Effects of Diglossia on Classical Arabic: Language Developments in Bilingual Learners

The purpose of this study is to investigate the effects of Arabic diglossia on the development of classical Arabic language acquisition amongst bilingual learners in a private school in Lebanon. The study compares the Arabic language (L1) performance to that of English as a Foreign Language (EFL) in accordance with Bialystok’s model. One hundred-forty participants; (n=140), ranged from five to eleven years of age, with a mean of eight years were sampled. A two-stage random sampling technique was applied, while ensuring that students with contrastive academic achievement were included within the study. The participants sampled for this correlational research were learners enrolled at a private middle school. Participants were given two standardized measures so as to establish the trend of development in oral skills for both classical and colloquial Arabic, determine the grade level at which convergence occurs between the two forms of acquisition and finally compare the degree of classical Arabic (L1) acquisition with respect to their EFL acquisition. The findings indicated interrelatedness between Arabic diglossia and the late oral development of classical Arabic, whereas participants showed a higher degree of comfort with English than with their mother language. Recommendations for future directions and research are also given.


INTRODUCTION
Low educational achievement and high rates of illiteracy across Arab countries can be attributed to inherent complexities within the Arabic language (Labanieh, 2019;Maamouri, 1998;Abu-Rabia, 2000;Kaye, 2001;Maamouri, 2007). In addition, further studies on bilingualism and language acquisition should be carried out since "children who speak a language with a similar orthography to a second language may have some advantage when learning to read that second language in comparison to children whose first and second languages and orthographies are unrelated" ( Winskel, 2020,p.2).The Arabic language is characterized by diglossia, an occurrence whereby two varieties of the same language exist (Abu-Rabia, 2000;Fedda & Oweini, 2012). Fusha, or classical Arabic, is a universal standard utilized in both oral and written expressions of formal communication, while ammiya is a derivation utilized in colloquial speech (Dakwar-Khamis, 2005). Ferguson (1959) describes the comparably stable language state of diglossia in Arabic, Swiss German, modern Greek, and Haitian Creole as one where two forms are distinguishable. Further definition of this state can be found in the work of Labanieh, 2019, Maamouri (1998, as well as that of Fedda and Oweini (2012), where each of the two forms of Arabic, classical and colloquial, possesses a high (H) and low (L) form/value. Classical Arabic represents the high form of Arabic, and is taught in schools for use in formal communication; while the colloquial low form of the language is reserved for informal speech, acquired in the home through daily social interaction, and dependent on local dialect (Ferguson, 1959;Abu-Rabia, 2000).
Beyond the function that each form of the Arabic language serves and the respective means of acquisition, fusha and ammiya are distinguishable by their grammatical and phonological divergences (Kaye, 2001;Labanieh, 2019;Maamouri, 2007). Ferguson (1959) points to the increased complexity of classical Arabic's grammar when compared to the colloquial low form, while Abu-Rabia (2000) notes the respectively richer syntax and lexicon of the former. In addition to encompassing all the phonological sounds of ammiya, classical Arabic includes exclusive phonemes that Kaye (2001) attributes to historical changes in the language. The most apparent point of differentiation between the two forms is the exclusion of ammiya in written expression, resulting in classical Arabic being the only institutionalized form in schools and universities (Zughoul, 1980). It is within the context of the educational system that the diglossic quality of the Arabic language precipitates a primary sociolinguistic and academic problem facing Arab countries. A wealth of literature exists demonstrating learners difficulties in the acquisition of basic academic skills; mainly due to Arab learners mother tongue differing from the more complex fusha form at the earlier stages of their learning (Ayari, 1996;Abu-Rabia, 2000;Saiegh-Haddad, 2003;Dakwar-Khamis, 2005;Maamouri, 2007). Critically, their two to fivethousand-word vocabulary forms the basis for their acquisition of reading and writing skills. While the development of these abilities in non-diglossic languages like English is a seamless one. Arab students' oral language experience does not provide a comparable platform of transition (Perfetti, 2007). Perfetti (2007) has also highlighted the extent to which fusha diverges from ammiya in form and structure to the point that any similarities between the variations do not lend themselves to neither fluid word recognition nor to easy language comprehension. The disparaging chasm between the high and low forms of the language, along with which form is the more habitual in daily usage, lies at the heart of a call for reforming educational policy in the affected region (United Nations Development Programme (UNDP), 2011). Decades of research yielded a consensus on the importance of education in the development of societies at the socio-economic level; furthermore, research indicated that pedagogy has been essential as early as the primary stage of schooling (Maamouri, 1998). The UNDP has stated in its Arab Knowledge Report 2010-2011 that the Arab region should formulate education reform plans and policies if they want to invest in future generations and contribute towards the knowledge society of the 21 st century.
By way of further complicating the challenges to the education system in Arabicspeaking countries, school programs typically promote bilingualism (Fedda & Oweini, 2012). Lebanon is a prime example of bilingual education, whereby students acquire either English or French as a compulsory second language at the early stage of learning. Fedda and Oweini (2012) have cited bilingual education as an additional complication for the acquisition of vocabulary words in diglossic Arabic. Their work has highlighted the deficiency attributed to reduced focus on the Arabic language in the Lebanese curriculum, which: 1) competes for time with the designated second language; 2) suffers from less attractive learning resources and tools than its Western equivalent and; 3) is given secondary billing by parents in the home environment in favor of the secondary language. The prevalence of bilingual education in Arab countries like Lebanon presents an opportunity to expand on research into diglossia within the context of a bilingual education system. While ample research exists for Arabic diglossia, less attention has been focused on the relevance of bilingual education as an affecting factor. More significantly, there is no existent literature examining the effect of diglossia on classical Arabic language and its interrelatedness from the acquisition of English as a foreign language. Furthermore, there are no studies in Lebanon comparing the Arabic language (L1) performance of Lebanese bilingual learners to that of English as a Foreign Language (EFL) in accordance with Bialystok's model which is an international backbone model.
The primary purpose of this study has been to investigate the effects of diglossia on the development of oral classical Arabic language in bilingual learners attending the primary and elementary cycles of a private school in rural Lebanon. As such, in order to investigate the effects of diglossia on the development of oral Classical Arabic, the present study has sought to: (1) explore oral language development, with respect to both classical and colloquial Arabic, for bilingual Lebanese students at the early stages of learning; (2) determine the grade level of education at which point there might be a convergence of acquisition between classical and colloquial Arabic; and (3) account for a comparable or contrasting degree of oral language acquisition between classical Arabic and a foreign language.
With the work of Fedda and Oweini (2012) providing the only comparable study in existent literature, this study is significant to contribute to the existing plethora of literature and to extend the previous findings on diglossia with respect to bilingual students. The study also highlights the effect of diglossia on classical Arabic Language causing poor acquisition of classical Arabic though Arabic is the mother tongue of the participants in contrast to the noticeable acquisition of English as a foreign language. Furthermore, it is the only study that intended to compare the Arabic language (L1) performance of Lebanese bilingual learners to that of English as a Foreign Language (EFL) in accordance with Bialystok's model. The study is unique being the first to employ the standardized measures in order to determine the trend of development in oral skills for both classical and colloquial Arabic and to compare the degree of classical Arabic (L1) acquisition with respect to their EFL acquisition in an attempt to examine if there is a possibility for interrelatedness between Arabic diglossia and the late oral development of classical Arabic when learners are at a higher degree of comfort with English than with their mother language.

LITERATURE REVIEW THEORETICAL FRAMEWORK
The present study adopts a sociolinguistic approach to "describe a situation where in a given society there is more than one language variety in use" (Saiegh-Haddad, 2012, p.2). Another theoretical framework for this study is the theory of diglossia hypothesized by Ferguson (1959) who proposed a classification of diglossic features to encompass literary heritage, acquisition, and standardization. Ferguson's theory of diglossia is determined by the diglossic context characterized by the co-existence of a high and a low linguistic variety (Ferguson,1959). The study was further framed by the threshold theory and Bialystok's model which highlighted the existence of levels of linguistic proficiency bilingual children should attain in order to be characterized by bilingual superiority that is interrelated with the high degree of bilingualism (Ricciardelli, 1992).

LEARNING DIFFICULTIES ARISING FROM ARABIC DIGLOSSIA
The Arabic language is characterized by its diglossic quality, whereby the colloquial derivation is distinctively different from the co-existent formal expression (Ferguson, 1959). While students are required to learn how to read and write in the formal/high form of the language, their daily social interactions only serve to promote the low colloquial form irrespective of their education (Farran, Bingham, & Mathews, 2010;Turkistani, 2019 ). Ibrahim (2009) goes as far as stating that the disparity between the two forms of Arabic expression is akin to acquiring two separate languages. Arabic is a Semitic language which derives its alphabetic system from Abjad script. In its pure form, Abjad "does not rely on vowels and uses written scripts adapted from Sumerian Cuneiformlogographic writing that preceded alphabetic writing" (Farran, et al., 2010, p. 8). In its current evolved state, Arabic is composed of 28 letters that include three long vowels, as well as short vowels denoted by diacritics. In certain Arabic education programs, the emphasis is on teaching students to first read and write the vowelbased form of Arabic. Alternatively, this method of teaching is utilized in conjunction with learning the vowel-free form of the languagedependent on Arabic sight words (Oweini & Hazoury, 2010).
Existent research into Arabic diglossia (Abu-Rabia, 2000;Saiegh-Haddad, 2003;Turkistani, 2019) has uncovered a link between the nature of the language and low educational achievements with respect to Arabic literacy. Literature delves into the negative attitude of Arab children with respect to their primary language, as well as diglossia's impact on grammatical judgment, basic reading proficiency and comprehension, as well as working memory.
A study by Abu-Rabia (2000) has established the interrelationship between Arabic diglossia and reading comprehension. His study sampled first and second graders, whereby one group had prior exposure to colloquial Arabic at a preschool level, with a second group only exposed to colloquial Arabic during the same period. By comparing the reading comprehension performance of each group, Abu-Rabia's findings demonstrated that prior exposure to classical Arabic had a positive impact on students' performance levels with respect to reading comprehension. Abu-Rabia's study underlined the importance of instruction in classical Arabic at even the earliest stages of learning and language development. The improvement of reading comprehension does not, however, extend to all mechanisms of Arabic language acquisition at this stage.
Two studies by Saiegh-Haddad (2003) examined the phonological aspect of Arabic diglossia and reported difficulties facing students during the early stages of acquisition. In the first study, Saiegh-Haddad (2003a) found that phonological chasm between the high and low forms of the Arabic language had a negative effect on oral language fluency. In the second study, Haddad (2003b) extended these findings with an investigation into the ability of preschoolers and first graders in decoding phonemes. By examining phonemic awareness and pseudo-word decoding scores, Saiegh-Haddad found that first graders displayed difficulty in isolating typical phonological structures, irrespective of whether they benefited from preschool exposure to formal language instruction or not. This finding opened the door for further investigation into whether difficulties in decoding phonemes were shown at all grade levels or whether the difficulties decreased at a higher level of learning. However, further investigation by Saeigh-Haddad, Levin, Hende, and Ziv (2011) has demonstrated that preschool learners' performance in recognizing classical Arabic phonemes was significantly less than their performance on colloquial test items.
The preceding work by Abu-Rabia (2000) and Saeigh-Haddad et al. (2003a, 2003b has aligned with studies on oral language and comprehension that were not specific to the Arabic language. Both Catts (1997) and Snowling (2005) have demonstrated that difficulties in oral language were predictors of future deficiencies in reading comprehension, resulting from limited phonological awareness, difficulties with word retrieval, poor verbal retention, and limited speech production or perception. Additionally, mild difficulties in oral language have resulted in mild difficulties in reading while major oral language deficiency eISSN: 2550-2131 ISSN: 1675-8021 192 severely impeded reading comprehension (Kaiser, Roberts, & McLeod, 2011). The implication for the Arabic language is that, in order for students to develop a high rate of literacy and strengthen their reading comprehension, it is vital that prior attention is given to the high form of oral language since the colloquial form is an inadequate substitute and future reading is dependent on formal expression, phonemes and vocabulary.

BILNUALISM: AN IMBALANCED DYNAMIC
Not confined to internal challenges presented to the Arabic language due to its diglossic quality, the Arab region is also faced with the external encroachment of English and French (Amara, 2010). Several research studies have concluded that there is a gradual loss of the native Arab language among the younger generations, itself a byproduct of previous colonization/protectorate periods as well current globalization (Bassiouney, 2009;Maamouri, 1998;UNDP, 2011). The twentieth century saw an increasing institutionalization of English and French in Arabic schools and universities (Al-Khatib, 2006). This can be directly attributed to periods of colonization by Britain and France, especially in the Middle East following the first World War (Owens, 2001). Depending on the division of territories in the region, English or French was instilled as a second language. Lebanon is a prime example of ingrained bilingualism, with students exposed to either language throughout their 14 years of schooling. The impact of this exposure is compounded by the prevalence of English as a universal language of communication (Al-Issa & Dahan, 2011;Turkistani, 2019). Consequently, the spread of English results in the suppression of an indigenous language, leading to its deterioration at both the social and educational level. Chin and Wigglesworth (2007) have stated that when the first language of any community is not frequently used and transmitted to the coming generation, it is endangered and can be lost. While the situation in the Arab world is not quite this apocalyptic, linguists and scholars are apprehensive because English is favored over Arabic in different channels of communication, ranging from academic environments to advertisements (Mejdell, 2008).
There are extolled positives to a bilingual education whose benefit is conditional upon the satisfaction of a certain dynamic. Nero (2005) has stated that high literacy levels in the first language (L1) build the second language's (L2) cognitive academic language proficiency (CALP). It can be concluded that the continuous development of L1 leads to an improved learning for L2 (Hornberger, 2003). Moreover, Cummins (2000) has indicated that the continuous transfer of academic skills between L1 and L2 will only happen if students are evenly exposed to L1 and L2 in reading and writing. Thus, the transfer of the latter skills is facilitated when the common underlying proficiency (CUP) of L1 and L2 exists. In either instance, the development of the second language is dependent on the strong foundation of the first. In the case of the Arabic language, the literature not only suggests that there are challenges resulting from its inherent diglossia, but that the second language is increasingly replacing Arabic as the primary means of communication.

METHOD RESEARCH DESIGN
The correlational research design was used for the purposes of this study. Fraenkel and Wallen (2010) referred to this design as associational research, due to the researchers' intent to study relationships between variables without manipulating them. While correlational research does not establish causality, it does depend on two or more quantitative variables in order to investigate the degree of association between them (Johnson & Christensen, 2011). In this eISSN: 2550-2131 ISSN: 1675-8021 193 study, there is a presumed association between diglossia and oral language development, based on the present literature. In applying the correlational research, the study investigated the degree of association between: (1) learners' scores on the Story Recall subtest in both classical and colloquial Arabic, and (2) oral language scores in English and classical Arabic to identify the degree of preference between first and second languages.

SAMPLING PROCEDURE
A sample of 140 students from a private rural school in Aley, Mount Lebanon, were selected at random for participation in this study; with an equal number of respondents (n=20) selected within each class from kindergarten through grade 6. The sample size was set in order to comply with the requisites prescribed by Ary, Jacobs, Razavieh, and Sorensen (2010), whereby a minimum sample size of 50 is required in order to establish valid association between variables in correlational research.

PARTICIPANTS
One hundred forty participants (n=140) ranged from five to eleven years of age, with a mean of eight years were sampled. A two-stage random sampling technique was applied, while ensuring that students with contrastive academic achievement were included within the study. Furthermore, the principal point of homogeneity amongst the participants was that all students had been enrolled in this private school as of the kindergarten grade level; with Arabic being their primary language (L1), followed by English as their foreign language (EFL).
According to Gay et al. (2012), random sampling methods have been techniques to obtain a sample representing the population of interest. The two-stage random sampling technique was commonly valuable and more representative because it combined cluster random sampling and individual random sampling (Ary et al., 2012). Moreover, this technique allowed the researchers to spend less time in testing the sample than in testing all the chosen grades. The researchers along with the teachers, who were familiar with the respondents, selected ten participants at random from each grade level (kindergarten grades 1, 2, 3,4,5 and 6), followed by another ten students at random resulting in a total of one hundred forty students selected as respondents. Upon entering each classroom, the students were cordially invited to participate in a project that would improve the school's teaching of the Arabic language, at which point the would-be respondents were able to opt into the project voluntarily. Reaction to the invitation was positive throughout the grade levels, allowing for the designated number of students from each class to be selected; (n=20). Students' names were clustered in groups of two as part of the first stage of random sampling. For the second stage of random sampling, one student from each cluster was selected at random, with the names of participants posted on a whiteboard with a schedule of assigned days for testing.

EHICS
Prior to testing the participants, the researchers obtained approval from the school's administration to approach the sample group. All respondents were asked individually for approval to administer the testing process. Students' names were not recorded on score sheets, with each assigned a number dependent on their grade level. The privacy rights of students were thus respected in order to protect the confidentiality of the scores.

INSTRUMENTS AND PROCEDURES
The following two instruments along with their respective procedures-both in the form of standardized tests-were utilized in the collection of data:

STORY RECALL SUBTEST
Appropriated from the Woodcock Johnson-III Tests of Achievement (WJ III ACH), the Story Recall Subtest consisted of ten stories arranged in increasing order of complexity. While the WJ III ACH consisted of twenty two oral language tests, the Story Recall Subtest was independently considered as "a single measure of a broad cognitive ability and a more specific narrow ability as defined by Catell-Horn-Carroll theory" (Schrank, 2006, p. 1). This instrument was capable of identifying the oral language ability of participants, in addition to their language development and significant memory. The Subtest was administered by the teachers who first read out the stories, before requiring the participants to demonstrate recall by way of recoding the construction of mental representations (Woodcock, Schrank, Mather & McGrew,2007). The Subtest was appropriated by reciting stories in both classical and colloquial Arabic translations, with each two paired together in a grouping to serve as a starting point. Before testing, the correct starting items in the participant's score sheet were marked as per the specific table found in the WJ III battery. The starting item for a participant in grade four, for example, was story five. Elements of the story were separated by slashes (/) on the Test Record. For each element of the story correctly identified by the respondent, a check mark was placed for the segment. For words marked in bold letters, an exact recall of the element was necessary for a correct marking, while the remainder of the elements would be counted as correct if paraphrased accurately. Students were not penalized for errors resulting from articulation miscues, while synonyms of bold words were marked down as correct. In order to account for the school's Arabic teaching program, scoring guidelines were tailored with the assistance of the school's Arabic language coordinator. This was important due to schools varying in the instruction of Arabic at lower grade levels, as per previously cited literature regarding the introduction of vowel-based or non-vowel-based Arabic (Oweini & Hazoury, 2010). In effect, this meant that the participants in this case were only penalized for errors resulting from incorrect diacritics (or short vowels) if they belonged to the upper elementary grade levels (Grades 4, 5, 6). The students were marked following their recollection of both stories within each grouping (classical and colloquial story pairing). In the instance where a student scored less than the designated number on the score sheet for the starting item, stories from the preceding group level were recited. In the event of a student scoring higher than the designated number, the Subtest moved on to the next degree of story complexity until the criteria for scoring were met or the tenth story was successfully recalled. The software scoring program of the Woodcock Johnson III Tests was then used to input the final score tally in order to get the standardized age equivalent scores necessary to study the correlation of variables in the research.

Bilingual Verbal Ability Tests (BVAT)
The BVAT (Picture Vocabulary test (PVT), Oral Vocabulary -Synonyms test (OV-ST), Oral Vocabulary -Antonyms Test (OVAT), and Verbal Analogies Tests (VAT) used comprised a series of examinations that measured the bilingual verbal proficiency and exclusive cognitive/academic language abilities that participants demonstrate between English and a second designated language (Muñoz-Sandoval et al., 1998). As the manual of BVAT states, each item in the non-English test was a translation of the equivalent item in the English test.
First, the PVT consisted of fifty-eight pictured objects used to measure the comprehension knowledge of the participant by way of identifying the objects according to order of familiarity (Muñoz-Sandoval et al., 1998). Participating Students were given two eISSN: 2550-2131 ISSN: 1675-8021 195 different pictures at a time and asked to place their finger on the object recited by the administering teacher. A correct identification of the designated object was considered a correct response.
Second, the OV-ST and OV-AT assessed the comprehension knowledge of participants by evaluating the ability to provide synonyms and antonyms for administered words. Upon providing a demonstration of what is expected from the student, the participants were then asked to provide synonyms for twenty printed vocabulary words and antonyms for twenty-four others. Responses denoted scores of either 1 for correct responses, or 0 for false answers.
Third, the VAT measured the extended evaluation of comprehension knowledge by also measuring reasoning and fluid intelligence (Muñoz-Sandoval et al., 1998). The test was carried out by evaluating the capacity of the participants to reveal hidden relationships among words and retrieve a suitable response. The participants were first given a point of reference with the recital of a statement, such as "a bird flies; a fish ...". The participants were instructed to complete the analogy by providing the adequate verb. While the words themselves were simple throughout, the complexity of the relationship increased in tandem with students providing correct responses. The starting point in terms of relationship complexity was determined according to the suggested guidelines for the student's grade level. Again, participants scored 1 for correct analogies, and 0 for false responses.
In order to determine whether classical Arabic or English represents the more comfortable language for students, the BVAT were administered to statistically identify degrees of significance between the following: 1) the mean raw scores of BVAT in classical Arabic and English for the lower elementary classes and upper elementary classes (following on the same grade level groupings as the Story Recall Subtest); 2) the mean raw scores of BVAT in classical Arabic and English between the lower and upper elementary classes; and 3) the mean standard scores of Verbal Bilingual Ability and English Oral Proficiency to identify whether or not the classical Arabic scores have gained from the additional English scores.
Validity and interpretation of the results of the administered tests relied on the teachers' and researchers' preparation and following the general procedures while implementing the tests that were found in the comprehensive manual of each test (Muñoz-Sandoval et al., 1998;Woodcock et al., 2007). Testing rooms were chosen to be accessible for all participants, with attention given to comfortable setting, adequate lighting and air circulation. Notices were placed on doors in order to avoid interruptions and sufficient time was allocated in order to complete each test at a natural pace. This afforded participants a reasonable time for responding to each item, and testing sessions were administered on different days to ensure that there was no conflict with exam periods or a build-up of pressure. Furthermore, the following modifications were made for two students with physical disabilities as permitted by the tests' respective manuals: 1) Participants could elect to type their responses, and 2) enlarged test pages were provided for students with visual impairments.

RESULTS
Values were reported in the form of mean ± SD for quantitative variables. The analysis of variance technique (ANOVA) was utilized to determine the statistical significance of differences among means of standard scores for quantitative variables between all grade levels. The SPSS Tukey post-hoc test was conducted in order to determine differentiation between grade levels, while the Paired-Samples T Test was applied to determine the statistical significance of differences between raw score means of two quantitative variables for students at the same grade level. Finally, the one-way MANOVA was applied in order to determine the statistical significance of differences existent within raw score means of two or more quantitative variables across all grade levels. If the p-value (i.e. the probability of an observed eISSN: 2550-2131 ISSN: 1675-8021 196 result arising by chance) is less than 0.05, then the difference between means was statistically significant.

STORY RECALL SUBTEST
The results of the Story Recall Subtest WJ III were analyzed to statistically identify the level of significance between the following: 1) the mean standard scores of the Story Recall Subtest in classical Arabic among all grade levels; 2) the mean standard scores of the Story Recall subtest in colloquial Arabic among all grade levels; 3) the mean standard scores of the Story Recall subtest in classical and colloquial Arabic between the lower elementary classes (Kindergarten through grade 3) and upper elementary classes (grades 4, 5, 6) and 4) the mean standard scores of Story Recall subtest in classical and colloquial Arabic within the same grade level. A breakdown of mean standard scores for this instrument is presented in the table below.

BILINGUAL VERBAL ABILITY TESTS (BVAT)
In order to determine whether classical Arabic or English represents the more comfortable language for students, the results of the BVAT were analyzed. A breakdown for the mean standard scores, as per examination within the BVAT, is presented in the tables below.

DISCUSSION
The findings of the present study align with those of Thomas-Sunesson, and Bialystok (2018) which indicated an additive view of bilingualism, where balanced proficiency in two languages; L1 and L2 could result in enhanced cognitive and linguistic functions irrespective of the socioeconomic background. Furthermore, with respect to the results of the Story Recall subtest, the findings of the present study align with those yielded by Turkistani (2019), Labanieh (2019), andFerguson (1959). The findings revealed that diglossia had an effect on the late development of oral language among the participants. By comparing the mean standard scores of classical and colloquial Arabic, the significance was established for both the lower and upper elementary school grades, but with a higher significance for those attending lower grade levels. The present study findings transpired that colloquial Arabic was more easily acquired than classical Arabic, but the difficulty of classical Arabic acquisition decreased as participants progressed through their schooling. These results supported the findings of Saeigh-Haddad et al. (2011), whereby colloquial Arabic phonemes were more readily recognized than classical ones. Diglossia's impairment of oral language fluency in Saeigh-Haddad (2003b) was equally confirmed. In reference to the colloquial Arabic scores of the Story Recall subtest, acquisition of informal language also improved as participants moved through grade levels, although the improvement was not as evident as the improvement in classical Arabic. This could be attributed to the fact that the conditions for acquiring colloquial language skills were less challenging due to informal speech not being as dependent on schooling as classical Arabic. In support of Fedda and Oweini (2012), the findings of this study showed that classical eISSN: 2550-2131 ISSN: 1675-8021 198 Arabic would ameliorate more significantly due to the school setting being the only environment where the high form of the language was practiced.
With respect to the BVAT findings, the raw scores of the English tests were significantly higher than those of the Arabic editions, indicating a clear preference of students towards English (L2) over their primary Arabic language (L1). This finding indicated a necessary reform plan for the acquisition of the Arabic language in order to counteract the encroachment of the secondary language in Arab countries. Moreover, the scores of the bilingual verbal ability (Arabic and English combined) were the same as the English language proficiency scores. This was possibly related, as identified in Cummins (2000) and Nero (2005), to the low level of acquisition of the Arabic language.
Studies have shown that the high literacy levels in L1 build L2's cognitive academic language proficiency (CALP). It can be concluded that the continuous development of L1 leads to an improved learning for L2 (Hornberger, 2003). Moreover, Cummins (2000) has indicated that the continuous transfer of academic skills between L1 and L2 will only happen if students are evenly exposed to L1 and L2 in reading and writing. The results of the BVAT have indicated that this balance is not present in the sample group, and that English is effectively supplanting Arabic as the primary L1; a language which already suffers from complications due to its diglossic nature. An expansion on this research could serve the purpose of providing a springboard for the development of new strategies to support Arabic teachers and to highlight the areas where policy reform can be made.

RECOMMENDATIONS AND IMPLICATIONS
The findings call for pressing reform plans regarding Arabic language teaching and learning in order to promote better classical oral language acquisition. Recommendations include instructional strategies in the classroom for the purpose of improving classical Arabic proficiency.
It is suggested that teachers, and more importantly curriculum designers emphasize the acquisition of classical Arabic in its oral form in order to improve the performance of learners throughout their stages of schooling. This is as important to the immediate acquisition of formal Arabic in its written form as it is for the long-term survival and transmission of this diglossic language in the face of encroachment by a second language. The presented literature has indicated that both the teacher's approach and parentally adopted attitudes at home have had a profound effect in motivating students to maintain the robustness of their first language. At the level of school education, research-based learning activities can be enacted in order to enhance acquisition and overcome pedagogical challenges. Fedda and Oweini,(2012) have previously suggested a reform to the instruction of the Arabic language in order to maintain cultural identity and overcome the youth's "mixed feelings about their own language" (p. 358). This has been accomplished by introducing students to enjoyable activities that improve oral language skills in tandem with written script in textbooks. Such activities include interactive Q&A sessions; whereby teachers directly engage their students with questions about their daily comings and goings, with an eye on emphasizing the correct usage of formal Arabic over the colloquial. By affording formal Arabic an oral, and thus social transmission, educators are thus less dependent on textbooks that are perceived as boring while strengthening the established connection between strong oral skills and future written proficiency. A related approach, and one that enhances picture vocabulary, involves having students either describe their own artwork in formal oral depictions, or having teachers organize games of discovery, whereby students need to extract clues from their classmates and detect the nature of mystery objects; all with the use of formal oral language skills.
Given that the development of oral skills goes hand in hand with the improvement of social skills, tactics such as the recreation of enjoyable real-life situations represent a means to achieve higher levels of interest and engagement in the classroom. Role playing activities, as Schrank and Woodcock (2001) suggest in the companion program of the WJIII, further enhance oral skills by way of social communication. The enforcement of formal language in social settings can be further enhanced by promoting socially oriented approaches such as the STORE strategy (Setting, Trouble, Order of Events, Resolution, and End). Further researchbased strategies that are dependent on social enhancement can be promoted by way of: 1) the use of image-schema-based instruction (Morimoto and Loewen, 2007); 2) task-basedlanguage-teaching with emphasis on form components (De La Fuente, 2006) and; 3) word manipulation (Fang and Xi-ya, 2009).
While the highest expectations are set upon teachers in the school setting, parents are equally encouraged to appropriate strategies to enhance their children's acquisition of Arabic in order to diminish the divide between the colloquial language used in the household and the more stately form of Arabic that is otherwise only promoted in the classroom. Donitsa-Schmidt et al. (2004) find that parental attitudes towards the Arabic language, vis-à-vis a second language such as English, is a strong predictor of how their children approach Arabic within the classroom. Parents are encouraged to recite stories to their children in the primary Arabic language, or even watch cartoons in their native tongue, before engaging them with conversations about those very stories/TV programs in formal oral language. While this may seem counterintuitive to parents, given that colloquial Arabic is the mainstay of social communication, such efforts to engage in formal language would not only help to avoid the caveats of an encroaching second languages, but face the difficulties of a diglossic Arabic language distressed by its own internal challenges.

LIMITATIONS OF THE STUDY
Two related limitations exist for this study. First, the participants were selected from a single private school, a scenario that could have enabled selection bias. Second, the results of the presented research sample cannot be generalized to the entire population. While the student curriculum in Lebanon is set by the state for public and private schools, differences in teacher training, expertise and attitudes between each educational institution cannot be accounted for without a wider sampling scope.

CONCLUSION
The purpose of this study was to investigate the effects of Arabic diglossia on the development of classical Arabic language acquisition among bilingual learners in a private school in Lebanon. The study found that: 1) diglossia had a negative impact on oral language development for bilingual Lebanese students; 2) the negative effect of diglossia regressed over time due to increased exposure; and 3) there was a convergence of acquired classical and colloquial Arabic skills as of the fourth grade of schooling.
The challenges that arose from the Arabic diglossic nature are well-investigated in the literature, whereas the compounding aggravation of a more attractive second language (English) had yet to be adequately evaluated. While the results of the Story Recall subtest demonstrated the relative difficulty of acquiring classical formal language in comparison to the colloquial low form, the BVAT revealed that the inherent difficulties of Arabic diglossia were further threatened by the more comfortable acquisition of English among students. Furthermore, the acquisition of the English language did not result in better bilingual ability for students; with the significant differences in acquisition leading to a deterioration of the