Open Journal of Modern Linguistics
2013. Vol.3, No.1, 40-46
Published Online March 2013 in SciRes (http://www.scirp.org/journal/ojml) http://dx.doi.org/10.4236/ojml.2013.31005
40
How Does Type of Orthography Affect Reading in Arabic and
Hebrew as First and Second Languages?
Raphiq Ibrahim1,2*, Asaid Khateb1,2, Haitham Taha1,2,3
1The Edmond J. Safra Brain Research Center for the Study of Learning Disabilities, Faculty of Education,
University of Haifa, Haifa, Israel
2Department of Learning Disabilities, University of Haifa, Haifa, Israel
3Sakhnin College for Teachers’ Education, Sakhnin, Israel
Received December 18th, 2012; revised January 2nd, 2013; accepted January 10th, 2013
This study aimed to examine the effects of visual characteristics of Arabic orthography on learning to
read compared to Hebrew among Arabic and Hebrew bilinguals in an elementary bilingual education
framework. Speed and accuracy measures were examined in reading words and non-words in Arabic and
Hebrew as follows: Arabic words and non-words composed of connected and similar letters, words and
non-words composed of connected and non-similar letters, and words and non-words composed of un-
connected letters. In Hebrew, words and non-words composed of similar letters and non-similar letters. It
was found that Arabic speakers showed an almost equal control in all reading tasks in both languages
whereas, Hebrew speakers showed better performance in their mother tongue in all reading tasks. In Ara-
bic, the best performance was in reading words and non-words that was unconnected. Based on these
findings, it was concluded that Hebrew speakers did not succeed in transferring their good ability in read-
ing their mother tongue to reading the second language, apparently due to the unique nature of the Arabic
orthography. Our findings with regard to the cross-linguistic research literature as well as the specific
features of Arabic language are discussed.
Keywords: Reading; Arabic; Hebrew; Orthography; Visual Complexity; Diglossia
Introduction
Research on bilingualism over the past three decades has fo-
cused on different issues including the effects of bilingualism
on cognitive and linguistic development. The consensus on this
issue today may be summarized thus: although bilinguals have
a more complex and possibly multi-structured mental lexicon
which is influenced by the idiosyncratic context in which the
languages have been learned and by the structural relations
between the languages (e.g., de Groot, 1992), this complexity
need not result in differential cortical organization of linguistic
abilities between bilinguals and monolinguals (Paradis, 2009).
Bialystok (2001) claimed that exposure to more than one lan-
guage at an early age results in heightened awareness of the
arbitrary and phonological aspects of language. In a previous
study, Arab children evince higher levels of phonological abili-
ties than monolingual Hebrew speakers (Eviatar & Ibrahim,
2000). Alphabetic orthography, like English and Hebrew, pho-
nological awareness is a very good predictor of success in read-
ing acquisition (Share, Jorm, Maclean, & Matthhews, 1984). In
turn to Arabic, the opposite finding has been reported in previ-
ous study that examined the relationship between phonological
abilities and various reading measures in first grade in Arabic
children learning to read Arabic and in Hebrew (monolinguals)
and Russian (bilinguals) children learning to read Hebrew
(Ibrahim, Eviatar, & Aharon Peretz, 2007). The authors sug-
gested that learning to read in Arabic is more challenging than
in Hebrew. One possible answer to this is that this is an
effect of the diglossia. In previous study it has been shown that
skilled adult readers of Arabic also read more slowly than
skilled adult readers of other languages (Azzam, 1993). There-
fore, diglossia cannot be the only reason for this pattern. In
addition, we wondered, what could be blocking the facilitative
effect of phonological awareness. In this study, this question
will be examined directly among children learning to read
Arabic and Hebrew and studying in similar conditions in a bi-
lingual education framework.
Arabic and Hebrew Orthographic Characteristics
Arabic is a typical case of diglossia. According to Saiegh-
haddad (2005), modern standard Arabic (MSA) is the language
used throughout the Arabic speaking world for writing and
some other formal functions, such as speeches and religious
sermons, while the spoken Arabic vernacular (SAV) is the lan-
guage used for everyday conversation. The classical literary
version is studied in school and is not acquired naturally with-
out formal learning. Ibrahim (2009) reported that learning LA
appears to be, in some respects, more like learning a second
language than like learning the formal register of one’s native
language. As opposed to the Arabic, diglossia does not exist in
Hebrew.
In addition to its diglossic nature, the orthography of Arabic
plays essential roles in assessing reading and examining the
predicative power of different processes.
The unique characteristics of Arabic (and to little extent in
Hebrew) orthography are many and complex, so only some,
those most pertinent to the present investigation will be
*Corresponding author.

R. IBRAHIM ET AL.
reviewed briefly here. These are the positional variants of let-
ters, the consonant diacritics, and letter ligature. Indeed, a num-
ber of letters (graphemes) share the same form (derived from
Nabatean which had fewer consonants) and are distinguished
only by the position and the number of consonant (dot) diacrit-
ics For example, the letters //, // and // represent the
consonants /t/, /b/ and /th/ respectively. Some adaptations of the
Arabic abjad (e.g., Sindhi in southern India), include up to 7 or
even 8 diacritical variants of the identical letter-form. An addi-
tional unique feature of Arabic orthography is that the majority
of letters vary in shape according to position in the word; word-
initial, medial or word-final position. It is worthy to emphasize
that letter position also imply a change on the letter variant
causing a little change ()//// or a large change (/)////
on the letter shape (for more examples see Appendix A). Six
letters, however, have only two variant shapes which depend
not only on the position in the word but also on the preceding
letter (//r/; //z/; //d/; //th/; //w/; and //a/). This subset of
letters may connect only from the right side (//Lawh/) but
not from the left /(/Walad/). This sub-group of letters, there-
fore, may appear to the reader as more “distinct”, because visu-
ally separated from adjacent letters. Unlike the Arabic orthog-
raphy, in Hebrew the majority of letters don't vary in shape
according to position in the word. Only five letters, n, , m, ,
p, , tz, , h, have two forms depending on whether they
appear in word-final or in other positions. Generally, the shal-
low version of the Hebrew orthography where every phoneme
is represented by either consonant or diacritical mark is ac-
quired relatively easily with typically developing first grade
children reaching decoding accuracy by the end of the first
grade (Shatil, Share, & Levin, 1999). This is in contrast to the
acquisition of basic reading skills in Arabic.
Besides, the written vowelization system is considered as one
of the unique features of the written Arabic. Written Arabic and
Hebrew words could be vowelized by diacritical marks added
above and below the letters within the word. In the case of
vowelized written words, the written patterns are considered as
shallow orthography, while in the case of non-vowelized writ-
ten words, the orthography is considered as a deep one. In this
case, the phonology is not reflected throughout the orthographic
pattern alone but other context cues are needed. Deep ortho-
graphic patterns usually appear in texts dedicated to adult read-
ers (Abu-Rabia, 2000).
Research on Orthography
Within the orthographic patterns of the written words, some
of the letters can be connected with former and subsequent
letter, while other letter can be connected only with the former
letters. As a result different types of written words can be pro-
duced: 1) fully connected; 2) partially connected; and 3) non-
connected words (i.e. words appearing with the basic forms of
the letters). Recent findings showed that these differences of the
internal connectivity of the written words in Arabic have an
impact on the time course of early brain electric responses dur-
ing the visual processing of letters (Ibrahim, Eviatar, & Aha-
ron-Peretz, 2002; Eviatar, Ibrahim, & Ganayim, 2004) and writ-
ten Arabic words (Taha, Ibrahim, & Khateb, 2012). These re-
sults complement those reported by Rao, Vaid, Srinivasan and
Chen (2011), where native Urdu readers read Urdu more slowly
than Hindi. These authors suggested that this is due to two fac-
tors: orthographic depth, where the relations between graph-
emes and phonemes are more regular in Hindi than in Urdu
(which has a consonantal script, like Arabic); and the greater
visual complexity of Urdu orthography than Hindi orthography.
A study by Taouk and Coltheart (2004), investigated learning
to read in Arabic between children and adults. In one experiment
they attempted to examine naming of real pronounceable “posi-
tion-illegal” words; which are words written with a wrong letter
variant according to its position. They found that children’s
word reading was significantly impaired when incorrect posi-
tional variants were substituted for the correct variants. This
finding provides evidence that positional variants of letters
affect word reading. A recent study by Abdelhadi, Ibrahim and
Eviatar (2011) examined how orthographic complexity can af-
fect vowel identification between 3rd and 6th grade. They used a
vowel identification task with stimuli in Arabic and Hebrew at
three levels of lexicality; real words, pseudo-words and non-
letters. Each level included three categories; separated letters,
ligatured or connected letters and connected letters with vowel
diacritics. The participants were required to identify a specific
vowel, fatHa or Patax. The highest performance was predicted
for separated letters, followed by connected letters, while the
poorest performance was expected on stimuli with connected
letters with vowel diacritics. The results were not consistent
with the hypothesis; unexpectedly, children from both grades
responded faster to letter strings with connected letters (both
words and pseudo-words). These results suggested that 3rd and
6th graders used a different perceptual strategy when the stimuli
were more word-like (e.g., comprised of connected letters) than
when they were less word-like (e.g., comprised of disconnected
letters).
The Linguistic Reality in Israel
The education system in Israel maintains the existing de-
tachment between Jews and Arabs: schools are nationally-
separated. Although Arabic is considered as a formal language
in the country, Jewish students are less exposed to Arabic lan-
guage in particular and to Arabic culture in general than Arab
students who study extensively Hebrew language, literature and
culture (Al-Hag, 2003; Amara & Mar I, 2002).
In Israel, bilingual schools emphasize the symmetry between
the two languages, Arabic and Hebrew, in every teaching meas-
ure. Therefore, reading constitutes a fundamental skill in the
acquisition of every language. Reading contains the ability to
connect reading symbols to meaningful word for the sake of
text decoding. From theoretical view, accepted theory concern-
ing the field of learning of second language and bilingualism
ascribes academic success and the acquisition of second lan-
guage in the level of the first language (Cummins, 1991, 2000).
According to Cummins (2000), as much as the level of linguis-
tic expertise in the first language increases (passes certain
threshold) over certain level of skillfulness higher than in the
second language (another higher threshold), appear the cogni-
tive growth.
The Current Study
In this study we postulate that the level of performance in
reading tasks (the speed and accuracy of reading) in Arabic and
Hebrew language (correspondingly) may reflect the different
nature of the transfer to second language. Native Hebrew speak-
ers may find more difficulty in learning Arabic as a second

R. IBRAHIM ET AL.
42
language than do native Arabic speakers when learning Hebrew,
this difference results from the complexity of the Arabic lan-
guage. According to the above mentioned postulation, an aca-
demic success in learning second language is only dependent
on skills acquired in first language, and therefore according to
Cummins’s interdependence hypothesis, we may expect that
there are no differences between the two bilingual groups in all
the other influencing conditions, i.e. age, the period of exposure
to second language and learning environment. In addition, this
study examines the unique characteristics of Arabic orthogra-
phy on the speed and accuracy measures when reading words
and non-words in Arabic and Hebrew as first and second lan-
guage.
The hypotheses of the study:
1) As far as accuracy and speed are concerned, we predict the
level of performance will be better in tests examining reading in
first language than reading in the second language. According
to Cummins interdependence hypothesis, academic success in
learning second language is only dependent on skills acquired
in the first language.
2) We predict the performance in reading tests in Hebrew
will be at a similar level or even better than reading tests in
Arabic beyond groups. In their study, (Ibrahim, Eviatar, &
Aharon-Peretz, 2002) explained that this mostly results from
two reasons: first, the orthographical complexity of the Arabic
language and second, the status of written Arabic (Literary)
which is considered by some authors as a second language
(Ibrahim & Aharon-Peretz, 2005; Eviatar & Ibrahim, 2001;
AbuRabia, 2000; Saiegh-Haddad, 2003, 2004).
3) Among native Arabic speakers, reading words with dis-
connected letters in Arabic is better (in terms of accuracy and
speed) than reading words with connected letters. According to
Azzam (1993), the major problem is that children have to re-
member tree or four shapes for each letter according to the
position in the word and the child’s ability to distinguish be-
tween letters based on tiny differences of a single characteristic
or the position of dots in letters.
Method
Participants: 49 participants (32 from 3rd grade and 17 from
4th grade) with an average age of 9.3 years participated in this
study. They were recruited from two bilingual schools in the
north and center of the country where the study was conducted.
Participants from the two groups were matched with regard to
age, general ability, and the level of reading (educators chose
children with commanding performance in reading in the two
languages Arabic and Hebrew and none of them was diagnosed
with learning disabilities). In the bilingual educational schools
participated, literacy in Hebrew and literacy in Arabic is ac-
quired in parallel from first grade. Children receive equal
amounts of instruction in each language daily by both Hebrew
and Arabic L1 teachers. Therefore the teaching method planned
to be similar in two languages.
Stimuli: Three types of lists of words in Arabic were con-
structed: words with connected and similar letters, words with
connected and non-similar letters and words non-connected
letters (see Table 1 for example). In Hebrew, two lists of words
were constructed: words with similar letters and words with
non-similar letters (see Table 2 for example). This difference is
due to the fact that Hebrew letters do not connect with each
other. For each word sub-list, a list of non-words was estab-
lished. This allowed examining the process of pure encoding of
reading in comparison with the process of structural reading of
familiar words.
In all list, non similar (non-cognate) words from literary and
non-literary Arabic (spoken) were chosen. This step was taken
to avoid the possibility that similar shapes cause to bias scores.
Before, conducting the experiment, the frequency of words in
stimulus list was assessed. For this purpose, the initial word list
included 240 words in Arabic (divided into three sub-lists de-
fined on the basis of their orthographic characteristics) and 160
words in Hebrew (divided into two groups based on their
orthographical characteristics). The initial Arabic and Hebrew
word lists established with the of language teachers from other
schools in the neighboring area. The words of all lists were then
introduced in a questionnaire of frequency which was filled by
children. Here, they had decide how frequent was each word
using a scale ranging between 1 (rare) and 5 (very frequent).
Based on the results of this assessment, 24 words whose mean
frequency ranged between 1.5 - 3.5 points were finally retained
for each sub-list. The rationale for using this middle frequency
range was to neutralize as much as possible the factor of word
frequency.
Table 1.
Experimental conditions in Arabic.
Words which consist of similar letters and
connected
Words which consist of dissimilar letters and
connected
Words which consist of letters that is
disconnected
word non-word word non-word word non-word

 
  
 
 
 
 

tantab Batnat methyaa’ thaya’am awzan azwan
Table 2.
Experimental conditions in Hebrew.
Words which consist of similar letters Words which consist of dissimilar letters
A word A non-word A word A non-word
    
harad dahor zahov zahol

R. IBRAHIM ET AL.
It should be noted that the number of syllables and length of
the words were controlled between Arabic and Hebrew. As for
the non-words (which also were comparable to words in length
and number of syllables), they were formed by changing one or
two letters in the word or by substituting the position of some
letters within each real word. All together, the study included
six sub-lists in Arabic (of 24 items each) and four sub-lists in
Hebrew (also of items).
Concerning the matter of examining the reliability of tasks,
an examination of the measure reliability of different tasks was
conducted based on Alfa Cronboch. (See Table 3).
From Table 3, it can be inferred that beyond the language
groups, grades’ reliability and speed of reading in different
tasks ranged between .76 and .97.
Procedure: The study took place inside the school in quiet
room, individually for each participant. The meeting with each
participant started with a short acquaintance aimed at creating a
comfortable atmosphere while providing a brief explanation for
the essence of the meeting in order to remove any concern, dis-
comfort or hesitation. When the participant confirmed his readi-
ness to start the tasks he/she was instructed to read ten sub-lists
(hereafter tasks) in Arabic and Hebrew as fast and accurately as
possible. The order of the presentation of the tasks was bal-
anced across participants. Four short training lists were pre-
pared so as familiarize the participants with the different tasks.
Before starting each task, they were informed that their reading
time of each tasks will be monitored using a stopwatch. In ad-
dition, a digital recorder registered their reading during the
successive tasks for assessing off-line their reading accuracy.
Results
The analysis of the reading accuracy and speed in the native
language showed that both reveal a good reading performance
among the two groups of participants (native Arabic speakers
and native Hebrew speakers) concerning speed and accuracy in
the different task in their native language. Furthermore, very
important results were found in relation to the difference status
of Arabic language for native Arabic speakers and as a second
language for native Hebrew speakers. This is in comparison to
the status of Hebrew language for the two groups of speakers.
Decisive effect of Arabic writing characteristics on the acquisi-
tion of this language was demonstrated by tyro readers. The
average time of reading and percentages of accuracy in all read-
ing tasks (reading separated words and reading text) in the two
languages were the dependent variables.
The results of each hypothesis mentioned at the introduction
are separately presented as the basic question which stands at
the core examination is: is learning to read in Arabic language
more difficult than Hebrew?
The average-time of reading and percentages of accuracy
were analyzed as the native language (Arabic speakers and
Hebrew speakers), the language of the test (Arabic and He-
brew) and the type of the test (reading words and reading text)
were served as independent variables. Regarding the issue of
the two types of tasks (reading words and reading text) two
analyses were conducted:
In the first analysis, the variable of the type of tests included
three levels of different types of words (connected and similar,
connected and dissimilar, and disconnected).
Examining differences in reading accuracy of native lan-
guage on the basis of native language:
One-way analysis of variance of the grade in reading words
in native language on the basis of native language was carried
out and significant results were found F(1,47) = 15.47, p < .001,
2
p

= .25. Native Arabic speakers are more successful (M =
97.62, SD = 1.78), than their counterparts (native Hebrew
speakers) (M = 93.73, SD = 4.26).
Examining difference in the speed of reading words in native
language on the basis of native language:
Here a one-way analysis of variance in the speed of reading
words in mother language on the basis of native language was
carried out and significant results were found F(1,47) = 19.80, p
< .001, 2
p

= .30. Hebrew speakers (M = 34.40, SD = 10.68),
are faster than their counterparts (Arabic speakers) (M = 50.15,
SD = 13.32). It should be pointed out that compatibility in the
number of syllables and the length of words in each task in the
two languages (Arabic and Hebrew) was conducted.
Examining differences in the speed of reading Arabic words
on the basis of native language and type of words:
Here are two-way analysis of variance in the speed of read-
ing words in Arabic on the basis of native language and type of
words was conducted with repeated measurements for type of
words and significant differences were found (F(1,47) = 55.35,
p < .001, 2
p

= .54). Arabic speakers (M = 50.15, SD = 13.32)
were faster than their counterparts (Hebrew speaker) (M =
128.57, SD = 53.80). In addition differences were also found on
the basis of type of words F(5,235) = 48.11, p < .001, 2
p

= .51. An interaction based on native language and type of
words was found F(5,235) = 20.26, p < .001, 2
p

= .30. To
examine the source of this interaction, researchers carried out a
Post-Hoc test. It was found that regarding to test 5 (words with
disconnected letters), Arabic speakers were faster than Hebrew
speakers in comparison to other types of words. In addition,
regarding test 6 (non-words with disconnected letters), Arabic
speakers were faster in comparison to tests 2, 3 and 4. Further-
more, it was also found that Hebrew speakers were faster in
reading words 4, 5 and 6, than words 1, 2 and 3, and they were
also slower in reading words from type 4 than words from types
5 and 6.
Examining differences of accuracy in reading words in Ara-
bic language on the basis of native language and the type of
word:
Table 3.
Accuracy and rates in reading Arabic and Hebrew words.
Total items SD Mean Task  (Cronbach)
6 .97 Accuracy in reading Arabic words 81.22 22.57
6 .96 Reading rate of Arabic words 83.76 53.32
4 .76 Accuracy in reading Hebrew words 95.73 4.29
46.62 20.23 .95 4 Reading rate of Hebrew words

R. IBRAHIM ET AL.
Two-way analysis of variance of grade in reading words in
Arabic on the basis of native language and the type of word
with repeated measurements for the type of word was con-
ducted. Significant differences was found accordingly F(1,47) =
34.32, p < .001, 2
p

= .42. Arabic speakers (M = 94.32, SD =
3.90), were more successful than Hebrew speakers (M = 67.53,
SD = 23.86). In addition differences in the type of words were
found F(5,235) = 30.54, p < .001, 2
p

= .39. Here there is an
interaction based on native language and the type of word
F(5,235) = 17.71, p < .001, 2
p

= .27.
To examine the source of this interaction, a post-hoc test was
carried out. It was found that Arabic speakers were more suc-
cessful in words of the type 5 (words with disconnected letters)
and they were also more successful in words of the type 6
(non-words with disconnected letters) in comparison to words
of the type 2 (non-words with connected and similar letters.
Examining difference in the speed of reading words in He-
brew language based on native language and type of word:
Researchers conducted a two-way analysis of variance in the
speed of reading words in Hebrew based on native language
and type of words with repeated measurement for the type of
word. Significant differences based on native language were
found F(1,47) = 18.24, p < .001, 2
p

= .28. Hebrew speakers
(M = 34.93, SD = 10.68), were faster than Arabic speakers (M
= 55.79, SD = 20.97). In addition, differences were found con-
cerning the type of words F(3,141) = 18.63, p < .001, 2
p

= .28. Therefore, there is an interaction on the basis of native
language and the type of word F(3,141) = 13.56, p < .001, 2
p

= .22. To examine the source of this interaction, a Post-Hoc test
was carried out.
It was found that among Arabic speakers, no differences
were found in the speed of reading based on type of word. It
was also found that Hebrew speakers were faster in reading
words with dissimilar letters than reading words with similar
letters or non words. They were also faster in reading words
with similar letters than reading non-words with similar and
dissimilar letters.
Examining differences in accuracy in reading words in He-
brew based on native language and type of word:
Tow-way analysis of variance in the grade of reading words
in Arabic based on native language and the type of word with
repeated measurements for the type of word was conducted.
Significant differences were found regarding native language
F(1,47) = 8.22, p < .01, 2
p

= .15. Hebrew speakers, (M =
97.62, SD = 1.78), were more successful than Arabic speakers
(M = 94.31, SD = 5.05), though no differences were found
concerning the type of word F(3,141) = .74, p > .05, 2
p

= .02.
Thus, there was no interaction on the basis of native language
and type of word F(3,141) = .43, p > .05, 2
p

= .02.
Examining differences in the speed of reading words based
on native language and the language of test:
Tow-way analysis of variance of the speed of reading words
based on native language and the language of test with repeated
measurement for the language of test was conducted. Signifi-
cant differences were found regarding native language F(1,47)
= 18.59, p < .001, 2
p

= .28. Arabic speakers (M = 52.97,
SD=16.57), were faster than Hebrew speakers (M = 81.48, SD
= 29.37), and here significant differences were found regarding
the language of test F(1,47) = 80.43, p < .001, 2
p

= .63.
Among all participants, it was found that they were faster in
reading in Hebrew M = 45.09, SD = 20.23 than Arabic (M =
89.36, SD = 53.32), and this is beyond the native language
F(1,47) = 102.22, p < .001, 2
p

= .69. Finally, an interaction
was found based on native language and the language of test
F(1,47) = 102.22, p < .001, 2
p

= .69. A post-hoc examination
suggests that for native Arabic participants, no differences were
found between reading words in Arabic (M = 50.15, SD =
13.32), and reading words in Hebrew (M = 55.78, SD = 20.97).
On the other hand, for native Hebrew participants, the speed of
reading words in Arabic (M = 128.57, SD = 53.80), is longer
than the speed of reading words in Hebrew (M = 34.39, SD =
10.68). It was also found that native Arabic readers’ speed of
reading Arabic is faster than native Hebrew readers. Conversely,
native Hebrew readers’ speed of reading Hebrew is faster than
the speed of native Arabic.
Examining differences in accuracy of reading words based
on native language and the language of test:
This research reveals another important finding that is con-
sistent with a line of earlier studies which suggested that the
status of literary Arabic is equal to the second language for
native Arabic readers. According to this finding, native Arabic
rehears’ performances in Arabic reading tasks were not differ-
ent from their performances in Hebrew reading tasks (Actual
second language). On contrary, native Hebrew readers’ per-
formance in reading Hebrew tasks was significantly better.
Here, researchers conducted a tow-way analysis of variance
for grade in reading words based on native language and the
language of test with repeated measurement, and significant
differences related to native language were found F(1,47) =
23.14, p < .001, 2
p

= .33. It was also found that native Arabic
speakers were more successful (M = 94.31, SD = 3.94), than
their Hebrew counterparts (M = 82.57, SD = 12.13). Therefore,
differences related to the language of test were also found
F(1,47) = 43.77, p < .001, 2
p

= .48 accordingly there is an
interaction based on native language and the language of test
F(1,47) = 43.84, p < .001, 2
p

= .48. Graph 5 shows the re-
ceived interaction (See Figure 1).
According to a Post-Hoc examination, for native Arabic par-
ticipants, no differences were found in accuracy of reading
words in Arabic (M = 94.32, SD = 3.90) and reading words in
Hebrew (M = 94.31, SD = 5.05). On the contrary, for native
Hebrew participants, the level of accuracy in reading words in
Hebrew (M = 97.62, SD = 1.78), was higher than reading words
in Arabic (M = 67.53, SD = 23.86). In addition, concerning
reading Arabic words, native Arabic readers were better than
Figure 1.
Interaction of reading words accuracy by the subject native language.
44

R. IBRAHIM ET AL.
native Hebrew readers. Yet, concerning reading words in He-
brew, no differences were found between the two groups.
Discussion
The analysis of the study’s results is conducted on the basis
of Cummins’s (1979-1981) interdependence hypothesis. Ac-
cording to this hypothesis a good linguistic ability in the first
language predicts similar ability in the second language. Con-
sequently, native Hebrew and Arabic speakers would have been
expected to read in second language at a similar level (of speed
and accuracy) to their native languages. Yet it was found that
native Arabic speakers read significantly better in Hebrew than
native Hebrew speakers read in Hebrew. These findings em-
phasize that learning to read in Arabic language more chal-
lenging and than reading in Hebrew even though Arabic and
Hebrew have a similar nominal basis.
In the course of the current study, tasks the level of perform-
ance was compared with reading separated words and non-
words in Arabic and Hebrew, when the unique characteristics
of Arabic orthography were greatly considered from a forma-
tive point of view. In this regard, the purpose was to clarify
emperically whether Arabic language is more difficult in proc-
essing than Hebrew. As it has been planned, the two popula-
tions whose mother languages are different are simultaneously
and at the same extent exposed to reading and writing in the
second language in the same educational frame; two bilingual
schools in Israel.
The first prominent finding in this study was that reading
Hebrew is faster and more accurate than reading Arabic beyond
the native language (Arabs as Jews) and beyond the type of test
(reading words and non-words). In this context, it is important
to note, that even though acquisition of the Hebrew orthography
is characterized by some level of complexity due to for example
the similarity in the shape of the letters, the acquisition of the
Arabic orthography is very much more complicated. Besides
for the diglossia of the Arabic language, there is an additional
complexity stemming from the similarity in the shape of letters,
number of dots above or below, the connection between the
letters, together with the changing shape of letters depending on
their location within the word. Thus, our data bring additional
evidence that this complexity influences both reading accuracy
and reading rate among the two groups (Arabic and Hebrew
speakers). This finding is important and is joined with earlier
findings that dealt with the status of literary Arabic language
for Arabic readers as well as the formative difficulty of Arabic
orthography (Eviatar, Ibrahim, & Ganayim, 2004; Ibrahim Eviatar
& Aharon Peretz, 2002; Maamouri, 1998; Saiegh-Haddad, 2003).
This central finding relates to fundamental differences be-
tween the two types of orthographies in Arabic (connected
words versus nonconnected words) and Hebrew (words with
similar versus words with dissimilar letters). In this compari-
son, it was found that among Arabic speakers, reading words
and non words in Arabic with disconnected letters is more ac-
curate and faster than reading words and non-words with con-
nected letters among either group. This finding emphasizes the
difficulty in reading words and non-words composed of con-
nected letters and basically of similar and connected letters—a
specific quality of Arabic orthography. This finding is in line
with the findings of Ibrahim et al. (2002) who compared the
spatial identification of Arabic letters with Hebrew letters. Even
so, in the same area (but inside participants), differences were
found among native Hebrew speakers; only in the speed based
on the type of word (groups of words and non-words separate
by the type of the orthographical information). However, there
were no differences in the accuracy of reading words and non
words (with similar and dissimilar letters). Among Hebrew
speakers, it was found that the speed of reading words with
similar letters is slower than reading words and non words with
dissimilar letters. The matter of the similarity of the shape and
feature of letters was also examined in Arabic language. Shim-
ron & Navon (1982), argued that in comparison to other lan-
guages, all letters in Hebrew are quadrangular and similar, such
as English. According to these researchers, letters in Hebrew
are mostly at one line this affects the ability to identify word.
When letters are more similar, it is more difficult to identify
them compared to dissimilar letters (Shimron & Navon, 1982).
Therefore in comparing Arabic and Hebrew, the speed of He-
brew speakers in reading words with similar letters in Hebrew
is faster than the speed of Arabic speakers and native Hebrew
speakers in reading words with similar and connected letters. It
should be noted that the similarity in the shape of letters in
Hebrew leads to a relative delay in indentifying words but not
making error. Although the similarity in Arabic orthography
(the shape of letters, the number of dots above and under, the
connection between them, together with the change of their
shapes based on their positions in the word) affects the accu-
racy and speed of reading among the two groups. This result is
in consistency with earlier studies conducted by Eviatar and her
colleagues (2004), who suggested that there are other factors
which may add special difficulty when reading in Arabic leads
to slower reading and making errors (Eviatar, Ibrahim, & Ga-
nayim, 2004). Beyond that, at initial presentation, the connected
items (which were matched for diacritical complexity) were
actually read more quickly than the non-connected items. This
finding is consistent with Abdelhadi et al. (2011) who also
found a speed advantage for words with connecting letters
among skilled adult readers of Arabic performing a visual
search task. The authors attributed this advantage to the fact
that most printed words in Arabic consist of ligatured letters
rather than non- ligatured letters. The present data, therefore,
replicates this finding and extends it to young readers perform-
ing a standard word pronunciation task. In a related issue, Ibra-
him and Eviatar (2012) tried to determine if the processing of
Arabic orthography seems to make different demands on the
cognitive system both in beginning and in skilled readers while
recognizing Arabic letters compared to Hebrew letters. The re-
searchers used behavioral measures of performance asymme-
tries in a divided visual field paradigm. There results show that
Arabic orthography specifically disallows the involvement of
the RH in letter identification, even while the RH of the same
participants does contribute to this process in English and in
Hebrew. The results were attributed to the additional visual
complexity that characterizes Arabic orthography.
Conclusion and Future Research
The present study investigated 3rd and 4th grade bilingual
Arabic/Hebrew-speaking learners in relation to visual factors
which are central to issues of second language learning. While
we did not aim to propose an explanatory model of learning to
read in second language, the findings of the present study sup-
port previous findings suggested that measures of speed and
accuracy of reading in Hebrew among Arabic speakers were

R. IBRAHIM ET AL.
46
significantly higher than measures of reading in Arabic among
Hebrew speakers. Our conclusion was that Hebrew speakers
did not succeed in transferring their proficiency and success in
reading in their mother tongue to success in reading the second
language. In view of the fact that bilingual schools use similar
method in teaching the two languages, this allow us to conclude
that Hebrew speakers did not succeed in transferring their good
ability in reading their mother tongue to reading the second
language, due to the unique nature of the Arabic orthography.
Turning to reading disabled (or dyslexic) children, Breznitz
(2003) argued that the inefficiency of reading process in this
population derives not only from problems of accuracy and
timing inside a specific processing system but also from inter-
action of information between systemic processes (i.e., the vi-
sual-orthographical system). The findings regarding learners’
proficiencies in second language (Arabic or Hebrew) contribute
to debates regarding the best methods or the strategies that
might be chosen to teach learning to read second language. The
findings of the present study rather suggest that, it is useful to
continue to explore the cognitive and neurocognitive basis in
decoding writing systems and how type of orthography might
affects some hitherto neglected aspects of second language
learning and L2 proficiency.
REFERENCES
Abdelhadi, S., Ibrahim, R., & Eviatar. Z. (2011). Perceptual load in the
reading of Arabic: Effects of orthographic visual complexity on de-
tection. Writing Systems Resear ch, 3, 117-127.
doi:10.1093/wsr/wsr014
Abu-Rabia, S. (2000). Effects of exposure to literary Arabic on reading
comprehension in a diglossic situation. Reading and Writing: An In-
terdisciplinary Journal, 13, 147-157. doi:10.1023/A:1008133701024
Abu-Rabia, S., & Siegel, L. S. (1995). Different orthographies, differ-
ent context effects: The effects of Arabic sentence context inskilled
and poor readers. Reading Psychology: An International Quarterly,
16, 1-19. doi:10.1080/0270271950160101
Amara, M., & Mar’i, A. (2002). Language education policy: The Arab
minority in Israel. Dordrecht/Boston/London: Kluwer Academic
Publishers.
Bialystok, E. (2001). Bilingualism in development: Language, literacy,
and cognition. New York: Cambridge University Press.
doi:10.1017/CBO9780511605963
Azzam, R. (1993). The nature of Arabic reading and spelling errors of
young children. Reading a nd Writing, 5, 355-385.
doi:10.1007/BF01043112
Breznitz, Z. (2003). Speed of phonological and orthographic processing
as factors in dyslexia: Electrophysiological evidence. Genetic, Social
and General Psychology Monographs, 129, 183-206.
Cummins, J. (1979). Linguistic Interdependence and the Educational
Development of Bilingual Children. Review of Educational Research ,
49, 222-251.
Cummins, J. (1981). The role of primary language development in pro-
moting educational success for language minority students. In Cali-
fornia State Department of Education (Ed.), Schooling and language
minority students: A theoretical framework (pp. 3-49). Los Angeles:
Evaluation, Dissemination and Assessment Center, California State
University.
Cummins, J. (2000). Language, power and pedagogy: Bilingual chil-
dren in the crossfire. Clevedon: Multilingual Matters.
Eviatar, Z., & Ibrahim, R. (2001). Bilingual is as bilingual does: Me-
talinguistic abilities of Arabic speaking children. Applied Psycholin-
guistics, 21, 451-471. doi:10.1017/S0142716400004021
De Groot, A. M. B. (1992). Determinants of word translation. Journal
of Experimental Psychology: Learning, Memory, and Cognition, 18,
1001-1018. doi:10.1037/0278-7393.18.5.1001
Ibrahim, R., & Aharon-Peretz, J. (2005). Is literary Arabic a second
language for native Arab speakers: Evidence from a semantic prim-
ing study. The Journal of Psycholinguistic Re s ear ch , 34, 51-70.
doi:10.1007/s10936-005-3631-8
Ibrahim, R., & Eviatar, Z. (2012). Multilingualism among Israeli Arabs,
and the neuropsychology of reading in different languages. Literacy
Studies, 5, 57-74.
Ibrahim, R., Eviatar, Z., & Aharon Peretz, J. (2002). The characteristics
of the Arabic orthography slow it’s cognitive processing. Neuro-
psycholgy, 16, 322-326. doi:10.1037/0894-4105.16.3.322
Ibrahim, R., Eviatar. Z., & Aharon Peretz, J. (2007). Metalinguistic
awareness and reading performance: A cross language comparison.
The Journal of Psycholinguistic Research, 36, 297-317.
doi:10.1007/s10936-006-9046-3
Ibrahim, R. (2009). The cognitive basis of diglossia in Arabic: Evi-
dence from a repetition priming study within and between languages.
Psychology Research and Behavior Manag em en t, 12, 95-105.
Maamouri, M. (1998). Language education and human development:
Arabic diglossia and its impact on the quality of education in the
Arab region. The Mediterranean Development Forum. Washington,
DC: The World Bank.
Paradis, M. (2009). Declarative and procedural determinants of second
languages (studies in bilingualism 40). Amsterdam: John Benjamins.
Rao, C., Vaid, J., Srinivasan, N., & Chen, H. C. (2011). Orthographic
characteristics speed Hindi word naming but slow Urdu naming:
Evidence from Hindi/Urdu biliterates. Reading and Writing, 24, 679-
695. doi:10.1007/s11145-010-9256-9
Roman, G., & Pavard, B. (1987). A comparative study: How we read
Arabic and French. In J. K. O’Regan, & A. Levy-Schoen (Eds.), Eye
movements from physiology to cognition (pp. 431-440). Amsterdam:
North Holland Elsevier.
Saiegh-Haddad, E. (2003). Linguistic distance and initial reading ac-
quisition: The case of Arabic diglossia. Applied Psycholinguistics, 24,
115-135. doi:10.1017/S0142716403000225
Saiegh-Haddad, E. (2004). The impact of phonemic and lexical dis-
tance on the phonological analysis of words and pseudowords in a
diglossic context. Applied Psycholinguistics, 25, 495-512.
doi:10.1017/S0142716404001249
Saiegh-Haddad, E. (2005). Correlates of reading fluency in Arabic: Di-
glossic and orthographic factors. Reading and Writing: An Interdis-
ciplinary Journal: An Interdisciplinary Journal, 18, 559-582.
doi:10.1007/s11145-005-3180-4
Share, D. L., Jorm, A. F., Maclean, F., & Matthews, R. (1984). Sources
of individual differences in reading acquisition. Journal of Educa-
tional Psychology, 7 6, 1309-1324.
doi:10.1037/0022-0663.76.6.1309
Shatil, E., Share, D. L., & Levin, I. (2000). On the contribution of kin-
dergarten writing to Grade 1 literacy: A longitudinal study in He-
brew. Applied Psycholinguistics, 28, 1-25.
Shimron, J., & Navon, D. (1982). The dependence on graphemes and
on their translation to phonemes in reading: A developmental per-
spective. Reading Research Quarterly, 17, 210-228.
doi:10.2307/747484
Taha, H., Ibrahim, R., & Khateb, A. (2012). How does Arabic ortho-
graphic connectivity modulate brain activity during visual word rec-
ognition: An ERP study. Brain Topography, 26, 292-302.
doi:10.1007/s10548-012-0241-2
Taouk, M., & Coltheart, M. (2004). Learning to read in Arabic. Read-
ing and Writing, 17, 27-57.
doi:10.1023/B:READ.0000013831.91795.ec