Content uploaded by Salim Abu-Rabia
Author content
All content in this area was uploaded by Salim Abu-Rabia on Feb 11, 2015
Content may be subject to copyright.
Reading, Syntactic, Orthographic, and Working
Memory Skills of Bilingual Arabic-English
Speaking Canadian Children
Salim Abu-Rabia1,3 and Linda S. Siegel2
This study assessed the reading, language, and memory skills of 56 bilingual Arab-Canadian children
age’s 9–14. English was their main instructional language, and Arabic was the language spoken at
home. All children attended a Heritage Language Program in Toronto where they were taught to read
and write Arabic. The children were administered word and pseudo-word reading, language, and
working memory tests in English and Arabic. The majority of the children showed at least adequate
proficiency in both languages. There was a significant relationship between the acquisition of word
and pseudo-word reading working memory, and syntactic awareness skills in the two languages. The
poor readers in Arabic had lower scores on all linguistic tasks, except the visual task. There were no
significant differences between bilingual English Arabic children and monolingual English-speaking
children on the reading, language, and memory tasks. However, bilingual English Arabic children
who had reading problems in English had higher scores on English pseudo-word reading and
spelling tasks than monolingual English-speaking children with reading disabilities, probably because
of positive transfer from the regular nature of Arabic orthography. In this case, bilingualism does
not appear to have negative consequences for the development of language reading skills in both
languages—Arabic and English—despite the different nature of the two orthographies.
KEY WORDS: Bilingual Arabic-English; bilingual reading disabilities; monolingual reading dis-
abilities; reading disabilities in different orthographies.
INTRODUCTION
Bilingualism is a common phenomenon for many children in Canada. The
children of immigrants speak and hear their first language in their homes
661
0090-6905/02/1100-0661/0 © 2002 Plenum Publishing Corporation
Journal of Psycholinguistic Research, Vol. 31, No. 6, November 2002 (
©
2002)
This research was supported by a grant form the Natural Sciences and Engineering Research
Council of Canada to Linda S. Siegel.
1Faculty of Education, University of Haifa, Mount Carmel, Israel.
2Educational and Counselling Psychology and Special Education, University of British
Columbia, Vancouver, B. C., Canada.
3To whom all correspondence should be addressed. Faculty of Education, University of Haifa,
Mount Carmel, Haifa 31905, Israel. Fax: 972-4-240911; email: salimar@construct.Haifa.ac.il
and neighborhoods, but they are educated in the language of the majority.
Canada has two official languages, English and French. In addition, multicul-
turalism is an official policy of Canada. In many communities across Canada
this policy means that children who come from homes in which neither
English nor French is spoken receive most of their schooling in English or
French (depending on the area of the country where they live), but they also
receive some instruction in their heritage or home language. Little is known
about the nature and extent of English (or French) and heritage language
proficiency of these children.
There are two major hypotheses about the relationship between skills
in first and second languages. According to the linguistic interdependence
hypotheses developed by Cummins (1979) concerning learning the first
language (L1) and the second language (L2), children who have learning
problems in L1 will show similar problems in L2. Further, academic skills
acquired successfully in L1 will be transferred to L2. Learning L1 and L2
does not hinder the progress of either, and, in fact, may enhance both.
Alternatively, there is the script-dependent hypothesis, namely, that the
characteristic of different scripts may result in different reading and writing
problems emerging in the two languages. For example, English does not
have a one-to-one relation between graphemes and phonemes; words are not
always pronounced as they are spelled and there are many irregularities.
However, Arabic has much more predictable grapheme–phoneme corre-
spondence rules than English, so the nature of the problems in the two
languages may be quite different.
There have been a number of studies of the relationship between
bilingualism and reading. Da Fontoura and Siegel (1995) examined the
nature of language, memory, and reading skills of bilingual Portuguese-
Canadian students. The children were administered word and pseudo-word
reading, language, and working memory tasks in English and Portuguese.
The majority of the children showed at least adequate proficiency in both
languages on those tasks. The children who had low reading scores in
English also had significantly lower scores on the Portuguese tasks. The
Portuguese-Canadian children who were normal-achieving readers did not
differ from a comparison group of monolingual English-speaking normal-
achieving readers on the above tasks. In both English and Portuguese,
reading difficulties appeared to be related to deficits in phonological pro-
cessing, working memory, and syntactic awareness.
In support of Cummins’ hypothesis in a study of Berber- and Arabic-
speaking children educated in Arabic in Morocco, Wagner, Spratt, and Ezzaki
(1989) found that early differences in reading and language skills of the two
groups of children disappeared by the later grades. Furthermore, a study by
Geva and Siegel (2000) found significant relationships among performances
662 Abu-Rabia and Siegel
on word identification, pseudo-word reading, working memory and syntactic
awareness tasks in English and Hebrew.
A study conducted by Durgunoglu, Nagy, and Hancin-Bhatt (1993)
investigated cross-language transfer of phonological awareness of bilingual
Spanish-English beginning readers. Participants were 31 Spanish-speaking
first-grade students. They were administered tests of letter naming, Spanish
phonological awareness. Spanish and English word recognition, and
Spanish and English oral proficiency. Performance on English word and
pseudo-word recognition tests were correlated with levels of both Spanish
phonological awareness and Spanish word recognition, thus indicating
cross-language transfer. These results confirm the importance of phonemic
awareness in first and second-language acquisition (e.g., Stanovich, 1982;
Tunmer, Herriman, & Nesdale, 1988; Wagner, 1988; Yopp, 1988).
In support of the script-dependent hypothesis, Ryan and Meara (1991)
reported a study that investigated the hypothesis that Arabic-speaking learn-
ers of English tend to rely heavily on consonants when attempting to recog-
nize English words, because of the different orthography of Arabic, their L1.
The task was to recognize the missing vowels in English words. The Arabic
speakers tended to confuse English words with similar structures; they also
made more errors in the reading and spelling tests than non-Arabic-speaking
individuals learning English as a second language and native English speak-
ers. The results of Ryan and Meara (1991) provide support for the idea that
different orthographies may lead to different types of errors in learning L2.
This highlights the importance of taking different writing systems into con-
sideration as a potential factor to be related to the development of L2.
Although Arabic is considered a transparent language with a pre-
dictable relationship between letters and sounds, there are still irregularities,
with several vowels and letter-sound pronunciations and irregular writing
rules for many letters. These considerations are serious for beginning read-
ers who have not yet mastered all vowels and letter-sound combinations,
and whose relative literacy weakness prevents them from relying on textual
context for word prediction and comprehension. Thus the script-dependent
hypothesis would lead to the prediction that the irregularity of grapheme–
phoneme conversion rules in English would create additional reading and
writing problems for these bilingual Arab students. Cummins’ (1979) hypoth-
esis, by contrast, leads to the prediction that similar problems should be
manifested in both languages because of central processing deficit.
The purpose of this study was to examine in a group of bilingual
Arabic- and English-speaking children the relationship between reading,
writing, phonological, syntactic, orthographic, and memory skills in English
and Arabic as two different orthographies. The children came form Arabic-
speaking countries. In Canada, their schooling was in English, and they
Reading, Syntactic, Orthographic, Working Memory Skills 663
received instruction in speaking, writing, and reading Arabic for 3 hours a
week in Arabic “Heritage” Language programs.
This study examined some processes that are significant in the devel-
opment of reading skills in English and have not been studied in the
Arabic orthography. These are phonological and syntactic awareness and
orthographic and working memory processes. Phonological processing
involves the association of sounds with letters, that is, the understanding of
grapheme–phoneme conversion rules and their exceptions, which is the basis
of decoding print (Stanovich 1988a, 1988b). Syntactic awareness refers to
the ability to understand the syntax of the language, which is critical for effi-
cient reading of text. Working memory refers to the retention of information
in short-term storage while processing incoming information and retrieving
information from long-term storage. Orthographic knowledge refers to the
ability of children to recognize the typical visual patterns (e.g., possible let-
ter sequences of the language). The development of these processes was
compared in normal and disabled readers in both English and Arabic.
METHOD
Subjects
The subjects were 56 students from metropolitan Toronto, a predomi-
nantly English-speaking area of Canada. They were 9–14 years old, 32
females and 24 males, in grades 4–8. All the children were born outside
Canada and had lived more than 2 years in Canada. The majority of the
children came from the lower socioeconomic level. In 10 families, the par-
ents worked in skilled jobs such as teaching and nursing or were the own-
ers of small businesses, and the remaining parents worked in unskilled or
semiskilled jobs or not at all. Most of the parents had very little education.
The language spoken at home was Arabic, but all children had English as
their instructional language in Canadian schools. The children attended
Arabic Heritage Language programs for approximately 3 hours per week,
where they received instructions in reading, writing, and speaking Arabic.
A comparison group of monolingual English-speaking children were
selected from a larger sample who had been administered the tests in
English (Siegel, 1988; Siegel & Ryan, 1989a, b; Stanovich & Siegel, 1994).
These children were matched on age to the Arabic-English bilingual chil-
dren in the 9–11-year-old range. There were 45 monolingual normal read-
ers (Wide Range Achievement Test, WRAT, Jastak & Wilkinson, 1984,
Reading Percentile ⬎30) and 20 monolingual reading-disabled children
(WRAT, Reading Percentile ⬍25). For the purpose of statistical analysis,
the children were divided into two age-groups, 9–11 and 12–14.
664 Abu-Rabia and Siegel
TESTS AND TASKS
The following tests and tasks were administered.
English Tasks
Wide Range Achievement Test-Revised (WRAT-R, Jastak & Wilkinson,
1984). The Reading, Spelling, and Arithmetic subtests were used. The
Reading subtest requires the child to read an increasingly difficult series of
words. In the Spelling subtest the child is required to write the correct
spelling of an increasingly difficult series of words. The Arithmetic subtest
requires the individual to perform an increasingly difficult series of arith-
metic calculations.
Woodcock Reading Mastery Test (Woodcock, 1973). The Word Attack
and Word Identification subtests were used. In the Word Attack subtest the
child is required to read an increasingly difficult series of pseudo-words
such as cyr, roo, and wrey according to the phonological rules of English.
English Oral Cloze Test (Siegel & Ryan, 1988). The children were
instructed to fill in the missing word in each of 20 sentences presented orally.
An example of a sentence is: “Jane her sister ran up the hill.”
English Working Memory Test (Siegel & Ryan, 1989a). The children
were presented with sentences orally with the final word missing; they had
to supply it and repeat all the missing words from the set. There were three
trials in each set size (2, 3, 4, and 5). Examples of sentence are: “In sum-
mer it is very .” “People go to see monkeys in a .” “With
dinner we sometimes eat bread and .” The child was required to
repeat the three words he or she selected, in this case “hot,” “zoo,” “but-
ter,” in the same order that they had been presented. To minimize word-
finding problems, the sentences were chosen so that the word was virtually
predetermined.
English Orthographic Test (Seigel, Share, & Geva, 1995). Seventeen
pairs of pseudo-words were presented. Only one of each pair could be the
spelling of a word. Examples of these paired pseudo-words are: filv-filk,
and tolz-tolb. The child was instructed to select the member or pair that
could be “or looked like” an English word.
English Phonological Test (Olson, Kliegel, Davidson, and Foltz, 1985).
Twenty-six pairs of pseudo-words were presented; one word of the pair was
a pseudo-homophone and one a pseudo-word. The child was instructd to
select the word in each pair that “sounds like a real word.” Examples of
these pairs are: saip-saif and seaf-seet.
English Visual Test (Olson et al., 1985). Twenty-six pairs of words
were presented: one word of the pair was a pseudo-homophone and one was
Reading, Syntactic, Orthographic, Working Memory Skills 665
666 Abu-Rabia and Siegel
correctly spelled. The children were instructed to select the correctly spelled
word. Examples of these stimuli are: sheep-sheap and face-fase.
Arabic Tasks
Arabic Word Reading. These test involved two lists of words that stu-
dents were asked to read aloud. One list contained 45 Arabic words with
vowels (parallel to the English WRAT-R list) and the other 106 Arabic
words with vowels (parallel to the English Woodcock test list). The words
were selected from Arabic readers used in Arabic primary and intermediate
schools in Israel.
Arabic Spelling Test. The children had to write an increasingly diffi-
cult series of 45 Arabic words (parallel to the English WRAT-R spelling
words). The words were selected from the Arabic readers used in Arab pri-
mary and intermediate schools in Israel.
Arithmetic (Arabic version). The WART-R Arithmetic test was pre-
sented to the children in Arabic.
Arabic Word Attack Test. The children were required to read an
increasingly difficult series of Arabic pseudo-words.
Arabic Oral Cloze Test. The children were isntructed to fill in the
missing words in each of the sentences presented orally. Critical features of
Arabic were tested.
Arabic Working Memory Test. The children were presented orally with
Arabic sentences lacking the final word, which they had to supply; they also
had to repeat all the missing words from the set. There were three trials with
each set size (2, 3, 4, and 5). The children were required to repeat the 2, 3,
4, or 5 words they selected in the same order that they had been presented.
Arabic Orthographic Test. Seventeen pairs of pseudo-words were pre-
sented. Only one of each pair could be the spelling of a word because the
incorrect one contained an orthographic combination that never occurred in
Arabic.
Arabic Visual Test. Twenty-six pairs of words were presented; one
word of the pair was a pseudo-homophone and one was correctly spelled.
Consistent with adaptation of the English language tests, we attempted
to develop an Arabic phonological test parallel to the English version.
Unfortunately, this particular test was very difficult to adapt because of col-
loquial and phonological Arabic inflections.
Procedure
The children were tested individually at home. The English and Arabic
tests were administered to all the children. The English tests were administered
Reading, Syntactic, Orthographic, Working Memory Skills 667
as follows: working memory, orthographic, oral cloze, phonological, visual,
Woodcock reading, and WRAT-R reading. The Arabic tests were adminis-
tered as follows: working memory, orthographic, oral cloze, visual condition
word attack, reading, and WRAT-R arithmetic. The order of language of pre-
sentation was randomly assigned to either English first or Arabic first.
RESULTS
The intercorrelations among tests are shown in Tables I, II, and III. As
these tables show, there were statistically significant correlations between
the English and the Arabic tests, except for the visual tests. There were sig-
nificant correlations between the word and pseudo-word reading tests within
and between the two languages, suggesting the operation of similar pro-
cesses in both languages.
The oral cloze tests were significantly correlated with all reading tests
in both languages; the Arabic and English working memory tests were also
significantly correlated with each other and with all reading tests, except for
the visual tests in both languages.
Cognitive Processes as a Function of a Reading Disability
The children were divided into two groups based on their score on the
WRAT-R reading subtest in English. Normal-achieving readers were
defined as those with scores greater than 30 percentile; there were 41 of
these children (ages 9–14). Eleven children (ages 9–14) whose scores were
less than 30 percentile on the WRAT-R reading subtest were considered
Table I. Intercorrelations of the English Tests
Word Word Phono- Ortho- Oral Working
identification attack logical Spelling Visual graphic cloze memory Arithmetic
WRAT .56 .44 .49 .55 n.s. .43 .57 .48 .61
reading
Word .59 .40 .57 n.s. .58 .66 .55 .49
identification
Word attack .50 .42 n.s. .50 .63 .49 .49
Phonological .55 n.s. .65 .73 .79 .50
Spelling n.s. .53 .39 .45 .43
Visual n.s. n.s. n.s. n.s.
Orthographic .61 .67 .39
Oral cloze .67 .44
Working .53
memory
aAll values are significant at the level of p⬍.01.
reading disabled. The children were also divided into two age-groups, 9–11
and 12–14. Because the number of the second age-group (12–14) was too
small to be divided in subgroups based on performance on English tests, the
differences on these tests were not calculated. The mean scores on the
English tests for the 9–11 gourp are shown in Table IV. There were signif-
icant differences between the groups in the English tests: oral cloze, word
attack, phonological task, arithmetic, spelling, and word reading. There
were no significant differences between the groups on the orthographic tests
and the visual task.
Table V presents the mean scores of the Arabic tests based on the defin-
ition of a reading disability using the WRAT and shows significant differences
between the groups in some of the Arabic tests: oral cloze, orthographic, word
identification, word reading, word attack, spelling, and arithmetic. There weas
no significant difference between the groups on the working memory test and
the visual test. Therefore the children who would be classified as reading-
disabled in English showed significantly poor performance in most of their
reading, spelling, and language tests in both languages.
The group was arbitrarily divided into good and poor readers in Arabic
based on their scores in Arabic word reading. Children (n⫽11) with scores
in the Arabic reading test less than 44 were labeled poor readers (age range
9–11); children (n⫽18) with scores in the Arabic reading test greater than
45 were labeled good readers (age range 9–11). Comparisons are shown in
Tables VI and VII. Table VI presents performances on Arabic tasks, and
Table VII presents performance on English tasks. Tables VI and VII show
that the good readers had significantly higher scores in every test, except
the visual tests in both English and Arabic.
668 Abu-Rabia and Siegel
Table II. Intercorrelations of the Arabic Tests
Word Word Phono- Ortho- Oral Working
identification attack logical Spelling Visual graphic cloze memory Arithmetic
WRAT .42b.34b.35c.42bn.s. .34b.45b.37b.43b
reading
Word .62c.60c.45bn.s. .49c.43b.35b.58c
identification
Word attack .49c.38bn.s. .49c.57c.35b.32b
Phonological .38bn.s. .51c.55c.34b.32b
Spelling n.s. .39b.39b.42b.37b
Visual n.s. n.s. n.s. n.s.
Orthographic .41b.60c.29a
Oral cloze .51c.30a
Working .32b
memory
ap⬍.05, bp⬍.01, cp⬍.001.
n.s., not significant.
Reading, Syntactic, Orthographic, Working Memory Skills 669
Table III. Intercorrelations of English Test Scores with Arabic Test Scores
English
Word Word Oral Working
Arabic WRAT identification attack Phonological Spelling Visual Orthographic cloze memory Arithmetic
WRAT .85c.40b.30a.27a.39bn.s. n.s. .42b.32a.54c
Word Identification .36b.82c.75c.48b.52cn.s. .46b.48b.56c.49b
Word attack .33a.69c.74c.40b.46bn.s. .42b.37b.34a.43b
Phonological .37b.36b.60c.50c.45bn.s. .40b.36b.34a.40b
Spelling .55c.56c.70c.75c.57cn.s. .48b.48b.39b.44b
Visual n.s. .69c.74cn.s. n.s. n.s. n.s. n.s. n.s. .43b
Orthographic .43b.39b.37b.60c.48bn.s. .74c.43b.53c.18a
Oral cloze .28a.45b.47b.76c.35an.s. .57c.67c.56b.26a
Working memory .47b.54c.40b.79c.60cn.s. .64c.65c.92c.44b
Arithmetic .53c.85c.59cn.s. .32a.52c.38b.44b.55c.90c
ap⬍.05, bp⬍.01, cp⬍.001.
n.s., not significant.
670 Abu-Rabia and Siegel
Table IV. Mean and Standard Deviations in the English Tests
as a Function of Group (Age Range 9–11)
English tests Normally achieving (n⫽20) Reading disabled (n⫽9)
WRAT reading 53.0c10.1
(17.1) (3.6)
Orthographic 12.3 11.0
(1.7) (2.0)
Visual 20.4 19.7
(1.8) (1.7)
WRAT spelling 45.9c20.7
(15.4) (6.4)
Oral cloze 15.6c9.7
(2.0) (3.0)
Working memory 8.5b5.1
(3.7) (2.1)
Phonological 21.0a16.9
(4.7) (4.9)
Woodcock word attack 47.3a24.4
(16.2) (6.8)
Woodcock word reading 46.1d12.7
(17.5) (6.3)
Arithmetic 30.1a27.2
(3.7) (3.6)
ap⬍.05, bp⬍.01, cp⬍.001, dp⬍.0001.
Table V. Mean and Standard Deviations in the Arabic Tests
as a Function of Group (Age Range 9–11)
Tests Normally achieving (n⫽20) Reading disabled (n⫽9)
Word Reading 54.3b43.2
(11.9) (8.9)
Orthographic 9.2b6.7
(2.2) (2.6)
Visual 13.3 12.9
(1.3) (1.7)
Spelling 549b27.8
(6.8) (7.0)
Oral Cloze 11.5c7.8
(2.0) (2.9)
Working memory 9.2 8.1
(3.7) (2.1)
Word attack 22.3b15.6
(5.3) (5.7)
Word identification 54.6b41.1
(11.6) (11.4)
Arithmetic 48.2a28.3
(18.3) (21.3)
ap⬍.05, bp⬍.01, cp⬍.001.
Reading, Syntactic, Orthographic, Working Memory Skills 671
Table VI. Comparison of Good and Poor Readers Defined by Their
Performance on Arabic Tests (Age Range 9–11)
Arabic tests Good readers (n⫽18) Poor readers (n⫽11)
Word reading 54.9b44.3
(11.2) (10.9)
Orthographic 9.7b6.4
(2.3) (1.7)
Visual 13.7 12.3
(1.1) (1.5)
Spelling 34.8b23.8
(6.6) (12.7)
Oral cloze 11.9b7.9
(2.3) (2.4)
Working memory 9.1a7.0
(4.4) (4.6)
Word attack 23.8b13.4
(4.4) (5.2)
Word identification 29.5a27.2
(3.6) (4.3)
Arithmetic 58.9d36.5
(7.9) (4.6)
ap⬍.05, bp⬍.01, cp⬍.001, dp⬍.0001.
Table VII. Comparison of Good and Poor Readers Defined by Their
Performance on English Tests (Age Range 9–11)
English tests Good readers (n⫽18) Poor readers (n⫽11)
WRAT reading 53.0c28.8
(19.5) (12.7)
Orthographic 12.6b6.4
(1.8) (1.7)
Visual 20.5 18.4
(1.9) (5.2)
WRAT spelling 45.9c36.7
(5.4) (8.5)
Oral cloze 15.79b7.9
(2.0) (2.4)
Working memory 9.1b6.3
(3.8) (4.6)
Phonological 21.5b12.3
(.80) (1.5)
Woodcock word reading 70.3a27.2
(9.0) (4.3)
Woodcock word attack 42.6a21.5
(3.7) (10.9)
Arithmetic 30.1c22.3
(3.6) (4.6)
ap⬍.05, bp⬍.01, cp⬍.001.
672 Abu-Rabia and Siegel
The other group (age range 12–14) was also arbitrarily divided into
good and poor readers in Arabic based on their scores in the Arabic word
reading. Children (n⫽9) with scores below 35 were labeled poor readers;
children (n⫽18) with scores above 36 were labeled good readers. The
comparisons are shown in Tables VIII and IX.
As seen in Table VIII, the good readers had significantly higher scores
in the orthographic, phonological, word attack, word identification, arith-
metic, and spelling tests, while the oral cloze, working memory, and visual
tests did not show any significant differences between groups. Further,
Table IX shows that there are similar results with the English tests, and the
results of the 12–14 age-group are similar to the results of the 9–11 age-
group, with significant differences on every variable except the visual and
working memory tasks.
Comparison with Monolingual English Normal and Disabled Readers
The performance of bilingual Arabic-English normal and disabled read-
ers was compared with the performance of English monolingual normal and
disabled readers for the 9–11-year-old age range. These results are shown in
Table VIII. Comparison of Good and Poor Readers Defined by Their
Performance on Arabic Reading Tests (Age Range 12–14)
English tests Good readers (n⫽14) Poor readers (n⫽9)
WRAT reading 53.0 28.8
(19.5) (12.7)
Orthographic 14.5a8.3
(2.1) (2.1)
Visual 22.5 20.5
(2.2) (4.4)
WRAT spelling 57.1c23.2
(7.7) (6.5)
Oral cloze 17.2 15.7
(2.2) (2.0)
Working memory 11.1 10.3
(4.4) (5.6)
Phonological 22.0a10.0
(4.1) (4.2)
Woodcock word attack 53.3b27.2
(15.7) (11.1)
Woodcock word reading 56.3b32.2
(12.9) (15.5)
Arithmetic 56.2b27.5
(12.5) (7.9)
ap⬍.05, bp⬍.01, cp⬍.001.
Reading, Syntactic, Orthographic, Working Memory Skills 673
Table X. Note that the majority of the children had no difficulties in English,
even though it was their L2, and they were defined as normal readers (⬎30
percentile on their English WRAT-R reading) in this study. Their scores in
the word and pseudo-word reading, spelling, and working memory tests were
equivalent to those of monolingual English normal readers.
When the bilingual reading-disabled children were compared with
monolingual English reading-disabled children the scores of the two groups
were not significantly different in the word reading, working memory, oral
cloze, visual, and orthographic tests. However, the bilingual reading-disabled
children had significantly higher scores than the monolingual children in the
WRAT spelling test and the Woodcock word attack test involving reading
pseudo-words. The bilingual reading-disabled children had significantly
lower scores in the oral cloze test than the monolingual children.
DISCUSSION
The relationships among phonological skills, syntactic awareness, and
working memory are consistent with the findings of Da Fontoura and Siegel
(1995), Geva and Siegel (2000), and Siegel and Ryan (1988). The strong
Table IX. Comparison of Good and Poor Readers Defined by Their
Performance on Arabic Reading Tests (Age Range 12–14)
Tests Good readers (n⫽14) Poor readers (n⫽9)
Word reading 63.4b25.8
(15.0) (16.7)
Orthographic 12.7a7.2
(2.6) (3.1)
Visual 14.1 14.4
(1.2) (1.2)
Spelling 26.7b13.9
(4.2) (5.8)
Oral cloze 13.4a8.9
(2.5) (3.2)
Working memory 6.0 5.1
(4.6) (1.2)
Word attack 26.9b13.1
(5.5) (5.7)
Word identification 48.0b22.9
(10.3) (5.1)
Arithmetic 30.4a27.7
(3.0) (7.4)
ap⬍.05, bp⬍.01.
relationship between pseudo-word reading and word recognition across lan-
guages indicates that the phonic skills are a significant component of read-
ing of the particular alphabetic languages examined in this study. The high
correlations between reading skills in English and Arabic found in this
study were similar to the correlations between reading in Berber and Arabic
found by Wagner et al. (1989) and are consistent with the results obtained
by Durunoglu et al. (1993) about cross-language transfer of phonological
awareness of bilingual Spanish-English beginning readers.
Because English and Arabic reading, language, and memory skills
were highly correlated, the relationships are suggestive of individual differ-
ence variables as the significant determinants of reading skills, rather than
difficulties that are language dependent. Thus the data from this study are
consistent with the linguistic interdependence hypothesis. Moreover, the
data are generally consistent with the hypothesis that bilingual children with
reading problems in English are likely to show problems in their other lan-
guage, in this case Arabic. Similar problems are evident in both languages
for the children who had difficulties. This close relationship between the
two languages is suggestive of general language deficit in some children.
674 Abu-Rabia and Siegel
Table X. The Mean Scores of English-Canadian and Arab-Canadian Normal-Achieving
Children and Reading-Disabled in Reading, Language, and Memory Tests (Age Range 9–11)
Normal achieving Reading disabled
English-Canadian Arab-Canadian English-Canadian Arab-Canadian
(n⫽45) (n⫽18) (n⫽20) (n⫽11)
Arithmetic 33.7 30.1 28.4 27.2
(5.7) (3.7) (6.2) (3.6)
WRAT spelling 44.2 42.1 10.5b20.7
(8.5) (5.5) (6.2) (6.4)
WRAT reading 58.0 53.1 8.9 10.1
(13.7) (17.1) (4.2) (3.6)
Woodcock word 48.3 46.1 14.1 12.3
identification (15.2) (17.5) (5.2) (6.3)
Woodcock word attack 51.1 47.30 17.2a24.4
(16.3) (16.2) (8.2) (6.8)
Working memory 7.9 8.5 5.8 5.1
(2.7) (3.7) (3.2) (2.1)
Oral cloze 16.4 15.7 17.1a12.0
(2.1) (1.9) (4.2) (3.0)
Visual 21.2 20.4 17.2 19.7
(2.3) (1.8) (4.2) (1.8)
Orthographic 16.3a12.3 10.4 10.8
(2.0) (1.7) (3.5) (2.0)
Phonological 20.1 21.0 13.2a16.9
(3.7) (4.7) (4.2) (4.9)
ap⬍.05, bp⬍.01.
The results are suggestive of the importance of certain cognitive pro-
cesses in the development of reading skills as outlined by Siegel (1993).
Phonological processing skills, as measured by pseudo-word reading, are
highly correlated with word recognition skills in both English and Arabic.
Disabled readers in Arabic show the same difficulties with phonological
processing as do disabled readers in English. In both English and Arabic,
deficits in working memory and syntactic awareness are also characteristic
of individuals with a reading disability.
The significant correlations among reading skills, syntactic awareness
tests, and working memory tests in the same language were significant, as
were cross-language correlations. Therefore, relationships between reading
and syntactic awareness were not language-specific, which supports the cen-
tral deficit hypothesis. However, the visual tests in English and Arabic did
not show significant correlations in the same-language and cross-language
correlations. These results highlight the role of phonological rather than
visual difficulties as the significant components of a reading disability.
The results of this study are analogous to those of Lambert and Tucker
(1972), Bank and Swain (1975), and McDougall and Bruck (1976) in their
studies of French immersion in Canada. In the French immersion programs,
native English-speaking children receive reading instruction in French.
English instruction is started several grades later, but these children catch
up with monolingual English-speaking children who receive instruction in
English. There appears to be some positive transfer from French to English
in the case of reading skills. However, the French skills of these children
were behind those of monolingual French-speaking children. For the bilin-
guals in the present study, if we consider Arabic as their first language and
English as their main instructional language, the results are even more
encouraging than those for the children in the French immersion programs.
For the children with normal reading skills, all the reading and language
skills that were measured were equivalent to those of monolingual English-
speaking children, with the exception of the English orthographic test,
which is not surprising because Arabic-English bilinguals had less exposure
to English orthography. However, they performed as well as the mono-
linguals in the spelling, word attack, oral cloze, and phonological tests.
Similarly, in a study of Turkish-speaking children in the Netherlands,
Verhoeven (1990) found that they were not as proficient as the native
Dutch-speaking children in word and pseudo-word reading and sentence
imitation tests. However, many of the Turkish-speaking children displayed
adequate proficiency.
The reading-disabled Arabic-English bilinguals had higher scores in
the English pseudo-word reading, word spelling, and some of the phono-
logical tests than a comparison group of monolingual English reading-
disabled children. This finding is similar to the results from a study by Da
Reading, Syntactic, Orthographic, Working Memory Skills 675
Fontoura and Siegel (1995) in which it was found that bilingual English-
Portuguese–speaking reading-disabled children had higher scores in English
spelling and pseudo-word reading tasks than monolingual English-speaking
reading-disabled children. This finding may reflect a positive transfer from
the more predictable grapheme–phoneme conversion rules of Arabic to the
very opaque orthography of English. It is unlikely that socioeconomic fac-
tors were responsible because the socioeconomic level of the Arab bilin-
guals was lower than that of the English monolinguals.
The results of this study show that bilingualism is clearly not an imped-
iment to the development of reading, syntactic, and memory skills. They pro-
vide support for the Heritage Language Program in Toronto (Cummins,
1979, 1989; Cummins & Danesi, 1990). Most of the children from Arabic-
speaking homes who were being educated in English but received some
instruction in Arabic performed well in the reading memory and language
tests in both English and Arabic. Some of the Arabic-speaking children did
show reading problems in English, but they showed similar problems in
Arabic. Of course, it is impossible to determine how the reading and spoken
language skills of these children compare with children in Arab countries
who have received all their instruction in Arabic.
Although this study was only conducted with one language group, the
results can be viewed as encouraging for the Canadian Heritage Language
programs and for multiculturalism in general. Obviously, more research must
be conducted before a conclusive statement can be made. It is important to
note that most of the bilingual Arab-Canadians in this study showed levels of
reading, syntactic awareness, and verbal working memory skills similar to
those of monolingual English-speaking Canadian children. The bilingual chil-
dren who did have problems had similar difficulties in English and Arabic.
In a meta-analysis of studies of bilingual education, Willig (1985) con-
cluded that “participation in bilingual educational programs consistently
produced small to moderate differences favoring bilingual education for
tests of reading, language skills, mathematics and total achievement when
the tests were in English, and for reading, language, mathematics, writing,
social studies, listening comprehension and attitudes towards school or self
when tests were in other languages” (p. 269). The results of this study sup-
port this conclusion.
REFERENCES
Adams, M. J. (1990). Beginning to read: Thinking and learning about print. The Reading
Research and Education Center. University of Illinois at Urbana-Champaign.
Bank, H. C., & Swain, M. (1975). Three-year evaluation of a large scale early grade French
immersion program: The Ottawa study. Language Learning, 25, 1–30.
676 Abu-Rabia and Siegel
Reading, Syntactic, Orthographic, Working Memory Skills 677
Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual
children. Review of Educational Research, 49 222–251.
Cummins, J. (1989). Empowering minority students. Sacramento, CA: California Association
for Bilingual Education, Sacramento, CA.
Cummins, J., & Danesi, M. (1990). Heritage languages: Toronto, Canada Our Schools
Ourselves Education Foundation.
Cummins, J., Swain, M., Nakajima, K., Handscombe, J., Green, D., & Iran, C. (1984). Linguistic
interdependence among Japanese and Vietnamese immigrant students. In: C. Rivera (Ed.),
Communicative competence approaches to language proficiency assessment: Research and
application (60–81). England: Multilingual Matters Ltd.
De Fontoura, H. A., & Siegel, L. S. (1995). Reading, syntactic and working memory skills of
bilingual Portuguese-English Canadian children. Reading and Writing: An Interdisciplinary
Journal, 7 139–153.
Durgunoglu, A. Y., Nagy, W. E., & Hancin-Bhatt, B. J. (1993). Cross-language transfer of
phonological awareness. Journal of Educational Psychology, 85, 453–465.
Ganschow, I., Sparks, R. L., Javorsky, J., Pohman, J., & Bishop-Marbury, A. (1991).
Identifying native language difficulties among foreign language learners in college: A
“foreign” language learning disability? Journal of Learning Disabilities, 24, 530–541.
Geva, E., & Siegel, L. S. (2000). Orthographic and cognitive factors in the concurrent devel-
opment of basic reading skill in two languages. Reading and Writing: An Interdisciplinary
Journal, 12, 1–30.
Jastak, S., & Wilkinson, G. S. (1984). The Wide Range Achievement Test-Revised Wilmington,
DE: Jastak Associates.
Lambert, W. E., & Tucker, G. R. (1972). Bilingual education of children. Rowley, MA:
Newbury House.
Lindgren, S. D., De Renzi, E., & Richman, L.C. (1985). Cross-national development dyslexia
in Italy and the United States. Child Development, 56, 1404–1417.
McDougall, A., & Bruck, M. (1976). English reading within the French immersion program:
A comparison of the effects of the introduction of English reading at different grade lev-
els. Language Learning, 26, 37–43.
Olson, R., Kliegel, R., Davidson, B. J., & Foltz, G. (1985). Individual and developmental dif-
ferences in reading disability. In: T. G. Waller (Ed.), Reading Research: Advances in the-
ory and practice (Vol. 4, 1–64). New York: Academic Press.
Ryan, A., & Meara, P. (1991). The case of the invisible vowels: Arabic speakers reading
English words. Reading in a Foreign Language, L 531–539.
Siegel, L. S. (1986). Phonological deficits in children with a reading disability. Canadian
Journal of Special Education, 1, 45–53.
Siegel, L. S. (1988). Evidence that IQ scores are irrelevant to the definition and analysis of
reading disability. Canadian Journal of Psychology, 42, 201–215.
Siegel, L. S. (1993). The development of reading. In: H. W. Reese (Ed.), Advances in child
development and behavior (Vol. 24, 63–97). San Diego: Academic Press.
Siegel, L. S., & Ryan, E. B. (1988). Development of grammatical sensitivity, phonological, and
short-term memory skills in normally achieving and learning disabled children.
Developmental Psychology, 24, 28–37.
Siegel, L. S., & Ryan, E. B. (1989a). The development of working memory in normally achiev-
ing and subtypes of disabled children. Child Development, 60, 973–980.
Siegel, L. S., & Ryan, E. B. (1989b). Subtypes of development dyslexia: The influence of
definitional variables. Reading and Writing: An Interdisciplinary Journal, 2, 257–287.
Siegel, L. S., Share, D., & Geva, E. (1995). Evidence for superior orthographic skills in dyslex-
ics. Psychological Science, 6, 250–254.
Stavonich, K. E. (1982). Individual differences in the cognitive processes of reading: 1. Word
decoding. Journal of Learning Disabilities, 15, 485–493.
Stavonich, K. E. (1988a). Explaining the differences between the dyslexic and garden variety
poor reader: The phonological-core variance-difference model. Journal of Learning
Disabilities, 21, 590–604, 612.
Stavonich, K. E. (1988b). The right and wrong places to look for the cognitive locus of read-
ing disability. Annals of Dyslexia, 38, 154–177.
Stavonich, K. E., & Siegel, L. S. (1994). The phenotypic performance profile of reading-dis-
abled children: A regression-based test of the phonological-core variable-difference
model. Journal of Educational Psychology, 86, 24–53.
Stevenson, H. W., Stigler, J. M., Lucker, G. W., & Lee, S. Y. (1982). Reading disabilities: The
case of Chinese, Japanese and English. Child Development, 53, 1164–1181.
Swain, M., Lapkin, S., Rowen, N., & Hart, O. (1990) The role of mother tongue literacy in
third language learning. Vox. 4, 111–121.
Tunmer, W. E., Herriman, M.L., & Nesdale, A. R. (1988). Metalinguistic abilities and begin-
ning reading. Reading Research Quarterly, 23, 134–158.
Venezky, R. I. (1970). The structure of English orthography. The Hague: Mouton.
Verhoeven, L. T. (1990). Acquisition of reading in a second language. Reading Research
Quarterly, 25, 90–114.
Wagner, D., Spratt, J. E., & Ezzaki, A. (1989). Does learning to read in a second language
always put the child at a disadvantage? Some counterevidence from Morocco. Applied
Psycholinguistics, 10, 31–48.
Wagner, R. K. (1988). Causal relations between the development of phonological processing
abilities and the acquisition of reading skills: A meta-analysis. Merrill-Palmer Quarterly,
34, 261–279.
Willig, A. C. (1985). A meta-analysis of selected studies in the effectiveness of bilingual edu-
cation. Review of Educational Research, 55, 269–317.
Woodcock, R. W. (1973). Woodcock Reading Mastery Tests. Circle Press, Minnesota:
American Guidance Society.
Yopp, H. K. (1988). The validity and reliability of phonemic awareness tests. Reading
Research Quarterly, 23, 159–177.
678 Abu-Rabia and Siegel