kahn-horwitz, schwartz & share (2011)
DESCRIPTION
OrtografiaTRANSCRIPT
Acquiring the complex Englishorthography: a triliteracy advantage?Janina Kahn-Horwitz
Oranim Academic College of Education; Gordon College of Education, Israel
Mila Schwartz
Oranim Academic College of Education; Edmond J. Safra Brain Research Centerfor the Study of Learning Disabilities, Department of Learning Disabilities,University of Haifa, Israel
David Share
Edmond J. Safra Brain Research Center for the Study of Learning Disabilities,Department of Learning Disabilities, University of Haifa, Israel
The script-dependence hypothesis was tested through the examination of the impact ofRussian and Hebrew literacy on English orthographic knowledge needed for spellingand decoding among fifth graders. We compared the performance of three groups:Russian–Hebrew-speaking emerging triliterates, Russian–Hebrew-speaking emergingbiliterates who were not literate in Russian (but only in Hebrew) and Hebrew-speakingemerging biliterates. Based on similarities between Russian and Englishorthographies, we hypothesised that Russian–Hebrew-speaking emerging triliterateswould outperform both other groups on spelling and decoding of short vowels andconsonant clusters. Further, we hypothesised that all groups would face similardifficulties with novel orthographic conventions. Russian–Hebrew-speaking emergingtriliterates demonstrated advantages for spelling and decoding of short vowels and fordecoding of consonant clusters. All three groups experienced difficulty with spellingand decoding the digraph th as well as the split digraph (silent e).
Going beyond second language (L2) acquisition to the acquisition of additional languages
is very common for many people all over the world. The development of third language
(L3) knowledge refers to the acquisition of a non-native language by learners who have
previously acquired or are acquiring two other languages. In this paper, we focus on
children for whom Russian is their heritage or first language (L1), and Hebrew is the
majority language and for them their L2. English for these children is an L3 that is
acquired at a relatively early age in third grade. All three languages are acquired
Journal of Research in Reading, ISSN 0141-0423 DOI: 10.1111/j.1467-9817.2010.01485.xVolume 34, Issue 1, 2011, pp 136–156
Copyright r 2011 UKLA. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ,UK and 350 Main Street, Malden, MA 02148, USA
Journal ofResearch in Reading
Journal ofResearch in Reading
sequentially. The current objective was to examine the contribution of the Russian
language and script to the acquisition of specific orthographic conventions in English as
an L3 versus challenges posed by the English orthography to Hebrew- and Russian-
speaking children.
English literacy is taught in Israeli elementary schools by starting with explicit
instruction of letter–sound correspondences (State of Israel Ministry of Education English
Inspectorate, 2009). The difference between long and short vowel sounds is not uniformly
emphasised (J. Kahn-Horwitz, unpublished data). English as an additional language
(EL2) textbooks that instruct early literacy acquisition include common digraphs (e.g. sh,
ch, th, ee, oo) and there is an instruction of these common digraphs. However, many of
these textbooks do not extensively cover the English orthographic conventions and it
would be inaccurate to say that the curriculum is phonics based or that instruction of
orthographic conventions is uniform (J. Kahn-Horwitz, unpublished data). This lack of
uniformity may also reflect a lack of orthographic knowledge on the part of EL2 teachers
(N. Roffman, unpublished data).
It has been hypothesised that access to two linguistic systems when acquiring an L3
represents a unique and complex phenomenon, which differs qualitatively from access to
only one linguistic system as in the case of L2 acquisition (Cenoz & Genesee, 1998;
Cenoz & Hoffmann, 2003). Some research supports this assumption by showing that
trilinguals differ from bilinguals in an increasing positive effect such as ‘an emergent
ability to focus attention’ (Aronin & OLaoire, 2003), cognitive effects such as
metalinguistic awareness (Jessner, 1999, 2006), general language proficiency and literacy
skills (language use, pragmatic skills, text reading accuracy and written expression) in
additional languages (Bild & Swain, 1989; Errasti, 2003; Valencia & Cenoz, 1992).
However, it could be that this evidence reflects proximity between the languages being
studied, which explains the positive effect of trilingualism and triliteracy (Cenoz &
Genesee, 1998; Cenoz & Hoffmann, 2003; Jessner, 2006; Leung, 2005). In this case, the
more typologically similar language and orthography (L1 or L2) might facilitate L3
literacy acquisition and vice versa. Accordingly, this facilitating effect may not be a
result of a general L1 and L2 and literacy advantage per se. In the following section, we
will address this issue of typological distance and closeness with regard to basic literacy
skill acquisition in EL2, specifically focusing on orthographic categories that are similar
and different to the Russian (L1) and Hebrew (L1, L2) orthographies.
Research focusing on the contribution of L1 and L2 phonemic awareness, decoding
and encoding to the development of L3 literacy skills among young bilinguals is still in
its infancy. The aim of this research was to investigate difficulties with novel digraphs in
English among Russian–Hebrew-speaking bilinguals who were literate in Russian and
Hebrew and were emerging triliterates, Russian–Hebrew-speaking bilinguals who were
literate in Hebrew only and were emerging biliterates and Hebrew-speaking monolingual
monoliterates who were emerging biliterates. These three groups were compared in their
L2 or L3 English spelling and word decoding regarding specific orthographic
conventions. An earlier study found Russian–Hebrew-speaking biliterate bilinguals in
fifth grade to be significantly superior to their Hebrew-speaking monolingual peers on
phonological processing skills (phoneme analysis, nonword decoding and spelling in
English L3; Schwartz, Geva, Leikin & Share, 2007). These findings provided support for
a positive impact of Russian (L1) basic literacy skills on English (L3) literacy acquisition.
In order to more finely examine the findings of this earlier study, the present study went
one step further by investigating specific English orthographic challenges emerging from
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 137
Copyright r 2011 UKLA
different literacy backgrounds. This study examined success in reading and spelling
specific English orthographic characteristics that exist in Russian but not in Hebrew as
opposed to English orthographic characteristics that do not exist in either Russian or
Hebrew for emerging biliterates as opposed to emerging triliterates.
What do we know about basic literacy skills acquisition from bilingual research?
Up to now, much of our knowledge concerning connections between written languages
with regard to spelling and decoding acquisition came from bilingualism research.
Relationships between basic literacy acquisition in L1 and L2 have been attributed to
underlying cognitive abilities (Geva & Ryan, 1993). Such abilities as, for example,
phonological awareness, and working memory are ‘thought to be part of one’s general
cognitive endowment and to be largely independent of specific language experiences’
(Genesee, Geva, Dressler & Kamil, 2006, p. 159). To complete the picture, phonological
segmentation ability was found to be a significant predictor of reading abilities among
children with monolingual, bilingual and trilingual backgrounds (Muter & Diethelm,
2001). This approach is rooted in Cummins’ interdependence hypothesis (Cummins, 1978),
and compatible with the linguistic coding differences hypothesis (Sparks & Ganschow,
1993), which attests to the transfer of literacy-related skills (e.g. phonological awareness
and word-recognition ability) between languages belonging to different linguistic families
with distinct orthographies (e.g. Hebrew–English, Geva & Wade-Woolley, 1998; Kahn-
Horwitz, Shimron & Sparks, 2005; Chinese–English, Gottardo, Yan, Siegel & Wade-
Woolley, 2001; Russian–Hebrew, Schwartz, Leikin & Share, 2005). Cross-linguistic
transfer has evolved conceptually and the current study uses August and Shanahan’s (2006)
broader conception of cross-language transfer whereby a dynamic interaction exists
between L1 and L2 component language and literacy skills. The former is a rich resource
from which to draw when acquiring a new language and literacy. Furthermore, L1 literacy
has been found to have a profound and long-lasting impact on the development of L2
literacy print acquisition and reading development (Koda, 1995, 2008). Understanding
cross-linguistic transfer for literacy acquisition may explain literacy developmental
trajectories in the case of different minority groups. In addition, pedagogical implications
would suggest highlighting relevant similarities or differences between L1 and the target
literacy in order to promote target literacy acquisition (Koda, 2008).
Note, however, that within the framework of bilingual research, there is growing
evidence to the role of script typology on cross-linguistic interaction between L1 and L2.
In this context, the script-dependent hypothesis (Geva & Siegel, 2000; Geva, Wade-
Woolley & Shany, 1993) claims that specific characteristics of L1 linguistic and
orthographic structure affect L2 literacy acquisition due to cross-linguistic transfer (Geva
& Siegel, 2000; Geva & Wade-Woolley, 1998). We suggest here that emphasis needs to
be placed not only on orthographic characteristics but also on the way the orthography
reflects the linguistic structure and the interplay between the languages involved. For
example, linguistic distance between Hebrew and English where the novel phoneme /y/
or /j/ represented by the grapheme th is both linguistically and orthographically
challenging due to the fact that digraphs do not exist in Hebrew. Specific characteristics
of any given orthography will determine the type of mapping with its underlying
processing that is required for successful decoding and encoding. Translated into practice
in any given literacy, children need to understand the orthographic-specific unit of word
138 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
analysis and how this maps onto phonemes or morphemes. This should then provide an
overall framework for subsequent literacy acquisition. If the unit of word analysis is
similar for L1 and the target orthography, for example consonant clusters in Russian and
English, learners will experience a similar process to their L1 literacy acquisition. In this
case, L2 basic literacy skills acquisition will be facilitated and strengthened by prior
literacy experience. A connectionist paradigm would explain the transition from effortful
mapping of orthographic units onto phonemes to automatic recognition in L2 literacy as
being a gradual and dynamic process (Koda, 2008). In this complex interplay between
developmental and cross-language factors, L2 literacy evolves, which provides a secure
basis for further L2 literacy acquisition (Genesee et al., 2006).
In the present study, the inclusion of two Russian–Hebrew-speaking groups who were
both exposed to Russian oral language from birth enabled us to study the interplay
between orthographic and linguistic characteristics of the language. A methodological
design that included a Russian–Hebrew-speaking emergent triliterate group facilitated
controlling for the role of Russian literacy over and above oral language. Accordingly, we
predicted that only the Russian–Hebrew-speaking emergent triliterates would benefit
from the complex syllabic structures (consonant clusters) in oral Russian. Their
experience with consonant clusters would act as a catalyst for greater phonemic
sensitivity. This sensitivity in turn would be expected to facilitate spelling and decoding
of English consonant clusters.
Basic literacy skills acquisition in EL2: an L1 Russian and Hebrew background
Phonemic awareness and word decoding
Phonemic awareness is an underlying mechanism of basic literacy skill acquisition.
Developing EL2 phonemic awareness reflects L1 orthographic experience (Bruck &
Genesee, 1995; Wade-Woolley, 1999). Thus, children with a background in an
orthography that is not analysed at a phonemic level may be expected to experience
phonemic awareness and synthesis as a greater challenge than children who have had
experience in phonemic awareness in their L1 orthography.
Like English, Russian Cyrillic script is a fully fledged alphabet, which is read from left
to right and has letters representing consonants and vowels. The Russian syllable
structure is very much more complex than the English syllable structure with Russian
syllables having up to four consonants in a single cluster (Akhmanova, 1971). This
phonological complexity has been found to have a positive influence on the development
of phonological awareness at the phonemic level. Thus, it has been found that Russian-
speaking preschoolers and first graders are able to perform more complex phonological
segmentation tasks than English- or Hebrew-speaking children (R. Ibrahim & Z. Eviatar,
unpublished data; Zaretsky, 2002).
The Russian orthography is considered to be relatively more transparent and less
challenging regarding letter–sound correspondence and decoding. However, spelling in
Russian is comparable with opaque orthographies with inconsistent grapheme–phoneme
correspondence such as English (Grigorenko, 2003). Reading instruction in Russian
schools stresses intensive phonemic awareness training by discrimination of phonemes,
phoneme blending and phoneme manipulation, distinguishing between consonants in
consonant clusters and between consonants and vowels, and segmentation of syllables
(Kerek & Neimi, 2009; Kornev, 1997; McEneaney, 1997).
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 139
Copyright r 2011 UKLA
Hebrew is a Semitic language with a consonantal orthography (or ‘abjad’), read from
right to left, and existing in two forms: pointed (fully vocalised by means of vowel
diacritics) and unpointed (partly vocalised) (Ravid, 2006; Share & Levin, 1999; Shimron,
1993). Both forms of the Hebrew script do not present vowels by fully fledged letters,
while initial acquisition of reading in pointed script is based on the usage of CV
subsyllabic segments as the basic units of decoding instruction. Early literacy instruction
in Hebrew does not stress the distinction between consonants and vowels (Bentin &
Leshem, 1993). Hebrew-speaking children may therefore be less sensitive to vowels as
separate phonemes. That is, these children struggle with progressing from the default
kamatz /a/ diacritic (e.g. 5 /ra/) CV segment to another CV segment with a different
diacritic mark (e.g. hirik /i/). This perhaps explains the fact that most reading errors in
Hebrew are vowel decoding errors (Share & Levin, 1999).
In addition, in contrast to Russian and English, Hebrew is characterised by relatively
simple syllable structures (CV and CVC). Consequently, we expected that Russian–
Hebrew-speaking emergent triliterates would experience success with phonemic
awareness and short vowel decoding in English (L3). This is due to their experience
with Russian (L1) literacy as a fully fledged alphabet with letters representing consonants
and vowels. We expected Russian–Hebrew-speaking emerging biliterates and Hebrew-
speaking emerging biliterates to find English phonemic awareness and short vowel
decoding to be challenging. This prediction is based on Schwartz et al.’s (2007) study,
which found that biliterate bilinguals performed significantly better than their
monoliterate bilingual and monolingual peers on a global score for phonemic awareness
in English as an L3.
Spelling acquisition
The English orthography is considered to be opaque (Seymour, Aro & Erskine, 2003;
Spencer, 2000, 2007). This implies that both decoding and encoding involve additional
mappings aside from direct translation of spelling to pronunciation. Acquisition of
symbol–sound correspondence of the English alphabet is merely a first step in literacy
acquisition. Additional orthographic processing involves automatic recognition of tens of
orthographic patterns making words graphemically complex (Davis & Bryant, 2006;
Spencer, 2007) as well as acquisition of words that have varying degrees of phonemic
irregularity, for example one where only the /n/ is phonemically regular. Successful
decoding and spelling include an understanding of the complex syllabic structure of
English, which facilitates pronunciation of multiple vowel combinations (e.g. closed
syllables with short vowel sounds such as, run, fun, up and phonemic exceptions such as
put where the /u/ does not have a short sound despite the syllable being closed;
Shankweiler & Fowler, 2004). According to Frost (2005), the multiple vowel system
(about 15 vowels) which is represented by fewer graphemes is what gives the English
orthography its opaque character. English word recognition is complex so that both
syllabic and morphological knowledge necessary for decoding multi-syllabic content
words of Greek and Latin origin need to be acquired. L1 English readers use this
knowledge from fourth grade onwards in order to ensure continued accurate and
independent decoding (Harm, McCandliss & Seidenberg, 2003; Henry, 2003;
McCandliss, Beck, Sandak & Perfetti, 2003). The various multiple-stage models that
describe English monolingual spelling development are testimony to the challenging
nature of the English orthography (Young, 2007). Indeed, it has become widely
140 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
recognised that English poses extraordinary difficulties for the young child learning to
read and write in English L1 (Davis & Bryant, 2006). This is even more so in the case of
EL2 and these difficulties appear to be directly rooted in the complexity of the
orthography (J. Kahn-Horwitz & Z. Goldstein, unpublished data; Spencer, 2000;
Wimmer & Goswami, 1994).
There has been less empirical examination of cross-linguistic transfer in spelling
between languages belonging to different families and represented by different scripts, as
in our case Hebrew, Russian and English. Existing research on development of EL2
spelling points out that children and college students transferred their knowledge of L1
phonology and phoneme-to-grapheme correspondence to their early EL2 spelling
(Fender, 2008; Tal, 2005; Van Berkel, 2005). In a review of L1 impact on EL2 spelling
development, Figueredo (2006) cites repeated evidence of facilitation and obstruction of
EL2 spelling depending on the characteristics of L1 orthographies and the distance
between L1 and L2 orthographies. At the pronunciation stage, English phonemes that do
not exist in L1 may be mispronounced and misrepresented graphemically in English
spelling. An example of this is the phoneme /y/ or /j/ represented by the consonant
digraph th that exists neither in Hebrew nor in Russian.
In accordance with the script-dependent hypothesis, concerning EL2 spelling
acquisition we predicted that Russian–Hebrew-speaking emerging triliterates would
outperform Russian–Hebrew-speaking emerging biliterates and Hebrew-speaking emer-
ging biliterates in the spelling and decoding of short vowels and consonant clusters. We
also hypothesised that all three groups would face similar difficulties in their spelling and
decoding of the following: foreign consonants, consonant digraphs, vowel digraphs and
foreign digraphs as novel orthographic conventions (see Table 1).
Method
Participants
A sample of 99 fifth-grade children were selected from five elementary schools located in
similar neighbourhoods in the northern region of Israel. The schools were characterised
by similar middle socioeconomic (SE) index. This index is calculated by the Israeli
Ministry of Education based on parents’ reports on their income, occupation and ranking
of family residential area. The government uses this index for the differential funding of
schools. In Israeli schools, the SE index is measured on a 10-point scale ranging from 1 to
10, with higher scores indicating lower SE status. In the present study, the SE index for
the schools was the following: 4.70, 4.90, 5.05, 5.05 and 5.1. Participant selection was
Table 1. Orthographic differences and similarities between Russian, English and Hebrew.
Orthographic conventions and patterns English Russian Hebrew
Consonant clusters Present Present Absent
Short vowel representation Present Present Absent
Foreign consonants (w, j) Present Absent Absent
Consonant digraphs (sh, ch) Present Absent Absent
Vowel digraphs (oo, ee) Present Absent Absent
Foreign digraphs (th) Present Absent Absent
Split digraph (silent ‘e’) Present Absent Absent
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 141
Copyright r 2011 UKLA
conducted in two stages. First, parents were approached during parent–teacher meetings.
The consent form for parents of both bilingual groups included questions concerning date
of arrival in Israel, ratings of language dominance at home, information on the level of
the child’s Russian literacy knowledge, length of exposure to Russian and English
literacy and whether children received private tutoring in English. The consent form for
parents of Hebrew-speaking emerging biliterates contained questions about parent
education, the age when the child began to receive English instruction and whether the
child received any private tutoring in English.
Next, based on the information provided by parents and teachers, we selected 53
Russian–Hebrew-speaking emerging triliterates who met the following exclusionary
criteria: (1) Russian is the first and dominant language in the home; (2) acquired basic
skills in Russian literacy between the ages of 5 and 6 years; (3) immigrated to Israel
before entering Grade 2; (4) began their English instruction in Grade 3; (5) did not
receive any supplementary private tutoring in English outside of school; (6) were not
diagnosed with learning disabilities, and severe hearing, visual and neurological
impairments (see Table 2). On the basis of the parental report, we selected 46
Hebrew-speaking emerging biliterates studying EL2 who shared the last three criteria
with their emerging triliterate peers. Finally, according to the information obtained from
parental reports of children’s Russian literacy knowledge, and the results of our own
testing of Russian decoding and spelling skills, the Russian–Hebrew speakers were
divided into two groups: (1) Russian–Hebrew-speaking emerging triliterates (were able to
read Russian and Hebrew with English as L3) (n 5 34) and (2) Russian–Hebrew-speaking
emerging biliterates (were able to read Hebrew with English as an L3) (n 5 19). Note that
the Russian–Hebrew-speaking emerging biliterates had rudimentary knowledge of the
Russian alphabet.
All Russian–Hebrew speakers were second-generation Russian–Jewish immigrants,
who lived in the north of Israel. Russian was the dominant home language, although
Hebrew was the primary language used outside of the home. Russian–Hebrew speakers’
reports regarding language practice at home and in the immediate environment described
the following: communication between children and grandparents occurred in Russian,
communication between parents and children was mostly in Russian, between siblings
communication was mainly in Hebrew for older siblings and in Russian for younger
siblings and between friends was mostly in Hebrew but with Russian-speaking friends
was sometimes in Russian.
Table 2. Group classification criteria.
Group L1
Russian
Russian literacy
acquisition
between 5 and
6 years old
Immigrated to
Israel before
Grade 2
Began English
instruction in
Grade 3
After school
English
tutoring
Diagnosed
learning
disabilities
Russian–Hebrew-
speaking emerging
triliterates
X X
Russian–Hebrew-
speaking emerging
biliterates
X X X
Hebrew-speaking
emerging biliterates
X X X X X
142 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
Research variables
Dependent variables were English literacy measures: pseudoword spelling and decoding.
Independent variables were Russian and Hebrew (L1 and L2) phoneme deletion,
pseudoword reading, spelling, and word reading accuracy and English phoneme analysis,
phoneme deletion and word recognition. General nonverbal ability was controlled for.
Measures
Baseline measures in Hebrew and Russian. These had a number of purposes. First, the
Russian literacy tests were conducted in order to distinguish between Russian–Hebrew-
speaking emerging triliterates and Russian–Hebrew-speaking emerging biliterates.
Second, Hebrew literacy tests facilitated controlling for differences in L1 or L2 between
the groups. Finally, the focus on Hebrew literacy tests enabled us to check for cross-
literacy links.
Phonemic awareness in Hebrew and Russian. Phoneme deletion (Hebrew version:
Shany, Lachman, Shalem, Bahat & Seiger, 2003; Russian version: Schwartz, 2006). This
measure consists of 20 words ranging from one to two syllables. The words represent the
following consonant/vowel structures: CVC, CCVC, CCVCV, CVCVC and CVCCV. For
test administration, the tester pronounced each word aloud. The participant repeated the
word after hearing it and then deleted a phoneme at the beginning, middle or end of the
word and pronounced the remaining (nonmeaningful) sequence. For example, in Hebrew,
‘Say matos (‘airplane’). Now, say matos without t’, correct answer would be ‘maos’; (In
Russian: shkola, ‘school’ without sh, correct answer would be ‘kola’). Two examples
were presented to the participants at the beginning of task administration. The total
number of correct responses was calculated. The maximum possible score is 20 for both
the Hebrew and the Russian task, respectively. The internal consistencies for the Hebrew
and Russian versions of this test (coefficient a) are .85 and .62, respectively.
Hebrew literacy tests
1. Word recognition accuracy (Shany et al., 2003). This test includes a list of 38 pointed
(vowelled) words varying in degree of difficulty, ranging from short, monomorphemic
words (e.g. dug, ‘fish’) to longer, multi-morphemic (e.g. , mitravxim, ‘they
get more comfortable’). The participants read these word after word aloud. Accuracy
scores were computed by allocating one point for each correctly pronounced word.
The maximum score on this task is 38. Internal consistency (coefficient a) is .90.
2. Pseudoword decoding (Deutsch, 1994). This list of 24 pseudowords was read aloud by
each participant. The number of errors (accuracy) was recorded. The maximum score
on this task is 24. Internal consistency (coefficient a) is .90.
3. Spelling (adapted from Shany et al., 2003). This test consists of 20 words representing
three different linguistic categories: (1) function words, (2) uninflected content words,
(3) inflected content words consisting of roots and word patterns different from those
used in the previous category. Participants listened to each target word, then to a
sentence including the word and again to the word in isolation. They were then
requested to write the target word. Internal consistency (coefficient a) is .87. The
maximum score for this task is 20.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 143
Copyright r 2011 UKLA
Russian literacy tests
The Russian literacy tests reflect the Russian orthography (Grigorenko, 2003; Inshakova,
2004; Kerek & Neimi, 2009) and were built on the basis of the Russian syllable structure
(Akhmanova, 1971; Kerek & Neimi, 2009).
1. Word recognition accuracy (Schwartz, 2006). Participants orally read a list of 20
words of increasing difficulty ranging from short, monomorphemic to longer, multi-
morphemic words with consonant clusters (e.g. grib [‘mushroom’], velosiped
[‘bicycle’]). Accuracy scores were computed by allocating one point for each
correctly pronounced word. The maximum score on this task is 20. Internal
consistency (coefficient a) is .63.
2. Pseudoword decoding (Schwartz, 2006). Participants decoded 20 pseudowords that
comply with Russian’s morpho-phonemic conventions. These words included mono-
and multi-syllabic pseudowords (e.g. porexi). Accuracy scores were computed by
allocating one point for each correctly pronounced word. The maximum score for this
task is 20. Internal consistency (coefficient a) is .90.
3. Spelling (Schwartz, 2006). Participants wrote 10 uninflected content words in L1
Russian. Participants first listened to the target word, then to a sentence including the
word and again to the word in isolation. They were then requested to write the target
word. One point was allocated for each correctly spelled word (e.g. kniga [‘book’]).
The maximum score for this measure is 10. Internal consistency (a) is .88.
General ability
In the Raven’s colored matrixes (Raven, Raven & Court, 1976, sets A, B, C, D and E), a
nonverbal ability test, participants matched one of six graphic patterns to a visual array.
English phonemic awareness and literacy tests
Baseline measures. In addition to Hebrew and Russian baseline measures, three English
measures were administered: phoneme deletion, phoneme analysis and word recognition.
1. Phoneme deletion (adapted from Rosner, 1975). Similar to Hebrew and Russian
phoneme baseline deletion tests, the adapted version of this test included 20 one-
syllable words that required the deletion of initial, middle or final phoneme. Each
word was presented orally, the participant repeated the word after hearing it and then
deleted a target phoneme and pronounced the remaining sequence, for example, ‘Say
gate. Now, say gate without g’. Internal consistency (coefficient a) is .74.
2. Phonemic analysis. Participants were asked to analyse words into constituent
phonemes. The tester pronounced the target word. The participant repeated the word
and was then asked to say the word like a robot would. Two examples were given
before the 20 target items were presented. The test was constructed to include 20
frequent monosyllabic words taken from the required vocabulary sections of fourth-
and fifth-grade elementary school English textbooks that the students studied from.
The test is made up of items including a closed syllabic structure including five basic
CVC items (e.g. dog), five items with a consonant cluster at the beginning of the word
CCVC (e.g. trip), five items with a consonant cluster at the end of the word CVCC
(e.g. milk) and five items with a consonant cluster at the beginning and end of the word
144 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
CCVCC (e.g. stand). Four sets of variables were used, each calculated separately: (1)
overall score was computed by allocating one point for each correctly analysed word
(maximum score for this task is 20); (2) a score was calculated for each correctly
analysed initial consonant cluster (maximum score for this task is 10); (3) a score was
allocated for each correctly analysed final consonant cluster (maximum score for this
task is 10); and (4) one point was allocated for analysing medial vowels (maximum
score for this task is 20). Internal consistency (coefficient a) was .87.
3. Word recognition (Woodcock reading mastery test-revised, form H: word recognition
subtest, Woodcock, 1987). This standardised test measures single-word reading skills.
Participants read aloud the list of words of increasing difficulty. After six consecutive
errors the test was discontinued. The maximum score on this task is 65. The test
manual reports split-half reliability coefficients ranging between 0.81 and 0.99.
Experimental measures. In order to check to what extent the linguistic and orthographic
structures of Russian versus Hebrew impacted on L2 spelling (script-dependent
hypothesis), a spelling test in English was constructed. The same items of this task
were reordered into a pseudoword reading task in order to examine the relationship
between spelling and decoding.
1. Pseudoword spelling (Kahn-Horwitz, Schwartz & Share, 2008). Two stages were
completed in our measure construction in order to increase the internal validity of this
tool. The first was screening 34 orthographic conventions that appear in four fifth-
grade English textbooks. Then based on our hypotheses, we analysed novel
orthographic conventions into those that were considered a challenge for all three
groups versus those that were considered a challenge for Russian–Hebrew-speaking
emerging triliterates as opposed to Russian–Hebrew-speaking and Hebrew-speaking
emerging biliterates (see explanation in introduction and Table 1). In the second stage,
12 experienced elementary school teachers were asked to rank to what extent students
at the end of fifth grade were familiar with each pattern. A word example was given
for each pattern. Teachers ranked each pattern from 1 (no exposure and instruction) up
to 5 (high level of exposure and instruction). Only patterns that received a ranking of 4
or 5 by at least 80% of teacher agreement were included as target items for dictations.
The following patterns were excluded based on this ranking: vowel diphthongs (ou,
oy, oi) foreign digraphs (oa) and r-controlled patterns (ur, ir). The final list included
eight orthographic conventions. Each group was represented by at least five items in
the final pseudoword list. Participants were asked to spell 34 pseudowords, which
were designed to investigate the script-dependent hypothesis. We distinguished
between two sets of items: the first set included English spelling conventions that exist
in Russian but not in Hebrew (examining Russian–Hebrew-speaking emerging
triliterate advantages over Russian–Hebrew-speaking and Hebrew-speaking emerging
biliterates). The second set incorporated English spelling conventions, which do not
exist in either of the target languages – Russian or Hebrew (in order to check for
general EL2 or L3 spelling difficulties) (see Table 1). Each target pseudoword was
pronounced twice by the English tester. After hearing the target pseudoword,
participants repeated the stimulus, and then wrote it. Scores were computed by
allocating one point for each correctly spelled item (maximum score for this task is
34). A separate score was calculated for each correctly spelled convention within each
of the categories (see Table 3). Internal consistency (coefficient a) is .72.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 145
Copyright r 2011 UKLA
2. Pseudoword decoding (Kahn-Horwitz et al., 2008). The same above items were
presented in a jumbled order for reading. Internal consistency (coefficient a) is .80.
Procedure
In this attribute group design study, we tested children’s L1 (Russian or Hebrew), L2
(Hebrew) and English (L2, L3) literacy measures at the end of Grade 5. At this point, all
children had been exposed to 3 school years of language and literacy instruction in
English (three lessons a week). The Russian literacy tasks were administered to the
Russian–Hebrew-speaking emerging triliterates. The Russian phonemic awareness task
was administered to both Russian–Hebrew-speaking groups. Hebrew and English
language and literacy measures were administered individually to all children.
For emerging triliterates and biliterates, Russian and Hebrew measures were
administered in the same session and counterbalanced. This session lasted about 45
minutes. For Hebrew-speaking emerging biliterates, Hebrew measures were administered
in one session lasting approximately 30 minutes. English measures were administered
during two separate sessions 2 weeks apart. We counterbalanced these measures in order
to present the pseudoword spelling and pseudoword decoding in separate sessions. The
Ravens test of nonverbal intelligence was administered in small groups of three children
at the end of one of the English sessions. The Hebrew tasks were administered by native
Hebrew speakers, while native Russian speakers administered the Russian tasks, and
native English-speaking English teachers and assessors administered the English tasks.
The instructions for each testing session were given in the language being tested except
for the English tasks where the instructions were administered in Hebrew in order to
avoid any misunderstanding.
Results
Table 4 presents the background data for bio-social and nonverbal ability for the three
groups. In order to examine differences between the groups on the variables mentioned
above, we conducted one-way analysis of variance followed up with post hoc
comparisons. In most analyses, the significant differences between the groups are
indicated by means of Latin superscripts (a, b), with different superscripts indicating a
statistically significant difference; groups sharing a common superscript do not differ
significantly.
Table 3. English orthographic category scoring for experimental pseudoword decoding and spelling.
Category Example
1. Consonant cluster ‘spr’ in sprit
2. Short vowel ‘u’ in snup
3. Foreign consonants ‘w’ in wid
4. Consonant digraphs (ch and sh) ‘sh’ in fosh
5. Foreign digraph (th) ‘th’ in swith
6. Vowel digraph (ee and oo) ‘oo’ in stoon
7. Split vowel digraph ‘e’ zome
8. r-controlled combination ‘ar’ in yarb
146 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
It can be seen that the three groups did not differ significantly in age and nonverbal IQ.
However, there was a significant difference between all three groups in gender with a
smaller number of girls in the Russian–Hebrew-speaking emerging biliterate group
compared with a preponderance of girls in the Russian–Hebrew-speaking emerging
triliterate group (w2(1) 5 12.3, p 5 .00). This difference will be addressed in the following
data presentation.
Group comparisons on baseline spoken and written English, Hebrew and Russian
language measures are presented in Table 5. Table 5 also includes L1 Russian literacy
scores.
The data reveal that all three groups received similar high scores for Hebrew and
English phoneme deletion. In addition, there were no significant differences between
groups for any of the English phoneme analysis tasks except for medial vowel analysis
where the Russian–Hebrew-speaking emerging triliterates outperformed the Russian–
Hebrew-speaking emerging biliterates but not the Hebrew-speaking emerging biliterates.
Because of the numerical differences between the Hebrew-speaking emerging biliterates
and the Russian–Hebrew-speaking emerging triliterates, we created a new emergent
biliterate group by combining Russian–Hebrew-speaking emerging biliterate and
Hebrew-speaking emerging biliterate groups. This combined group was compared with
the Russian–Hebrew-speaking emerging triliterate group on medial vowel analysis.
Independent t test results showed significant differences between the groups on vowel
analysis F(1, 97) 5 4.15, po.05.
The Russian–Hebrew-speaking emerging triliterates outperformed Russian–Hebrew-
speaking emerging biliterates on the phoneme deletion task in Russian (L1). Concerning
groups’ performance on Hebrew (L1, L2) measures, it is notable that the Hebrew-
speaking emerging biliterates performed better than both Russian–Hebrew-speaking
groups on Hebrew spelling. These results can be attributed to the complexity of spelling
in Hebrew (Ravid, 2006). More specifically, because the Hebrew orthography is
characterised by many-to-one phoneme-to-grapheme relationships, with a number of
pairs of graphemes representing the same phoneme (e.g. /k/), it seems that the
findings point to a possible delay in Hebrew spelling among Russian–Hebrew-speaking
children. Hebrew spelling ability has been found to be strongly related to
morphosyntactic awareness and lexical knowledge (Levin, Ravid & Rapaport, 1999;
Ravid, 2006). We did not collect data on children’s morphosyntactic and lexical
knowledge, so we can only speculate that these second-generation immigrants from the
former Soviet Union still have relatively less-developed Hebrew (L2) vocabulary.
Finally, Table 5 shows that the performance of Russian–Hebrew-speaking emerging
triliterates on Russian reading measures was near to maximum. These same triliterate
Table 4. Means and standard deviations (in parentheses) on background bio-social factors and nonverbal
intelligence.
Variables RHS emergent
triliterate (n 5 34)
RHS emergent
biliterate (n 5 19)
HS emergent
biliterate (n 5 46)
F w2
Age (years:months) 10.8 (0.56) 10.8 (0.42) 10.7 (0.51) .53 –
Gender (boys:girls) 12:23 16:3 24:22 – 12.30*
Nonverbal intelligence 67.9 (17.56) 70.9 (13.30) 68.3 (10.94) .30 –
Note: RHS 5 Russian–Hebrew speaking; HS 5 Hebrew speaking.*po.005.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 147
Copyright r 2011 UKLA
children were less successful at spelling in L1 compared with reading. A possible
explanation for this may be the complexity of Russian spelling, which is characterised by
highly opaque phoneme-to-grapheme correspondence (Grigorenko, 2003; Inshakova,
2004; Kerek & Neimi, 2009) compared with decoding, which reflects relatively
consistent phoneme–grapheme correspondence. Russian–Hebrew-speaking emerging
triliterates may have also had more practice with reading as opposed to spelling as
they reported informally to the researchers.
For English phoneme analysis, all three groups reached close to the ceiling effect for
initial consonant clusters and high scores were obtained for final consonant cluster
analysis. Numerically, the Russian–Hebrew-speaking emerging triliterates outperformed
the two emerging biliterate groups on vowel analysis as predicted. However, significant
differences were only found between the Russian–Hebrew-speaking emerging triliterates
(M 5 67.9, SD 5 27.86) who were at an advantage to the Russian–Hebrew-speaking
emerging biliterates (M 5 46.6, SD 5 32.75), whereas Hebrew-speaking emerging
biliterates (M 5 57.3, SD 5 33.35) fell in between these two, F (2, 96) 5 2.92, po.05.
The performance of the three groups on English pseudoword spelling is reported in
Table 6.
As predicted, there were no significant differences between groups on the novel
orthographic conventions in English that were considered a challenge for all participants.
Table 5. Means and standard deviations (in parentheses) on baseline spoken and written English, Hebrew
and Russian language measures as percentage values.
Variables (range) RHS emergent
triliterate (n 5 34)
RHS emergent
biliterate (n 5 19)
HS emergent
biliterate (n 5 46)
F
English
Phoneme deletion – Rosner (1–20) 91.3 (9.72) 89.0 (11.38) 84.7 (15.29) 2.70
Phoneme analysis – initial
consonant clusters (1–10)
95.6 (8.23) 94.7 (13.48) 95.6 (11.28) 0.05
Phoneme analysis – final consonant
cluster (1–10)
81.8 (30.00) 89.5 (13.93) 89.8 (22.75) 1.19
Phoneme analysis – medial vowel
(1–20)
67.9a (27.85) 46.6b (32.75) 57.3a,b (33.35) 2.92*
Phoneme analysis – overall (1–20) 63.7 (29.27) 43.7 (31.53) 55.8 (33.37) 2.43
Hebrew
Phoneme deletion (1–20) 87.5 (12.87) 79.0 (12.76) 84.0 (14.93) 2.33
Word recognition accuracy (1–38) 93.8 (5.38) 91.8 (6.20) 93.0 (5.12) 0.82
Pseudoword decoding accuracy
(1–24)
87.5 (10.81) 86.8 (12.37) 82.3 (15.28) 1.75
Spelling accuracy (1–20) 82.7b (22.87) 79.7b (17.83) 90.7a (9.17) 3.80*
t
Russian
Phoneme deletion (1–20) 96.2 (5.31) 90.3 (7.90) – 3.25*
Word recognition accuracy (1–20) 95.0 (6.12) – – –
Pseudoword decoding accuracy (1–20) 92.9 (9.10) – – –
Spelling accuracy (1–10) 65.8 (22.48) – – –
Note: RHS 5 Russian–Hebrew speaking; HS 5 Hebrew speaking.Because of close to significant p value, we conducted a one-way analysis of variance with LSD post-testresulting in significant differences between groups which are marked by means of Latin superscripts (a, b).*po.05.
148 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
However, the Russian–Hebrew-speaking emerging triliterates outperformed both
comparison groups on spelling of short vowels, which, as we expected, is considered a
challenge for emerging biliterates. At the same time, there were no significant differences
between the Russian–Hebrew-speaking emerging triliterates and the two other groups on
consonant cluster spelling.
For the pseudowords decoding measure (see Table 7), the Russian–Hebrew-speaking
emerging triliterates outperformed both comparison groups on the decoding of consonant
clusters. For short vowel decoding, the Russian–Hebrew-speaking emerging triliterates
were significantly superior to the Hebrew-speaking emerging biliterates. In addition, they
Table 6. Means and standard deviations (in parentheses) for pseudoword spelling in English as percentage
values and w2 values for nonparametric comparisons between the groups.
Variables RHS
emergent
triliterate
(n 5 34)
RHS
emergent
biliterate
(n 5 19)
HS emergent
biliterate
(n 5 45)
F w2 Number
of test
items
Pseudoword spelling (overall) 41.4 (13.81) 37.8 (13.38) 37.7 (19.03) 0.56 – 1–34
Consonant clusters spelling 92.3 (10.02) 91.5 (10.22) 87.5 (17.38) 1.29 0.85 1–13
Short vowels spelling 60.5a (13.07) 47.4b (20.79) 48.9b (24.32) 5.45* 6.06* 1–14
Foreign consonants spelling 69.4 (25.22) 66.3 (26.71) 72. 9 (27.02) 0.45 1.29 1–5
Consonant digraphs (‘ch’, ‘sh’) 82.4 (31.05) 82.1 (28.20) 66.7 (35.42) 2.78 8.2* 1–5
Consonant digraphs (‘th’) 30.0 (35.93) 34.7 (39.35) 26.2 (35.37) 0.38 0.6 1–5
Vowel digraphs (‘ee’, ‘oo’) 33.5 (27.29) 22.1 (22.99) 34.7 (25.01) 1.74 3.58 1–5
Split digraph (silent ‘e’) 21.2 (29.10) 23.2 (26.05) 22.2 (24.58) 0.04 0.59 1–5
‘r’ controlled 83.5 (15.15) 69.5 (23.45) 78.7 (25.37) 2.50 4.87 1–5
Note: RHS 5 Russian–Hebrew speaking; HS 5 Hebrew speaking. Superscripts (a,b) are used to show significantdifferences between groups.*po.05.
Table 7. Means and standard deviations (in parentheses) for pseudoword decoding and word recognition
accuracy in English as percentage values and w2 values for nonparametric comparisons between the groups.
Variables RHS emergent
triliterate
(n 5 34)
RHS emergent
biliterate
(n 5 19)
HS emergent
biliterate
(n 5 45)
F w2 Number
of test
items
Woodcock word recognition 61.0a (11.96) 56.0a,b (14.60) 51.6b (13.75) 4.87* – 1–65
Pseudoword decoding (overall) 63.7a (15.96) 55.4a,b (17.49) 52.2b (21.82) 3.58* – 1–34
Consonant clusters decoding 95.7a (5.42) 88.7b (14.59) 90.5b (12.89) 3.07*c
3.60* 1–13
Short vowels decoding 75.8a (14.29) 68.4a,b (14.74) 63.8b (18.89) 5.08* 8.30* 1–14
Foreign consonants decoding 96.1 (9.23) 93.0 (16.02) 93.8 (11.31) 0.54 1.12 1–5
Consonant digraphs (‘ch’, ‘sh’) 88.2a (15.66) 85.3a,b (21.95) 73.5b (35.79) 3.05*c
1.97 1–5
Foreign consonant digraphs (‘th’) 38.2 (39.19) 36.8 (33.51) 34.4 (35.16) 0.11 2.5 1–5
Vowel digraphs (‘ee’, ‘oo’) 80.0 (25.11) 68.4 (30.78) 68.7 (30.30) 1.76 3.22 1–5
Split digraph (silent ‘e’) 56.5 (35.67) 46.3 (30.59) 38.7 (33.84) 2.69 5.04 1–5
‘r’ controlled 87.7 (13.94) 79.0 (22.58) 77.4 (24.44) 2.46 3.26 1–5
Note: RHS 5 Russian–Hebrew speaking; HS 5 Hebrew speaking.cBecause of close to significant p value, we conducted one-way analysis of variances with an LSD post-testresulting in significant differences between groups which are marked by means of Latin superscripts (a, b).*po.05.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 149
Copyright r 2011 UKLA
showed close to significant differences with the Russian–Hebrew-speaking emerging
biliterates. As both emergent biliterate groups showed similar results on short vowel
decoding, it was decided to create a new emergent biliterate group by combining these
two groups. This was carried out in order to examine the effect of triliteracy on the
decoding of short vowels. The results of the t test with a set at .01 showed that as
hypothesised the emerging triliterates outperformed the combined emergent biliterate
group, F(1, 97) 5 1.49, p 5 .003.
The Russian–Hebrew-speaking emerging biliterates showed a significant advantage
over the Hebrew-speaking emerging biliterates on the overall measure of pseudowords
decoding and word recognition accuracy. Finally, there was a marginally significant
difference between the Russian–Hebrew-speaking emerging triliterates and the Hebrew-
speaking emerging biliterates on the decoding of consonant digraphs. These data could be
attributed to the fact that one of the two target phonemes , which is represented by
the digraph ‘ch’, does not exist in the Hebrew language and therefore is novel for the
Hebrew-speaking emerging biliterates but not for Russian–Hebrew-speaking emerging
triliterates.
Because of the different number of items for spelling and decoding of short vowels as
opposed to other orthographic conventions, we ran nonparametric comparisons (Kruskal–
Wallis tests) to examine whether we would find that the differences we observed for the
parametric tests remained consistent for the w2 results (see Tables 6 and 7). We found a
similar pattern of results for nonparametric tests. In addition, because the previous
analyses had identified a significant difference in the gender composition of the groups,
the analyses of Tables 6 and 7 were rerun with boys only. These analyses yielded the
same pattern of results. To summarise, this research points to clear-cut evidence that
Russian may facilitate spelling and decoding in English.
Discussion
This study, which was inspired by the script-dependent hypothesis, aimed to check the
following two hypotheses. The first hypothesis examined whether Russian–Hebrew-
speaking emerging triliterates were superior to Russian–Hebrew-speaking emerging
biliterates and Hebrew-speaking emerging biliterates in the spelling and reading of short
vowels and consonant clusters. The second hypothesis examined whether all three groups
would experience similar challenges regarding their spelling and reading of English:
foreign consonants, consonant digraphs, vowel digraphs and foreign digraphs as novel
orthographic conventions.
Our first hypothesis was fully supported with regard to spelling and decoding of
English short vowels. Because of similarities between the Russian and English
orthographies, in the present study, we expected and indeed observed a more direct
interaction between the target scripts. The advantage of Russian–Hebrew-speaking
emergent triliterates over the two emergent biliterate groups may be explained by the
facilitating effect of knowledge of the Russian orthography. It is notable, however, that
even the most advantaged group, the Russian–Hebrew-speaking emerging triliterates did
not achieve anywhere close to ceiling results for either spelling or decoding of short
vowels after 3 years of English literacy instruction. This clearly reflects objective
challenges presented by the English orthography (Wimmer & Landerl, 1997; Young,
2007).
150 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
Our results that all three groups obtained relatively high scores for consonant cluster
decoding and spelling were unexpected. These results reflect the ceiling effects obtained
by all for phoneme analysis of initial and final consonant clusters. For decoding
consonant clusters, there was a marginally significant advantage for Russian–Hebrew
emerging triliterates over the other two groups; however, no difference was observed for
consonant cluster spelling. We may have observed our originally expected advantage for
Russian–Hebrew emerging triliterates had we examined their reading and spelling at the
very beginning of their English literacy acquisition in third grade. Another explanation
could be that the Hebrew syllabic structure also includes consonant clusters albeit
consisting only of two consonants. To complete the picture, as a result of globalisation
and English being a lingua franca internationally there has been a continual influx of
English words into languages around the world of which Hebrew is no exception
(Lundberg, 2010). These borrowed words that children are exposed to daily and in
different contexts include brand names (e.g. Sprite, McDonalds), commercial words (e.g.
bank) and particularly words connected to the Internet (e.g. Skype). It could be that a
combination of exposure to consonant clusters from the aforementioned sources coupled
with experience with English over 3 years may have boosted their abilities to decode and
spell consonant clusters. The three groups’ success with consonant clusters may also
reflect that this aspect of the English orthography is less challenging as opposed to the
multiple graphemic and phonemic vowel patterns.
Despite the similar high scores by all three groups for initial and final consonant cluster
analyses, they showed a notably lower result for medial vowel analysis. The significant
advantage for the Russian–Hebrew-speaking emergent triliterates over the other two
groups on vowel analysis reflects that biliterates were less sensitive to vowels. We
supposed that this might be due to the fact that children who were exposed only to
Hebrew before English had less experience with an orthography that demands phonemic
sensitivity and which distinguishes between consonants and vowels. This does not occur
for Hebrew literacy instruction where CV blocks are the most basic phonological and
orthographic unit. This advantage converges with spelling and decoding advantages
shown by the Russian–Hebrew-speaking triliterates, which was supported by correlations
of around .45, po.05 between phonological analysis, reading and spelling.
The second hypothesis predicted equal difficulties for all three groups in their reading
and spelling of foreign and novel consonants, consonant digraphs, vowel digraphs and
foreign digraphs. Digraphs contribute to English grapheme complexity and have been
found to impact significantly on the correct spelling and decoding of English L1 children
of different ages (Davis & Bryant, 2006; Spencer, 2007). It seems that this complexity
impacts similarly and has greater delaying effects on young children acquiring EL2
literacy who have not had L1 experience with these novel phonemes and the concept of a
digraph.
Our study found that the foreign consonant digraph th was extremely challenging for
all three groups and this could be because for all participants, /y/ and /j/ are novel
phonemes and th is a novel orthographic pattern. All three groups also experienced
difficulty reading and spelling the split digraph (silent ‘e’), which appears to be a
challenging digraph for English L1 speakers as well (Davis & Bryant, 2006). In addition,
all three groups had difficulty spelling vowel digraphs ee and oo, yet experienced greater
success reading these digraphs. At the same time, the consonant digraphs sh and ch were
easier for reading for the Russian–Hebrew emerging triliterates and biliterates as opposed
to the Hebrew-speaking emerging biliterates and for spelling for all three target groups.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 151
Copyright r 2011 UKLA
The differences in decoding may be a result of the fact that the digraphs ch and sh exist in
Russian as phonemes, even though they are represented by single graphemes and
not digraphs. As for Hebrew, there is only the phoneme , which is also represented
by a single grapheme. It appears that when EL2 students have exposure to a phoneme in
their L1s, it is easier to acquire its grapheme correspondence. This seems to be the case
despite them not having experience with digraphs in their L1 literacy. When they have
neither experience with a novel phoneme nor its orthographic representation in the
additional language, for example ch in the case of Hebrew, or th and the split digraph
(silent e) in the case of both Russian and Hebrew, then this appears to be a very
significant obstacle in literacy acquisition in the additional language even after 3 years of
exposure. In contrast to digraphs, our results found that singleton consonants even when
they were novel as in the case of foreign consonants w and j as well as ‘r-controlled’
syllables with or and ar did not appear to challenge these emerging bi- and triliterate
children.
Within a broader theoretical context, our findings regarding trilinguals support existing
evidence from bilinguals where differences and interactions between not only
orthographic but also linguistic features (e.g. syllable structure and novel phonemes)
impacted acquisition of the L2 or L3 orthography. This supports proposing that the script-
dependence hypothesis be viewed in a broader sense, which includes linguistic
dimensions as well as emphasising the impact of linguistic and orthographic proximity
between target languages. For example, the case of the phoneme represented
orthographically by ch. In the present study, the Russian–Hebrew-speaking children were
exposed to this phoneme, although it is not represented in the Russian script by a digraph
as in English. They exhibited an advantage because of this linguistic proximity as
opposed to the Hebrew-speaking emerging biliterates. In Hebrew, the phoneme only
appears in a few borrowed words. These findings lead us to suggest the linguistic and
orthographic proximity hypothesis that should be explored experimentally between
different linguistic and orthographic contexts. This kind of research would examine
languages and scripts that have phonemic properties in common but greater orthographic
representation differences and vice versa.
Concerning a triliteracy advantage, it might be asked whether the gains observed
among our Russian–Hebrew-speaking emerging triliterates are attributable to their access
to two orthographic systems when acquiring L3 literacy. Alternatively, perhaps these
gains are due to the fact that the two orthographies in question were typologically related
(Russian and English) versus rather distant scripts (Hebrew and English). In other words,
does triliteracy represent a unique and complex phenomenon, which differs qualitatively
from biliteracy? We conclude that the specific advantage of emerging Russian–Hebrew-
speaking triliterates on short vowel decoding and spelling is a result of the similar way
the Russian and English scripts graphically represent vowels and not due to a triliteracy
experience per se. In order to further investigate the proposed language and orthographic
proximity hypothesis, additional research within a triliteral context is needed to examine
the languages that are linguistically and orthographically distant (e.g. Arabic [L1],
Hebrew [L2] versus English [L3]).
With regard to implications for the field, our results suggest that teachers might stress
similarities and differences between the scripts and use children’s knowledge of L1
literacy to enhance their cross-linguistic awareness (Cummins, 2008). Teachers could
assist multiliterate students in exploring their L1 literacy as a supporter language due to
typological closeness (in the case of Russian literacy) with English as a lingua franca.
152 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
The identification of orthographic patterns, which are difficult for both emerging
triliterate and biliterate children and which we have evidence for being challenging for
L1 English children, should be taught in a very much more explicit and gradual manner.
This would involve repetition of the target phonemes, identification of key lexical items
that would act as a mnemonic for these patterns, practice writing the target graphemes as
well as reading and spelling of words that include target patterns until automatic
recognition is achieved. This type of research-based teaching may serve to combat
students reaching seventh grade without being able to decode and encode effectively in
EL2 (Kahn-Horwitz & Ressissi, 2008).
Acknowledgements
This research was jointly supported by a grant from The MOFET Institute and the
Department of Teacher Education at the Ministry of Israel, 2009–2010, to the first author,
and a grant from Oranim College of Education Graduate Department to the first two
authors.
References
Akhmanova, O.S. (1971). Phonology, morphophonology, morphology. The Hague: Mouton.
Aronin, L. & OLaoire, M. (2003). Multilingual students’ awareness of their language teachers’ other languages.
Language Awareness, 12, 204–219.
August, D. & Shanahan, T. (Eds.) (2006). Developing literacy in second-language learners: A report of the
National Literacy Panel on Language-Minority Children and Youth (0-8058-6077-0). Mahwah, NJ: Erlbaum.
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.
Bild, E.R. & Swain, M. (1989). Minority language students in a French immersion program: Their French
proficiency. Journal of Multilingual and Multicultural Development, 10, 255–274.
Bruck, M. & Genesee, F. (1995). Phonological awareness in young second language learners. Journal of Child
Language, 22, 307–324.
Cenoz, J. & Genesee, F. (1998). Psycholinguistic perspectives on multilingualism and multilingual education.
In J. Cenoz & F. Genesee (Eds.), Beyond bilingualism: Multilingualism and multilingual education
(pp. 16–32). Clevedon: Multilingual Matters.
Cenoz, J. & Hoffmann, C. (2003). Acquiring a third language: What role does bilingualism play? (Introduction).
International Journal of Bilingualism, 7, 1–5. doi:10.1177/13670069030070010101.
Cummins, J. (1978). Bilingualism and the development of meta-linguistic awareness. Journal of Cross-Cultural
Psychology, 9, 131–149.
Cummins, J. (2008). Teaching for transfer: Challenging the two solitudes assumption in bilingual education. In
J. Cummins & N.H. Hornberger (Eds.), Bilingual education (pp. 65–75). New York: Springer.
Davis, C. & Bryant, P. (2006). Causal connections in the acquisition of an orthographic rule: A test of Uta
Frith’s developmental hypothesis. Journal of Child Psychology and Psychiatry, 47, 849–856. doi:10.1111/
j.1469-7610.2006.01597.x.
Deutsch, A. (1994). Operating of attention systems in the field of the syntax in children with severe reading
disabilities. PhD thesis (in Hebrew), The Hebrew University, Jerusalem.
Errasti, M.P.S. (2003). Acquiring writing skills in a third language: The positive effects of bilingualism.
International Journal of Bilingualism, 7, 27–52. doi:10.1177/13670069030070010301.
Fender, M. (2008). Spelling knowledge and reading development: Insights from Arab ESL learners. Reading in
a Foreign Language, 20, 1–25. Retrieved from http://nflrc.hawaii.edu/rfl/April2008/fender/fender.pdf
Figueredo, L. (2006). Using the known to chart the unknown: A review of first-language influence on the
development of English-as-a-second-language spelling skills. Reading and Writing, 19, 873–905.
doi:10.1007/s11145-006-9014-1.
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 153
Copyright r 2011 UKLA
Frost, R. (2005). Orthographic systems and skilled word recognition processes in reading. In M.S. Snowling &
C. Hulme (Ed.), The science of reading: A handbook (pp. 272–295). Oxford: Basil Blackwell.
Genesee, F., Geva, E., Dressler, D. & Kamil, M. (2006). Synthesis: Cross-linguistic relationships. In D.L.
August & T. Shanahan (Ed.), Developing literacy in a second language: Report of the National Literacy
Panel (pp. 153–174). Mahwah, NJ: Erlbaum.
Geva, E. & Ryan, E.B. (1993). Linguistic and cognitive correlates of academic skills in first and second
languages. Language Learning, 43, 5–42.
Geva, E. & Siegel, L.S. (2000). Orthographic and cognitive factors in the concurrent development of basic
reading skills in two languages. Reading and Writing, 12, 1–30.
Geva, E. & Wade-Woolley, L. (1998). Component processes in becoming English-Hebrew biliterate. In A.Y.
Durgunoglu & L. Verhoeven (Eds.), Literacy development in a multilingual context: Cross-cultural
perspectives (pp. 85–110). Mahwah, NJ: Erlbaum.
Geva, E., Wade-Woolley, L. & Shany, M. (1993). The concurrent development of spelling and decoding in two
different orthographies. Journal of Reading Behavior, 25, 383–406.
Gottardo, A., Yan, B., Siegel, L. & Wade-Woolley, L. (2001). Factors related to English reading performance in
children with Chinese as a first language: More evidence of cross-language transfer of phonological
processing. Journal of Educational Psychology, 93, 530–542. doi:10.1037/0022-0663.93.3.530.
Grigorenko, E.L. (2003). The difficulty of mastering reading and spelling in Russian. In N. Goulandris (Ed.),
Dyslexia in different languages cross-linguistic comparisons (pp. 92–111). London: Whurr.
Harm, M.W., McCandliss, B.D. & Seidenberg, M.S. (2003). Modeling the success and failures of interventions
for disabled readers. Scientific Studies of Reading, 7, 155–182.
Henry, M.K. (2003). Unlocking literacy: Effective decoding and spelling instruction. Baltimore, MD: Paul H.
Brookes.
Inshakova, O. (2004). Dyslexia in Russian. In I. Smythe, J. Everatt & R. Salter (Eds.), International book of
dyslexia: A cross language comparison and practice guide (pp. 173–179). Chichester: John Wiley.
Jessner, U. (1999). Meta-linguistic awareness in multilinguals: Cognitive aspects of third language learning.
Language Awareness, 8, 201–209.
Jessner, U. (2006). Linguistic awareness in multilinguals: Cognitive aspects of third language learning.
Edinburgh: Edinburgh University Press.
Kahn-Horwitz, J. & Ressissi, N. (2008). A multi-year intervention program and its impact on young English-
as-an-additional-language students in three municipal outlying areas: An evaluation research. Paper
presented at the 9th annual research conference of Oranim Academic College of Education, Tivon, Israel.
Kahn-Horwitz, J., Schwartz, M. & Share, D.L. (2008). Pseudo word spelling. Unpublished test, Oranim
Academic College of Education, Tivon, Israel.
Kahn-Horwitz, J., Shimron, J. & Sparks, R.L. (2005). Predicting foreign language reading achievement in
elementary school students. Reading and Writing, 18, 527–558. doi:10.1007/s11145-005-3179-x.
Kerek, E. & Neimi, P. (2009). Russian orthography and learning to read. Reading in Foreign Language, 21,
1–21. Retrieved from http://nflrc.hawaii.edu/rfl/April2009/articles/kerek.pdf
Koda, K. (1995). Cognitive consequences of L1 and L2 orthographies. In I. Taylor & D.R. Olson (Eds.), Scripts
and literacy (pp. 311–326). Dordrecht: Kluwer.
Koda, K. (2008). Impacts of prior literacy experience on second-language learning to read. In K. Koda & A.M.
Zehler (Eds.), Learning to read across languages: Cross-linguistic relationships in first- and second-language
literacy development (pp. 68–93). New York: Routledge.
Kornev, A.N. (1997). [Difficulties in reading and
writing among children: Assessment, intervention and prevention] (in Russian). Saint-Petersburg: Science.
Leung, Y.I. (2005). L2 vs L3 initial state: A comparative study of the acquisition of French DPs by Vietnamese
monolingual and Cantonese–English bilinguals. Bilingualism: Language and Cognition, 8, 39–61.
Levin, I., Ravid, D. & Rapaport, S. (1999). Developing morphological awareness and learning to write: A two
way street. In T. Nunes (Ed.), Learning to read: An integrated view from research and practice (pp. 248–
289). Amsterdam: Kluwer.
Lundberg, G. (2010). The impact of out-of-school language exposure on early language learning. Paper
presented at the 18th Congress of the Socio-linguistic Symposium, Southampton, England.
McCandliss, B.D., Beck, I.L., Sandak, R. & Perfetti, C. (2003). Focusing attention on decoding for children with
poor reading skills: Design and preliminary tests of the Word Building Intervention. Scientific Studies of
Reading, 7, 75–104.
McEneaney, J.E. (1997). Teaching them to read Russian: Four hundred years of the Russian bukvar. The
Reading Teacher, 51, 210–226.
154 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA
Muter, V. & Diethelm, K. (2001). The contribution of phonological skills and letter knowledge to early reading
development in a multilingual population. Language Learning, 51, 187–219.
Raven, J., Raven, J.C. & Court, J.H. (1976). Raven’s progressive matrices. Oxford: Oxford Psychologists Press.
Ravid, D. (2006). Hebrew orthography and literacy. In R.M. Joshi & P.G. Aaron (Eds.), Handbook of
orthography and literacy. (pp. 339–363). Mahwah, NJ: Erlbaum.
Rosner, J. (1975). Helping children overcome learning difficulties: A step-by-step guide for parents and
teachers. New York: Walker and Company.
Schwartz, M. (2006). The impact of literacy acquisition in L1 Russian on literacy acquisition in L2 Hebrew and
in L3 English among Russian-speaking (L1) children: Bi-literate bilingualism versus mono-literate
bilingualism. PhD thesis (in Hebrew), University of Haifa, Haifa.
Schwartz, M., Geva, E., Leikin, M. & Share, D.L. (2007). Learning to read in English as L3: The cross-linguistic
transfer of phonological processing skills. Written Language and Literacy, 10, 25–52.
Schwartz, M., Leikin, M. & Share, D.L. (2005). Bi-literate bilingualism versus mono-literate bilingualism:
A longitudinal study of reading acquisition in Hebrew (L2) among Russian-speaking (L1) children. Written
Language and Literacy, 8, 179–207.
Seymour, P.H.K., Aro, M. & Erskine, J.M. (2003). Foundation literacy acquisition in European orthographies.
British Journal of Psychology, 94, 143–174.
Shankweiler, D. & Fowler, A.E. (2004). Questions people ask about the role of phonological processes in
learning to read. Reading and Writing, 17, 483–515.
Shany, M., Lachman, D., Shalem, T., Bahat, A. & Seiger, T. (2003). Reading test: Word recognition fluency
(words-per-minute) and accuracy [in Hebrew]. Israel: Nitzan.
Share, D.L. & Levin, I. (1999). Learning to read and write in Hebrew. In M. Harris & G. Hatano (Eds.), Learning to
read and write: A cross-linguistic perspective (pp. 89–111). Cambridge: Cambridge University Press.
Shimron, J. (1993). The role of vowels in reading: A review of studies of English and Hebrew. Psychological
Bulletin, 114, 52–67.
Sparks, R.L. & Ganschow, L. (1993). The impact of native language learning problems on FL learning: Case
study illustrations of the linguistic coding deficit hypothesis. The Modern Language Journal, 77, 58–74.
Spencer, K. (2000). Is English a dyslexic language? Dyslexia, 6, 152–162.
Spencer, K. (2007). Predicting children’s word-spelling difficulty for common English words from measures of
orthographic transparency, phonemic and graphemic length and word frequency. British Journal of
Psychology, 98, 305–338. doi:10.1348/000712606X123002.
State of Israel Ministry of Education English Inspectorate (2009). Guidelines for the teaching of English at the
pre-foundation level. Retrieved from http://cms.education.gov.il/EducationCMS/Units/Mazkirut_Pedagogit/
English/Publications/
Tal, M. (2005). An analysis of the English spelling difficulties of native Hebrew speakers. MA thesis, University
of Haifa, Israel.
Valencia, J.F. & Cenoz, J. (1992). The role of bilingualism in foreign language acquisition. Journal of
Multilingual and Multicultural Development, 13, 433–449.
Van Berkel, A. (2005). The role of phonological strategy in learning to spell in English as a second language. In
V. Cook & B. Bassetti (Eds.), Second language writing systems (pp. 97–121). Clevedon: Multilingual
Matters.
Wade-Woolley, L. (1999). First language influences on second language word reading: All roads lead to Rome.
Language Learning, 49, 447–471.
Wimmer, H. & Goswami, U. (1994). The influence of orthographic consistency on reading development: Word
recognition in English and German children. Cognition, 51, 91–103.
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 (pp. 81–96). Mahwah, NJ: Erlbaum.
Woodcock, R. (1987). Woodcock reading mastery test-revised. Circle Pines, MN: American Guidance.
Young, K. (2007). Developmental stage theory of spelling: Analysis of consistency across four spelling-related
activities. Australian Journal of Language and Literacy, 30, 203–220.
Zaretsky, E. (2002). Effects of oral language on sound segmentation skills: Cross linguistic evidence. In W. Fay
& K.M. Louise (Eds.), Investigations in clinical phonetics and linguistics (pp. 201–212). Mahwah, NJ:
Erlbaum.
Janina Kahn-Horwitz, PhD, is a lecturer at Oranim and Gordon Colleges of Education. Herresearch interests include individual differences in language learning and their impact on English as
ACQUIRING THE COMPLEX ENGLISH ORTHOGRAPHY 155
Copyright r 2011 UKLA
a Foreign Language (EL2) acquisition, EL2 reading and spelling development, development of EL2assessment measures and intervention studies in EL2.
Mila Schwartz, PhD, is a senior lecturer at Oranim College of Education. Her research interestsinclude studying of biliteracy; early bilingual education; language and cognitive development ofearly sequential bilinguals; family language policy; and immigrant teachers’ pedagogicaldevelopment.
David L. Share gained his PhD in Psychology at Deakin University in Australia. He is currentlyfull professor at the Department of Learning Disabilities, and Edmond J. Safra Brain ResearchCenter for the Study of Learning Disabilities at the Faculty of Education, University of Haifa,Israel. His main areas of research are reading acquisition and reading disabilities with a specialemphasis on the role of writing systems in literacy learning.
Received 6 October 2010; revised version received 19 November 2010.
Address for correspondence: Janina Kahn-Horwitz, Manof 46, D.N. Misgav, 20184,
Israel. E-mail: [email protected]
156 KAHN-HORWITZ, SCHWARTZ and SHARE
Copyright r 2011 UKLA