of dyslexia in Chinese. The number of controlled studies is still small, and
26
the studies that have been carried out with poor readers have focused on a wide
variety of questions. To our knowledge, studies of writing development among
Chinese children with dyslexia have not yet been published, at least in English.
Studies to date suggest that phonological processing skills are associated with reading
difficulties in Chinese as they are in other writing systems. In addition Chinese poor
readers may also suffer impairments in semantic skills, though whether this is true to
a greater extent than would be observed in other writing systems is not yet known (So
& Siegel, 1997). Similarly, little is known about the timing and duration of the
various impairments in the component skills of Chinese reading. However, the study
of Chan and Siegel (2001) suggests that poor readers and younger normal readers rely
mainly on visual skills in the early stages of learning to read, and phonological
processing skills become important after the first year or two of schooling.
Summary and conclusion
Evidence from studies of dyslexia across different languages and writing
systems suggests, that in broad terms, reading impairments present similarly in
English, in other alphabetic writing systems, and in the Chinese logography. That is,
across a wide variety of writing systems, poor readers may be sensitive to the
statistical properties, and to the internal structures of printed words that they already
know, but they are not able to make use of this knowledge productively and fluently
in reading unfamiliar or infrequent words (e.g. Chan & Siegel, 2001) or to read words
more efficiently by processing larger orthographic units (Ziegler et al., 2003).
Although the degree of orthographic transparency influences the rate of learning to
read in both normal and dyslexic readers, dyslexic readers in all languages appear to
have particular difficulties in learning the inconsistencies and irregularities of writing
systems (e.g. Alegria & Mousty, 1994, 1996, Tseng et al., 1995, Wimmer, 1993).
With respect to the cognitive profiles of dyslexic children, the impairments of
those learning alphabetic systems appear more readily comparable than those of
Chinese children. This is to some extent determined by the structure and nature of the
spoken and written languages, as well as by differences in prevalent teaching
methods. For example, due to the importance of semantic information in the Chinese
writing system, Chinese dyslexic individuals may experience semantic difficulties to a
greater extent than their counterparts learning alphabetic orthographies. The converse
27
may be true of phonological deficits in writing systems based on the alphabetic
principle. Also, Chinese children who are taught new characters by rote
memorization may depend more on visual strategies in word recognition than German
or Czech children who are taught to read by phonics methods. However, more studies
in Chinese, and comparisons with English and other alphabetic languages are required
in order to answer these questions about relative differences.
Nevertheless, current evidence suggests that some of the cognitive deficits
underlying dyslexia are universal. First, evidence suggests that visual processing
problems are not a core cause of reading problems, and this seems to be true even in
the visually complex system of Chinese. Second, two major measures of phonological
processing – phonological awareness and verbal short-term memory – have been
found to be deficient to lesser or greater degrees in the vast majority of populations
with dyslexia. Third, a deficit in rapid naming speed (RAN) has been observed in all
studies that have employed this measure. Whether the RAN deficit represents simply
another aspect of a phonological processing deficit, or whether it indicates
dysfunction in a separate phonology-orthography timing mechanism needs to be
resolved by future research.
The study of dyslexia in languages other than English is relatively recent, and
many questions remain. Some important challenges for future research are to clarify
whether dyslexia may occur in the absence of phonological processing deficits in
some languages, or in some subpopulations of dyslexics. Similarly, it will be
important to determine whether or not mild phonological deficits are functionally
important in the profiles of dyslexic individuals, or whether in the context of regular
writing systems or in logographic systems for that matter, they become irrelevant to
the difficulties of dyslexics. Future research may ultimately reveal whether specific
reading and spelling difficulties can be reduced to a common set of cognitive
processing difficulties in all languages. It appears at the moment that the hope of
relating reading problems in all languages to a universal impairment in phonological
processing is nearer than many might have dared to hope a decade ago.
28
Shallow Deep
En
g
lish
French
Dutch
German
Czech
Hebrew
(pointed)
Finnish
Figure 1. A schematic or the relative depth of several alphabetic writing systems.
29
References
Alegria, J., & Mousty, P. (1994). On the development of lexical and nonlexical
spelling procedures of French-speaking normal and disabled children. In G. B.
a. N. Ellis (Ed.), Handbook of spelling: Theory, process and invervention (pp.
211-226). Chichester: John Wiley & Sons.
Alegria, J., & Mousty, P. (1996). The development of spelling procedures in French-
speaking, normal and reading-disabled children: Effects of frequency and
lexicality. Journal of Experimental Child Psychology, 63, 312-338.
Bast, J., & Reitsma, P. (1998). Analyzing the development of individual differences in
terms of Matthew effects in reading: results from a Dutch longitudinal study.
Developmental Psychology, 34, 1373-1399.
Ben-Dror, I., Bentin, S., & Frost, R. (1995). Semantic, phonological, and
morphological skills in children with reading disabilities and normally
achieving readers: Evidence from perception and production of spoken
Hebrew words. Reading Research Quarterly, 30, 876-893.
Bentin, S., & Leshem, H. (1993). On the interaction of phonological awareness and
reading acquisition: it's a two-way street. Annals of Dyslexia, 43, 125-148.
Bowers, P. B., Sunseth, K., & Golden, J. (1999). The route between rapid naming and
reading progress. Scientific Studies of Reading, 3(1), 31-53.
Bowers, P. G., & Wolf, M. (1993). Theoretical links among naming speed, precising
timing mechanisms and orthographic skill in dyslexia. Reading & Writing, 5,
69-85.
Brady, S., Shankweiler, D., & Mann, V. (1983). Speech perception and memory
coding in relation to reading ability. Journal of Experimental Child
Psychology, 35, 345-367.
Breznitz, Z. (1997). Enhancing the reading of dyslexic children by reading
acceleration and auditory masking. Journal of Educational Psychology, 89,
103-113.
Bruck, M., & Waters, G. (1988). An analysis of the spelling errors of children who
differ in their reading and spelling skills. Applied Psycholinguistics, 9, 77-92.
Byrne, B. (1998). The foundation of literacy: The child's acquisition of the alphabetic
principle. Hove: Psychology Press.
Byrne, B., Freebody, P., & Gates, A. (1992). Longitudinal data on the relations of
word-reading strategies to comprehension, reading time, and phonemic
awareness. Reading Research Quarterly, 27, 142-151.
Caravolas, M. (in press). Spelling development in alphabetic writing systems: How
orthographic consistency may influence the acquisition of literacy skills. In M.
Joshi (Ed.), Handbook of orthography and literacy: Lawrence Erlbaum
Publishers.
Caravolas, M., & Bruck, M. (1993). The effect of oral and written language input on
children's phonological awareness: a cross-linguistic study. Journal of
Experimental Child Psychology, 55, 1-30.
Caravolas, M., Bruck, M., & Genesee, F. (2003). Similarities and differences between
English- and French-speaking poor spellers. In N. Goulandris (Ed.), Dyslexia
in different languages: Cross-linguistic comparisons (pp. 157-180). London:
Whurr.
30
Caravolas, M., Hulme, C., & Snowling, M. (2001). The foundations of spelling
ability: Evidence from a 3-year longitudinal study. Journal of Memory and
Language, 45, 751-774.
Caravolas, M., & Volín, J. (2001). Phonological spelling errors among dyslexic
children learning a transparent orthography: The case of Czech. Dyslexia, 7,
229-245.
Chan, C. K. K., & Siegel, L. (2001). Phonological processing in reading Chinese
among normally achieving and poor readers. Journal of Experimental Child
Psychology, 80, 23-43.
Cossu, G., Shankweiler, D., Liberman, I., Katz, L., & Tola, G. (1988). Awareness of
phonological segments and reading ability in Italian children. Applied
Psycholinguistics, 9, 1-16.
de Jong, P. (1998). Working memory deficits of reading disabled children. Journal of
Experiemental Child Psychology, 70, 75-96.
de Jong, P., & van der Leij, A. (1999). Specific contributions of phonological abilities
to early reading acquisition: Results from a Dutch latent variable longitudinal
study. Journal of Educational Psychology, 91(3), 450-476.
de Jong, P., & van der Leij, A. (2002). Effects of phonological abilities and linguistic
comprehension on the development of reading. Scientific Studies of Reading,
6(1), 51-77.
deJong, P., & van der Leij, A. (in press). Develpmental changes in the manifestation
of a phonological deficit in dyslexic children learning to read a regular
orthography. Journal of Eductational Psychology.
Durgunoglu, A. Y., Nagy, W. E., & Hancin-Bhatt, B. J. (1993). Cross-language
transfer of phonological awareness. Journal of Educational Psychology, 85,
453-465.
Frost, R., Katz, L., & Bentin, S. (1987). Strategies for visual word recognition and
orthographical depth: A multilingual comparison. Journal of Experimental
Psychology: Human Perception and Performance, 13, 104-115.
Geva, E., Wade-Wooley, L., & Shaney, M. (1993). The concurrent development of
spelling and decoding in two different orthographies. Journal of Reading
Behavior, 25, 383-406.
Goulandris, N. (Ed.). (2003). Dyslexia in different languages: Cross-Linguistic
Comparisons. London: Whurr Publishers.
Ho, C. S., & Bryant, P. (1997). Phonological skills are important in learning to read
Chinese. Developmental Psychology, 33(6), 946-951.
Ho, C. S.-H., Chan, D. W.-O., Lee, S.-H., Tsang, S.-M., & Luan, V. H. (in press).
Cognitive profiling and preliminary subtyping in Chinese developmental
dyslexia. Cognition.
Ho, C. S.-H., Chan, D. W.-O., Tsang, S.-M., & Suk-Han, L. (2002). The cognitive
profile and multiple-deficit hypothesis in Chinese developmental dyslexia.
Developmental Psychology, 38, 543-553.
Ho, C. S.-H., Law, T. P.-S., & Ng, P. M. (2000). The phonological deficit hypothesis
in Chinese developmental dyslexia. Reading and Writing, 13, 57-79.
Huang, H. S., & Hanley, J. R. (1994). Phonological awareness and visual skills in
learning to read Chinese and English. Cognition, 54, 73-98.
Huang, H. S., & Hanley, J. R. (1997). A longitudinal study of phonological
awareness, visual skills, and Chinese reading acquisition among first-graders
in Taiwan. International Journal of Behavioral Development, 20(2), 249-268.
31
Kozminsky, L., & Kozminsky, E. (1995). The effects of early phonological awareness
training on reading success. Learning and Instruction, 5, 187-201.
Landerl, K. (2001). Word recognition deficits in German: More evidence from a
representative sample. Dyslexia, 7, 183-196.
Landerl, K., & Wimmer, H. (2000). Deficits in phoneme segmentation are not the
core problem of dyslexia: Evidence from German and English children.
Applied Psycholinguistics, 21, 243-262.
Landerl, K., Wimmer, H., & Frith, U. (1997). The impact of orthographic consistency
on dyslexia: A German-English comparison. Cognition, 63, 315-334.
Liberman, I. Y., & Shankweiler, D. (1979). Speech, the alphabet and teaching to read.
In L. Resnick & P. Weaver (Eds.), Theory and Practice of Early Reading.
Hillsdale, NJ: Lawrence Erlbaum Associates.
Mattingly, I. G. (1992). Linguistic awareness and orthographic form. In R. Frost & L.
Katz (Eds.), Orthography, phonology, morphology, and meaning (pp. 11-26).
Amsterdam: Elsevier Science Publishers.
McDougall, S., Hulme, C., Ellis, A. W., & Monk, A. (1994). Learning to read: the
role of short-term memory and phonological skills. Journal of Experimental
Child Psychology, 58, 112-123.
Muller, K., & Brady, S. (2001). Correlates of early reading performance in a
transparent orthography. Reading and Writing, 14, 757-799.
Oney, B., & Goldman, S. R. (1984). Decoding and comprehension skills in Turkish
and English: Effects of the regularity of grapheme-phoneme correspondences.
Journal of Educational Psychology, 76, 557-567.
Patel, T., Snowling, M., & de Jong, P. (in press). A cross-linguistic comparison of
children learning to read in English and in Dutch. Journal of Educational
Psychology.
Porpodas, C. D. (1999). Patterns of phonological and memory processing in
beginning readers and spellers of Greek. Journal of Learning Disabilities, 32,
406-416.
Rack, J. P., Snowling, M. J., & Olson, R. K. (1992). The nonword reading deficit in
developmental dyslexia: a review. Reading Research Quarterly, 27, 29-53.
Schatschneider, C., Carlson, C., Francis, D., Foorman, B., & Fletcher, J. M. (2002).
Relationship of rapid automatized naming and phonological awareness in
early reading development: Implications for the double deficit hypothesis.
Journal of Learning Disabilities, 35, 245-256.
Schneider, W., Roth, E., & Ennemoser, M. (2000). Training phonological skills and
letter knowledge in children at risk for dyslexia: A comparison of three
kindergarten intervention programs. Journal of Educational Psychology, 92,
284-295.
Seymour, P. H., Aro, M., & Erskine, J. M. (2003). Foundation literacy acquisition in
European orthographies. British Journal of Psychology, 94, 143-174.
Share, D. (2003). Dyslexia in Hebrew. In N. Goulandris (Ed.), Dyslexia in different
languages: Cross-linguistic comparisons (pp. 208-234). London: Whurr
Publishers.
Shu, H., & Anderson, R. C. (1997). Role of radical awareness in the character and
word acquisition of Chinese children. Reading Research Quarterly, 32(1), 78-
89.
Shu, H., Chen, X., Anderson, R. C., Wu, N., & Xuan, Y. (2003). Properties of school
Chinese: Implications for learning to read. Child Development, 74(1), 27-47.
32
Snowling, M., Gallagher, A., & Frith, U. (2003). Family risk of dyslexia is
continuous: Individual differences in the precursors of reading skill. Child
Development, 74(2), 358-373.
Snowling, M. J. (2000). Dyslexia (2nd ed.). Oxford: Blackwell.
Snowling, M. J., & Hulme, C. (1989). A longitudinal case study of developmental
phonological dyslexia. Cognitive Neuropsychology, 6, 379-403.
So, D., & Siegel, L. (1997). Learning to read Chinese: Semantic, syntactic,
phonological and working memory skills in normally achieving and poor
Chinese readers. Reading and Writing, 9, 1-21.
Sprenger-Charolles, L., Colle, P., Lacert, P., & Serniclaes, W. (2000). On subtypes of
developmental dyslexia: Evidence from processing time and accuracy scores.
Canadian Journal of Experimental Psychology, 54, 87-103.
Stanovich, K. E. (1992). Speculations on the causes and consequences of individual
differences in early reading acquisition. In P. Gough & L. Ehri & R. Treiman
(Eds.), Reading Acquisition (pp. 307-342). Hillsdale, NJ: Erlbaum.
Stanovich, K. E., Nathan, R. G., & Zolman, J. E. (1988). The developmental lag
hypothesis in reading: longitudinal and matched reading level comparisons.
Child Development, 59, 71-86.
Torgesen, J. K., Wagner, R., Rashotte, C. A., Burgess, S., & Hecht, S. (1997).
Contributions of phonological awareness and rapid automatic naming ability
to the growth of word-reading skills in second to fifth grade children.
Scientific Studies of Reading, 1, 161-185.
Tseng, O. J. L., Zhong, H. L., Hung, D. L., & Lee, W. L. (1995). Learning to be a
conspirator: A talel of becoming a good Chinese reader. In B. d. Gelder & J.
Morais (Eds.), Speech and reading: A comparative approach (pp. 22-246).
Hove, UK: Erlbaum.
Vellutino, F. R., Scanlon, D. M., Fletcher, J. M., & Snowling, M. (in press). Specific
reading disability (dyslexia): What have we learned in the past four decades?
Journal of Child Psychology and Psychiatry.
Vellutino, F. R., Scanlon, D. M., Sipay, E., Small, S., Pratt, A., Chen, R., & Denckla,
M. (1996). Cognitive profiles of difficult-to-remediate and readily-remediated
poor readers: Early intervention as a vehicle for distinguishing between
cognitive and experiential deficits as basic causes of specific reading
disability. Journal of Educational Psychology, 88(4), 601-638.
Vellutino, F. R., Scanlon, D. M., & Spearing, D. (1995). Semantic and phonological
coding in poor and normal readers. Journal of Experimental Child Psychology,
59, 76-123.
Wesseling, R., & Reitsma, P. (2000). The transient role of explicit phonological
recoding for reading acquisition. Reading and Writing, 13, 313-336.
Wimmer, H. (1993). Characteristics of developmental dyslexia in a regular writing
system. Applied Psycholinguistics, 14, 1-33.
Wimmer, H. (1996a). The early manifestation of developmental dyslexia: evidence
from German children. Reading and Writing, 8, 171-188.
Wimmer, H. (1996b). The non-word reading deficit in developmental dyslexia:
Evidence from children learning to read German. Journal of Experimental
Child Psychology, 61, 80-90.
Wimmer, H., & Goswami, U. (1994). The influence of orthographic consistency on
reading development: word recognition in English and German. Cognition,
91-103.
33
Wimmer, H., & Landerl, K. (1997). How learning to spell German differs from
learning to spell English. In C. A. Perfetti & L. Rieben & M. Fayol (Eds.),
Learning to Spell: Research, theory and practice across languages. Mahway,
New Jersey: Lawrence Erlbaum Associates.
Wimmer, H., Landerl, K., Linortner, R., & Hummer, P. (1991). The relationship of
phonemic awareness to reading acquisition: More consequence than
precondition but still important. Cognition, 40, 219-249.
Wimmer, H., & Mayringer, H. (2002). Dysfluent reading in the absence of spelling
difficulties: A specific disability in regular orthographies. Journal of
Educational Psychology, 94(2), 272-277.
Wimmer, H., Mayringer, H., & Landerl, K. (2000). The double-deficit hypothesis and
difficulties in learning to read a regular orthography. Journal of Educational
Psychology, 92(4), 668-680.
Woo, E. Y., & Hoosain, R. (1984). Visual and auditory functions of Chinese
dylsexics. Psychologia, 27, 164-170.
Zhou, Y. G. (1978). To what extent are the "phonetics" of present-day Chinese
characers still phonetic. Zhonggou Yuwen, 146, 172-177.
Ziegler, J. C., Jacobs, A. M., & Stone, G. O. (1996). Statistical analysis of the
bidirectional inconsistency of spelling and sound in French. Behavior
Research Methods, Instruments, & Computers, 28(4), 504-515.
Ziegler, J. C., Perry, C., Ma-Wyatt, A., Ladner, D., & Schulte-Korne, G. (2003).
Developmental dyslexia in different languages: Language-specific or
universal? Journal of Experimental Child Psychology, 86, 169-193.
