Arabizi Among Kuwaiti Youths : Reshaping the Standard Arabic Orthography

Arabizi is a trendy language phenomenon utilized by young Arabs to communicate across various social platforms. Young Kuwaitis seem to not be any exception in that regard. This paper aims mainly at investigating the linguistic features of Arabizi as produced by the young generation in Kuwait, and the reasons for which the practice has been persistent amongst the youth community. The main corpus data was collected from spontaneous WhatsApp chats of 35 young Kuwaiti respondents who provided 400 of their e-messages to be linguistically analyzed by the researcher. A digital questionnaire was also implemented to illicit respondents’ responses on the reasons for which young Kuwaitis use Arabizi in their e-messages. Due to the heterogeneity of the spontaneous corpus, supplemental data was provided from a story writing that was sent to the respondents to be re-written in the style they choose when they normally chat on WhatsApp. From a linguistic point of view, the study reveals a number of tendencies that place Arabizi as a unique method of communication with a profile that employs both transcription and transliteration in the way it represents its consonants vs. vowels, Kuwaiti dialectical phoneme shifts and the wide use of extralinguistic features. Intensive code-switching and mixing has also been displayed. The present study also signifies a number of sociolinguistic reasons for which Kuwaiti users of Arabizi employ the script in their e-communication across social platforms.


Preliminaries
"We live in a digital world.Just short of being pixelated or virtual ourselves, the things around us continue to get smarter, faster, more connected, and increasingly digital".The blogger Jean Turgeon (2017) has pictured our life in the new digital era, where technology has reshaped every aspect of the human's daily life.With the emergence of various social media platforms, people no longer have to wait for days and hours to get a response from others.
According to the newsletter of the University of Iowa, IowaNow (2013) states young generations have been labelled "digital natives" because of their heavy reliance on the use of internet and digital technology.Young people tend to, therefore, spend less time interacting with their peers face-to-face and increasingly more time screen-to screen.Social media platforms have invaded almost every household worldwide, and the younger generations seem to be affected the most.Virtual communication has mostly replaced the traditional communicative styles, and with the fast pace of the new forms, innovative writing styles emerged in many languages and cultures, if not all.
Replications of such reshaped environment seem to have its huge impact on languages in both spoken and written versions.Text messaging and online communication has increasingly become the fundamental part of social discourse as the new technologies integrated into smartphones enable its users to send off chunks of written information by just a few keyboard taps.Gordon (2011) relates some serious changes of the orthography features of written languages to the volume, brevity and popularity of the current digital mode of communication.Referred to as computer-mediated communication (CMC), the mode is defined as any communicative exchange through a computer device (Mcquails, 2010).Gordon however, prefers to use the term electronically-mediated communication (EMC) to acknowledge a line of mobile devices (tablets, smartphones and the like) on which this mode of communication is consumed.
According to Crystal (2001), a newly established academic area in the field of linguistics has emerged, namely "Internet Linguistics", due to the rise and development of internet communication.The field focuses on a number of issues including the style of different formats in the written and spoken language, the likelihood of a language change, alongside other sociolinguistic features and consequences of what Crystal refers to as "Netspeak".While a significant amount of research revolves around English, only a few studies have focused on online Arabic (Warschauer et al. 2002;Palfreyman & Khalil, 2003;Al-Tamimi & Gorgis, 2007;Farrag, 2012).

Significance of the Study
Messaging in the Arab World (more specifically amongst Kuwaitis as being the population for this study) excessively use Latin symbols to represent Arabic words.Several terms were used in the literature to describe this phenomenon.Terms such as Arabizi, Arabish, hybrid language, Romanized Arabic, Latinized Arabic, and Arabic Chat Alphabet have been used to represent the phenomenon.Nevertheless, in Kuwait, the term that most frequently used to describe a hybrid form of Kuwaiti only and/ or Kuwaiti-English code switching (KE CS) where Latin characters are used to replace Arabic orthography symbols in e-communication is well-known by the young generation as Englizee Mu3arrab (Arabized English).In fact, in itself, such term seems to indicate that Kuwaitis perceive the innovative writing system to be mainly English that has been converted to Arabic rather than vice versa.However, due to the misleading nature of the term, the researcher prefers to use the term Kuwaiti Arabizi as it best reflects a variety code of Kuwaiti Arabic (KA) and/ or Kuwaiti-English code switching (KE CS) written in English characters within various e-communication platforms for a number of reasons that will be explored further in the current study.

Review of Relevant Literature
A division among researchers in the field has been instigated on whether CMC/ EMC reflects a written or a spoken language variety.Baron (2000) views the variety as being partially speech and partially written.Crystal (2001) however, rejects the idea of considering EMC as a fusion of both writing and speaking, as it displays features of both.He therefore suggests EMC to represent a third medium.Gordon (2011) poses a dilemma of using EMC when one's native language lacks a standardized writing system.Gordon further describes EMC as remaining largely unexplored despite the heterogeneity of the spoken Arabic dialects used among Arabs.As a result, Gordon views the new ways of encoding previously unwritten dialects in such a highly digital world as representing a transition from a strictly spoken language to a regularized written code, composed largely in the Roman alphabet, particularly when one considers the significance and popularity of online social media in the Middle East (Gordon, 2011).

Arabizi: Linguistic Features
The techniques utilized in text messaging differs greatly from ordinary writing.Kul (2007) explores the nature of letter deletion in text messages in what he referred to as "textese", with the general prediction being that text messages will be decoded phonemically (via mental reading), as governed by the phonological principles of the semiotic "figure and ground" principle, predicting that figures to be foregrounded, grounds to be further backgrounded, with the phonological convention of consonants being figures and vowels being grounds (Dressler, 1996).Another phonemic principle investigated in Kul's study is the "rich-get-richer" principle, arguing that figures that appear in strong positions (such as consonants appearing initially, or at the initial position of stressed syllables) are preserved or strengthened by the process of stopping, whereas grounds appearing in weak positions are weakened; vowels, for example, in unstressed position tend to be reduced to schwa or even a zero; consonants in word final positions are deleted (Donegan, 1985).Kul's findings suggest that while the semiotic figure-and-ground principle seems to be in force, deletions in "textese" fail to observe Donegan's rich-gets-richer principle.In a similar vein, linguists list a number of linguistic features of text messages including shortening, contractions, clipping, acronyms, letter-number homophones, phonetic respellings, unconventional capitalization, and symbol replacing word (Sgerstad, 2002;Thurlow, 2003;Lupez Rua, 2005).Such downgraded linguistic properties highlight, as claimed by Bralczyk (2004) andSutherland (2002), a lesser than writing style that establishes a deviated mode of communication.Kul (2007) calls such claims depreciation of the productivity of texting, and an underestimation of texting impact on people's everyday communication.Crystal (2008) stresses the predominance of English orthographic rules, due to its global widespread, even when EMC is composed in other languages, particularly in the ones where Roman alphabet is utilized.In this regard, Crystal claims that English texting abbreviations such as lol (laughing out loud) and brb (be right back) have been widely spotted in EMC corpus of non-native English speakers.Crystal further illustrates an extensive use of numerals to replace the spelling of certain syllables in EMC of languages like Italian, German, Spanish and French.The numeral 2, dos, for example replaces the syllable in the word saludos (greetings), just as a similar approach has been employed by English EMC users for the words for (replaced by 4) and to (2). Gordon (2011) claims that replacing syllables with numerals tends to be more extensively used in EMC of languages with syllabic orthographies such as Japanese and Chinese.
Users of Greek EMC (Greeklish), on the other hand, tend to employ a mixture of Arabic numerals and Roman characters to replace Greek characters based on their graphical similarity (similar form), and phonetic similarity (similar sound) (Tseliga, 2004).With that said, the Greek graph η is replaced by the Roman alphabet h based on graphical similarity between the two symbols, I based on phonetic similarity, and 8 to replace θ based on visual similarity.Based on that pattern, the word Athena (Aθηva in Greek orthography) could be written in EMC as A8hva (2004).
In the light of the emergence of the phenomenon and its excessively used language variety across e-communications, Latinized Arabic has recently started to attract the attention and the efforts of linguists.The term "ASCII-ized Arabic" (AA) to refer to the "American Standard Code for Information Interchange" is presented to describe symbols that represent Arabic script in e-communication.AbuSa'aleek (2014) describes AA as a character encoding scheme originally based on the English alphabet which encodes 128 specified characters-the numbers 0-9, the letters A-Z, alongside few basic punctuation marks.
Driven by the urgency of establishing a transliteration scheme of Arabic names within the English speaking and reading communities (particularly for official documentary businesses), Gorgis (2010) lists a number of deficiencies in the current proposed transliteration schemes for Arabic (specifically Arabic names), a few of which are: 1.The difficulty of finding a one-to-one correspondence between Arabic and English spelling-pronunciation norms; 2. The inability to normalize regional variations in the pronunciation of words across the Arab world; and 3. that an Arabic name is not transliterated randomly, but rather determined by a cognitive process that conforms to the English phonology system.In this regard, Gorgis proposes what he refers to as a cognitive-pragmatic alternative scheme which requires firstly the formation of a committee that includes a number of Arab linguists and computer scientists interested in the field, as well as English informants from the main three English speaking countries (Australia, Briton, USA); and secondly a sufficient budget to work on the project.Haggan (2007Haggan ( , 2010) ) claims that up to date, studies on the prevalence of the Romanization of Arabic script in e-communication, and its inter-regional and inter-media variations have been missing from the field.In an attempt to redefine terms used to represent the linguistic variety, Gorgis makes a distinction between transliteration, Romanization, Latinization and Arabicization.In this vein, Gorgis explains that while transliteration refers to the process of representing native words of language X in the script of language Y, Romanization refers to the process of representing native words of language X in Roman orthography only (English, French or German).Latinization however, describes the process of, not only using the script, but also using the language and its vocabulary.Arabicization, on the other hand, involves transliterating foreign writing systems into Arabic orthography.Gordon (2011) sheds light on the linguistic features of what he referred to as the Arabic Electronically-Mediated Communication (EMC), an orthographic system devised by Arabic speakers of different dialects across the Arab World to facilitate Arab users of electronic platforms in making the required transition from a spoken language only into a regularized written code through the aid of the Roman alphabet.Utilizing the skeleton of EMC features of novel phrases and abbreviations, vowel deletion, typographic representation of extralinguistic features such as facial expressions and prosodic features, Gordon investigated Arabic EMC, while considering the particular challenges and existing resources to Arabic speaking users of electronic media.Such extralinguistic representations seem to be at the core of communication in the current research in particular as Kuwaitis rely heavily on tone and intonation to express variable meanings.A word like "zain" in Kuwaiti for instance may express a diverse set of responses including an affirmation (OK), threat (you'll see!), indifference (so what!), empathy (oh my god what happened next?), astonishment (really!), assurance (with pleasure), uncertainty (let me think about it), and enthusiasm (hopefully).Bearing in mind the lack of face-to-face interaction across e-communication, such diverse response would definitely be misleading.Unlike many researchers on the topic, Gordon views the innovative script to have great implications on the literacy in the Arab World, as it broadens the possibilities of what Arabic text can be.Sebba (2009) defines Romanization as an orthography system that maps the graphemes, phonemes, or syllables of one language into the graphemes of Latin alphabet.The practice however, tends to exhibit high variations amongst different source languages.While Romanized Greek, for example, presents a phonologically/ orthographically-based approach (Androutspoulos, 2006), Romanized Arabic seems to mainly rely on phoneme-grapheme mapping, a system that is described by a number of researchers of the phenomenon in the Arab World as being closer to transcription than transliteration (Palfreyman & Khalil, 2003;Al-Badrashiny et al, 2014).When phoneme-grapheme mapping is unmanageable, users of the variety tend to rely on the orthographic similarity of the characters in the native language script.Accordingly, numerals such as 2 for ‫,ء‬ 3 for ‫,ع‬ 7 for ‫ح‬ are utilized in the script.
Both code-switching and code-mixing between English and the spoken Arabic dialects have been reported in Arabizi.AbuSa'aleek (2014), for instance, reports a frequent recurrence of code-switching in the structure and composition of Arabizi.His data reveal a systematic switch between English and Saudi Arabic at a discourse level.Within the same data, the author also reports a regular manifestation of code-mixing between English and Saudi at a clause/ word level.Along the same vein, Al-Khatib and Sabbah (2008) list factors that motivate Arabizi users to switch and/ or mix between English and Jordanian among Jordanian university students as ease and swiftness of writing, less space consuming, and most importantly, communicatively function governed.In this regard, the authors report a switch to Jordanian to serve cultural/ religious functions and when quoting someone who originally used Arabic.Switching from Arabic to English, on the other hand, takes place to serve the function of prestige, academia and taboo related topics.

Arabizi: Reasons for Romanization
Bou Tanios (2016) notes that the process of Romanization should not be viewed as the result of technological advances.She further explains, although technological restrictions of the computer encoding systems in Latin script may have initiated the phenomenon of Romanization, they can never explain the script's persistence during the time when technology encoded the main World Wide writing systems.Investigating reasons for the persistence and the progressive widespread of Romanized script, researchers on the topic report "going with the flow" (Essawi, 2011); reconstructing a community's identity in a way that establishes an affiliation with Latin alphabet user nations (Sebba et al, 2012); indicating high educational status and prestige (Palfreman & Khalil, 2003); or reflecting the modernity of Latin-alphabet Western World users (Androusopoulos, 2009).
In a multilingual context, where EMC users have the knowledge of multiple writing systems, a switch between different alphabets is more complicated on EMC than on paper.(Cassany, 2014).The process of shifting between one keyboard and another, with more familiarity with a particular one, is described as time-consuming within a "fast-paced nature of online communication" would lead EMC users to resort to Latin script as a cost effective solution for time and technological constraints" (Bou Tanios, 2016, p. 13).Within the Lebanese context, Bou Tanios reports "familiarity" as another reason for the practice of Romanization among the older EMC users, as it used to be the norm during online chatting.In that respect, Standard Arabic is described as being "unacceptable" and "weird" (Bou Tanios, 2016, p. 13).Al-Khatib and Sabbah (2008) stress that "euphemism", topics that might be offensive to talk about in such conservative societies, plays a central role in driving Jordanian university students to codeswitch between their dialect and English in their e-communications.

Kuwait Getting Global
Kuwait is a country which has been increasingly affected by the globalization of English due to its leading role in the world of technology, academia and business.For decades, the country has been undergoing a tug of war between the nationalists and the globalists.Mahgoub ( 2007) for example, describes Kuwait as a country "rushing towards modernization without comprehending its drawbacks".In her study on the effects of global English on culture and identity in the UAE, a country with a very close cultural and sociolinguistic environment of Kuwait, Hopkyns (2014) expresses the nation's concerns about the negative effects of such a powerful language on the local languages, cultures and identity, despite the great powers and knowledge global English would bring to the country.She further claims that the worldwide "homogeneity-heterogeneity debate" seems to be even stronger in the Arabian Peninsula.Hopkins concludes that, although her study of Emirati undergraduates' and teachers attitudes towards English seem to be positive, the study reveals an association between the growing usage of English within the Emirati society and a cultural fragility.She adds, "this needs to be taken into consideration with regard to teaching and learning English in Gulf universities, with a view to making longer term changes" (p: 12).
In Kuwait, English has been officially introduced to Kuwaitis through its education system.The language is introduced in the main stream State Schools as a foreign language, whereby English is taught as a major subject for an hour a day as soon as the child starts schooling.Advanced levels of proficiency in English are required to excel at the university level, especially in the colleges of science, engineering, medicine and business, which tend to be mainly instructed in English.On top of that, recruitment in prestigious jobs in the country require exceptional proficiency in English, but surprisingly not Arabic, to gain a post.English is also widely spoken in business and diplomatic circles.Akbar (2007) describes Kuwaitis as therefore facing a double standard national linguistic ideology; an overt ideology of the community's preservation of its national language variety (Standard Arabic), and a covert one of the sought after native-like proficiency in English that enables Kuwaitis to succeed in entering the State most prestigious walks of life.As a result, Kuwaiti parents have been progressively enrolling their children into English private schools from which they can gain the required native-like proficiency in the language.Reports released by the annual education censuses indicate that the number of Kuwaiti children enrolled in such schools tends to grow increasingly higher from only 3% during the year 1994, to 25% in the year 2003 (Ministry of Education, 1998Education, , 2003)).According to Statista (statista.com),the number has reached 257,405 students in the private schools, comprising over 35% of the total number of students by the year 2014/ 2015.Dashti (1997) claims that English has changed into a "more central code in the lives of Kuwaitis", which has significantly changed the status of English from a foreign language into an ESL status.Dashti adds, such a linguistic status change has resulted in the emergence of a code-switching variety between English and Kuwaiti amongst the younger generation in the country.
Technology-wise, Kuwait offers a dynamic telecoms sector with a strong bias towards mobile infrastructure and services.The country offers opportunities due to its high penetration of mobile and vast LTE coveragecombined with an emerging mobile content and services sector.
Three main mobile operators in Kuwait are forging ahead and developing both infrastructure and services.All three are offering LTE services and progressively attempting to explore 5G opportunities.
Kuwait has demonstrated strong growth in Internet users due primarily to the high ownership of smart phones amongst the population as well as a large number of households with access to either a computer or tablet.The central agency for Information Technology (CAIT) provided the data on the household usage, claiming that smartphones are found in around 99.7% of households, 60% own a computer or a tablet, and around 50% own their personal tablets.The institution further reports that computer use in the households stands 28% below smartphone usage during 2017.CAIT further claims that 8 out 10 households have access to internet, while the non-internet households can access the net through their mobile phones.When considering such intensive consumption of emerging communication media, one would anticipate a significant impact on language change and usage amongst smartphone users in Kuwait.
Just like its Arab neighboring countries, Kuwait displays a case of diglossia, with Standard Arabic (SA) being a "high variety" used in formal contexts, that tends to coexist with Kuwaiti Arabic dialect (KA) as the community's "low variety", used in informal contexts (Ferguson, 1972).Thus, SA is widely used in educational, religious and official documents, while KA is used in everyday speech.Similar to other Arabic dialects, KA features its unique phonology, syntax, and semantic features that are distinctively different from SA.
When considering the ongoing change of status of English from an EFL into an ESL in Kuwait, and its empowering socio-economical effect within the Kuwaiti society, the community is believed to consistently move towards bilingualism.Such a bilingual setting, I believe, provides Kuwaiti speakers with a number of language choice options that tends to extend to online settings.Lee (2015) claims that when English is an option, internet users will more likely use it in their online communications, especially when the users do not share the same first language.Nevertheless, some researchers in the field claim that even speakers who share the same first language may opt for using English over a local language variety due to cultural, social or personal motivations.In the Jordanian community, for example, English, or a mixture of English and Jordanian tend to be chosen over Jordanian dialect only when the users negotiate sensitive topics such as "homosexuality" (Bianchi, 2012).Lee (2007) attributes the choice between English only or a mixture of English and a local dialect to the concept of synchronicity. Lee further illustrates, in informal CME contexts, such as chatting and instant messaging platforms conforming to face-to-face exchanges, code mixing would more likely be chosen over the English only.Nevertheless, Internet users may use English only in official CME platforms such as e-mails and Facebook, where the speech gets less spontaneous due to the lack of synchronicity.Standard Arabic on the other hand is barely used in informal CMC contexts; its usage seems to be more reserved for formal communication (Warschauer et al. 2002).

Codeswitching or Codemixing
Codeswitching is a linguistic phenomenon that broadly refers to the systematic alternation between two or more languages or varieties of the same language during oral/ and/ or written discourse (Mahootain, 2006).Gumperz describes the phenomenon as a juxtaposition of two different linguistic systems within the same speech exchange (Gumperz, 1982).In contrast to the layman perception of the practice as reflecting incompetence in one or both languages, researchers in the field view the practice as a natural result of acceptable proficiency levels in more than one language (or language variety) (Mahootain, 2006).A few researchers distinguish between codeswitching and codemixing, where the former is defined as a switch taking place at the discourse/ sentence level, while the latter signaling a switch at a clause/ word level.Singh (1985) describes a switch within the same statement as codemixing, while a switch at a different unit of speech to indicate codeswitching.
In his investigation of e-discourse, AbuSa'aleek (2014) describes his e-discourse sample to be predominated with both codeswitching and codemixing, suggesting the practice to be at the core of the linguistic profile of e-communications as practiced by Arab young generations.Taha (2015) states that the technology advancement world-wide has intensified the use of codeswitching as Latin letters are easier and more accessible than Arabic letters on mobile keypads.Such a heavy usage of Arabizi on smartphones by Arab youngsters might be gradually posing a threat on the young Arab identity, Taha claims.
For the sake of the present study, the researcher will use the term codeswitching when the switch takes place at the discourse or sentence level.Codemixing, however, will be used to signal a switch within the same clause/ word level.

Objectives of the Study
This paper aims mainly at investigating the linguistic features of Arabizi as produced by the young generation in Kuwait.The study also aims to identify the reasons for which the practice has been persistent amongst the youth community in the country.With this in mind, the present study is an ambitious attempt to answer the following research questions: A. What are the main linguistic features of Kuwaiti Arabizi as performed by young Kuwaitis in relation to,

Data Collection & Tools
The primary data for this study was a corpus text collected from group e-conversations as well as private exchanges on the WhatsApp of 35 Kuwaiti young smartphone users aged mainly between 18 and 30.The informants were briefed on why and how their contributions will be used and assessed in the study.They were accordingly asked to capture all their WhatsApp data written within a time frame of two weeks after deleting any identity revealing cues.A family WhatsApp group dynamics was additionally utilized to examine the utilized code context, sender and recipient.All in all, samples of 400 e-messages ranging between the length of a clause, a sentence, several sentences and a series of paragraphs were sent to the researcher.Each linguistic aspect deduced will be illustrated with a number of examples as they appeared in the corpus followed by an English translation for the sake of clarification.
Due to the heterogeneity of the collected sample, the researcher decided to illicit another corpus text that tends to represent a more unified nature in terms of its content, setting and purpose.Such controlled corpus data will be utilized to explore the consistency of the linguistic profile represented by the spontaneous primary data, when other factors such as topic, setting and content are strictly controlled.The researcher has therefore sent a short story to her informants and asked them to rewrite it in a style of their own and send it to a friend over their smartphones.They were then asked to capture their rewritten story and forward it to the researcher.
A self-report questionnaire designed by the researcher has also been utilized to elicit data on both language use in e-communication scripts and the reasons behind such choice.The questionnaire has mainly included questions to elicit socio-economic/ socio-linguistic profiles of the informants, alongside a list of reasons for using the script of their own choice to stimulate the informants' feedback.A room has additionally been provided for any other reasons that are not included within the list.The questionnaire was designed on the online SurveyMonkey digital tool and was advertised on Twitter with a short tweet addressing the required sample of respondents.The questionnaire link was also sent to the young Kuwaitis within the social circle of the researcher via WhatsApp.All in all, 116 respondents filled in the questionnaire and posted the results on the website.

E-communication Corpus Results
The linguistic/ sociolinguistic analysis of the corpus data reveals the following tendencies: A thorough investigation of how vowels are presented in Kuwaiti Arabizi as presented in the spontaneous WhatsApp texts signals a clear tendency of the deletion of unstressed vowels as they occur in Kuwaiti words, while stressed vowels are almost always preserved.A word like "7alch" meaning "your situation" for instance includes two vowels, a long stressed vowel in the first syllable of the word "7a" and a short unstressed vowel in the second syllable "lɪch".In most cases, the data shows that Kuwaitis tend to preserve the long stressed vowel, slightly shorten it at the expense of the short unstressed vowel, which is regularly deleted.The following actual discourse may illustrate the point clearer: 7alch 7al 5watch (if you let this go) nafs kl mara.Meaning, Your situation is just like your sisters' if you let this go as usual.
According to Standard orthography rules, the sentence should be written as follow: 7aalɪch 7aal 5awaatɪch (if you let this go) nafs kɪl marra.
However, as a result of a clear tendency of shortening the long stressed vowels at the expense of deleting unstressed short vowels (may be as a result of the fast-paced spontaneous conversations on the WhatsApp), short vowels were mainly deleted, while long stressed vowels were slightly shortened.Rather than associating this tendency to the expected mental reading of the written messages by the recipients as structured by phonological principles, Gordon (2011) claims the tendency is more likely governed by the effect of "character-to-character transliteration" of Arabic orthography, in which unstressed short vowels are typically absent from its original orthography, while only stressed long vowels are represented.

A Distinction Between Long Stressed Vowels and Short Unstressed Vowels
The investigation of the vowel system in Kuwaiti Arabizi has also shown a good number of cases of unstressed short vowels being preserved.A closer investigation of those cases signals the application of another vowel orthography system; long stressed vowels are doubled (breaking the orthography vowel system of Standard Arabic), while short unstressed ones are preserved.A word like "aakher" for example, is written with a double "a" vowel to be distinguished from the short unstressed second vowel "e" (syllables preceded with an apostrophe indicate stress): 9ad'geeni 'aakher hammi il'7een si'waalif dala3 meaning 'Believe me, my last concern for now is your bratty behaviour'.
Such unexpected finding seems to challenge Gordon's claims of corresponding Arabizi vowel deletion to the influence of Arabic orthography.The finding therefore suggests three factors to be at play here; mental reading via phonemic decoding, Standard Arabic orthographic rules, as well as the communication pace.In other words, Kuwaiti Arabizi users might adhere to a phonology based mental reading when they have enough time to do so.The concept, however, requires further confirmation in a study that covers a wider range of corpus and focuses specifically on the vowel system in Kuwaiti Arabizi.

Consonants Are Always Preserved
Our main corpus shows that Kuwaitis do not delete word consonants, regardless of their stress or position when they use Kuwaiti words.In fact, even silent consonants that are uniquely related to the spoken language only are never dropped in e-communication.The definite article in Arabic for example has two versions; one that is pronounced called "al qamariya", and another version where the [L] sound gets silent "al shamsiya".Utilization of one or the other is governed by the attached sound in Arabic.Our corpus shows a clear representation of the [L] sound regardless of it being one or the other.Yet, they are more likely to adhere to dropping silent consonants which do not cause ambiguity when they code switch to English.To illustrate my point, the following are a few examples from the main corpus data (words in bold are the English ones with dropped consonants, words between brackets are Kuwaiti ones with silent preserved underlined consonants):

Rawaaan food w9aal Comin
Rawan, the food is here Coming I shudnt even reply to ur ma9khara I shouldn't even reply to your nonsense.

Ana bghait agolich somthin
I wanted to tell you something.

La (titwa83ain) mny askt o I accept all the bullshit you throw my way
Don't expect me to shut my mouth up and accept all the bullshit you throw my way.

Im talkin sm3eny
I'm talking, listen to me!

Haatha illy gaala (elryal) when I was talkin to him!
This was what the guy said when I was talking to him!

Phoneme Dialectal Shifts Tend to Be Expressed in the Spelling of Arabizi
Arabic is a term that has been linguistically used to refer to a wide range of dialects, many of which are considered to be mutually unintelligible (Gordon, 2011).Depending on the geographical region, a group of dialects seem to mirror more common linguistic properties including lexical items, prosody, and phoneme dialectical shifts.The lexical item "abeek" meaning "I want you"in Kuwaiti for example is represented in the term "3awzaak" in Egyptian, "badiyaak" in Levantine.Gordon claims that, while we can hardly view such various dialects as a unified class, EMC orthographic systems reflect a similar set of linguistic processes such as numeric substitution for Arabic consonants and vowel deletion.Gordon further proposes that a linguistic investigation of one dialect, Levantine in his study, presents a comprehensive view of EMC in the Arab World as a whole "as seen through the lens of one particular dialect" (2011, p.14).
An essential linguistic variation in Arabic dialects is the deployment of a range of phoneme shifts across various dialects.Kuwaiti for example more likely uses the consonant shift of / j/ to represent the phoneme / dʒ/ in SA in the spoken language of the sedentary Kuwaitis.Levantine shifts to the glottal stop / ʔ/ to represent the phoneme / q/ as in the word / ʔlb/ meaning "heart" originally pronounced in SA as / qalb/ .One of the noted features in WhatsApp corpus was to find out whether the users abide by conventional Arabic orthography rules or innovate their eccentric spelling system that corresponds to their spoken form.I therefore looked closely into how their e-messages represent the widely phoneme shift utilized in Kuwaiti dialect including the following: a. / dʒ/ → / j/ b. / q/ → / ɣ/ , / g/ or / dʒ/ c. / k/ → / tʃ/ Our corpus shows that Kuwaiti youths tend to adhere to the phoneme dialectical shift in its written form as shown in the following examples (each shifted consonant will be underlined, bolded and followed by its original phoneme): R7 2yeb (2dʒeb) ill 2ghradh ilbait I'll bring the stuff home.

Chm marra 2gollk la tkhaleni 2st3yl (2st3dʒl)
How many times do I tell you to not make me rush!

Ana fi library o ma3ndy wgt (waqt)
I'm in the library.I have no time.

Kani wagf (waqf) jddam (qddam) il bait
He was standing in front of the house.
The finding indicates that Arabizi as used by Kuwaitis appear to more faithfully represent its dialectical phoneme shifts rather than the conventional standard Arabic orthography.In that respect, Arabizi is more likely considered as transcription than transliteration.The finding seems to contradict Levantine EMC which revealed much more adherence to the word conventional orthography rather than the dialectical phonemic shifts (Gordon, 2011).
3.1.5Basic or Advance Arabizi?Kenalli et al. (2016) classifies two levels of Arabizi, each with its own character encoding system for sounds that do not exist in the basic Latin alphabets.While an advance Arabizi utilizes a combination of English figures incorporated with an apostrophe sign, basic Arabizi uses conventional Latin alphabets to represent similar sounds.
Our corpus suggests Arabizi as practiced by the young Kuwaiti generation tends to implement an advance system.Yet, there is also evidence of a few cases of basic Arabizi implementation, especially by the older users (26+).It is worth mentioning at this point that our corpus has been mostly provided by 18-25 years old respondents, which could explain the low recurrence of basic Arabizi orthography as opposed to its advance system.Users of 26+ of age have once been the first generation of digital chats via SMS, which could explain its users' abidance to the American Standard Code for Information Interchange (ASCII) that does not employ numerical style (See tables of Basic & Advance Arabizi).(Knas, 2009).
Utilizing emoticons have also been detected in the corpus.Three main features were widely signaled: capitalization represents anger, excitement and loudness; letter repetition represents emphasis; vowel deletion/ preservation represents the pace of communication.When in rush, vowels are more frequently deleted than preserved."9a7" meaning "right" for example has been spotted more frequently as "97", "9a7" sometimes, and "9a7777" in a few cases, the latter to imply emphasis.The area would certainly benefit from further research that focuses mainly on this aspect of Arabizi.
Kuwaiti users of Arabizi also tend to rely heavily on Emoji images.The frequency and meanings of each Emoji image is however out of the scope of this study.Future research the use of Emoji images in Arabizi would be very helpful in that respect.

Abbreviations or Shortcuts
Arabizi as deployed by Kuwaitis lack a very prominent feature of English EMC, namely abbreviations.Our data reveals a complete absence of abbreviations in their Kuwaiti words.Yet, the same users rely heavily on abbreviations whenever they switched to English in their conversations.Crystal (2008) explains the relative lack of phonetic shortcuts in Arabic EMC to the nature of Arabic which, unlike English, has a high sound-to-character correlation.The present study's findings in that regard supports Crystal claims of the universality of English, not only as a spoken and written language, but also as an EMC tool across various languages.The following are a few examples of the English abbreviations spotted in the WhatsApp chats:

Lol (laugh out loud)mama
Lol mom.

Brb (be right back).
I'll be back soon.

Mn 9jch! OMG (Oh My God)
Are you serious!Oh My God!

Rj3 mn wain, her bday (birthday)?
Where did he come back from, her birthday?5alaa9!G2g (get to go) Enough!Get to go! Ok (Okay), tyt (take your time) Okay, take your time.

Codeswitching, Codemixing or Both?
One of the pivotal linguistic features the current study investigates is the range of codeswitching/ codemixing consumption in Arabizi amongst Kuwaiti bilingual community.A quantitative analysis of the corpus shows a high frequency of alternation between Kuwaiti and English in the informants' WhatsApp messages.The prime vs complementary language (Kuwaiti or English) was specified based on the "bulkiness" of the elements of either language in the messages.The findings in that respect demonstrate a higher tendency of using English/ Kuwaiti texts more often than Kuwaiti/ English, Kuwaiti only or English only.An overall of 73% of the text messages was an alteration between English and Kuwaiti, with English being the matrix language in 52% of the corpus embedded with Kuwaiti, and approximately 10% Kuwaiti embedded with English.Kuwaiti only comprised 9%, while English only comprised approximately 15% (12% of which were written by male participants).A few of the messages (2%) were neither considered as Kuwaiti/ English nor as English/ Kuwaiti due the even distribution of both languages in the investigated texts (See the following figure for illustration).
Another si and 37% intersenten lexical sw English an By the sam intersenten Kuwaitis.male parti be more e respectivel .orgignificant find respectively).ntial (between witching, the sw nd sparkled wit me token, an e ntial and phra Among the in cipants as opp excessively pe ly).Henceforth, the present study introduces a strong variability between the codeswitchers and codemixers in relation to gender.The finding might reflect a distinguishing style of Kuwaiti/ English alternation in relation to a key sociolinguistic factor, namely gender.Nevertheless, due to the small size of the study's sample (35 respondents only), it might be too early to overgeneralize the finding beyond this research.
Age-wise, the younger participants (18-25) were heavier code-mixers, with nearly the bulk of their texts written in English mixed with a few Kuwaiti lexical items.The investigation displays around 78% of the code-mixers were from the younger age group.In view of the uneven distribution of the various age groups across the study's sample, once again, it would be unwise to draw a general sociolinguistic rule on the topic.
The following are a selected cases of examples from the actual collected corpus data (Kuwaiti words, phrases and clauses are bolded and followed by their English translation for the sake of clarification).Further examples of the collected corpus data can be found in Appendix B. a. Code-mixing (Bulk in English sparkled with a few Kuwaiti tags and words): O ba3dein (and after that) u work out a ratio of expense according to the ma3aash (salary).

But if he's studying moo wgta (it's not the right time)
La maa kallamta (no I didn't talk to him), iv been busy.

Rja3 min wain (where did he come back from), her bday?
Ilmushkla (the problem is) im being shwy (a bit) difficult, abi killshy (I want everything) sterilized and cleaned 3dl (thoroughly) ijel.ccsenet.Two further reasons were also expressed by two informants, one of which is having a difficulty in spelling words in Arabic correctly, and as a result using Arabizi as an escape.Another reason relates the use of Arabizi to the initial lack of Arabic font in the technological devices when users used to chat on MSN messenger and the like.This reason in particular was expressed by a 36 year old informant.

Story Writing Responses
As mentioned earlier, there was a need for a unified corpus of Arabizi to find out how closely the described linguistic profile of Arabizi conforms to reality.The story (which was written in Kuwaiti with Arabic alphabet only) was presented to the respondents of the online questionnaire as an optional item in order to avoid burdening the respondents with an exhausting task.Only 17 out of 116 respondents managed to accomplish the task.However, bearing in mind that the data has been used for further examination of Arabizi linguistic profile, the researcher believes that the 17 responses were enough to serve the purpose.
2. Older respondents (26+) stick more frequently to the Basic Arabizi symbols rather than advanced (apostrophized symbols) Arabizi.Younger ones on the other hand follow the symbols of what has been described as Advanced Arabizi.It is important to clarify at this point that such age-structured rule has been broken in some rare cases: Wathi7 vs wa'67 meaning "clearly" Gha9b vs '39b meaning "forcefully" Khosh vs '5osh meaning "cool" 3. Respondents from Private/ English schools tend to frequently alternate between Kuwaiti and English intersententially, interphrasially and lexically.Respondents from Public/ Arabic schools on the other hand tend to stick to Kuwaiti.When they mix codes, they will most likely alternate between Arabizi (Romanized alphabet) and Arabic (Arabic alphabet).

Discussion
The present research data represents a linguistic profile that tends to uniquely distinguish Arabizi as used by Kuwaiti youths in Kuwait, with the following most prominent features:

Transliteration & Transcription
Kuwaitis tend to transliterate their e-conversation when they use the Arabic definite article "al" meaning "The".Yet, they tend to rely heavily on transcription to add up some extralinguistic features due to the heavy reliance of Kuwaiti dialect on the tone of the expression.Emoticons, all word capitals, repeated vowels or consonants are frequently used to reflect the required extralinguistic features.Kuwaitis' phoneme shifts also reflect another aspect of transcription as opposed to transliteration.The shift from [dʒ] to [j], and [q] to [g] are frequently represented in the data.

Phonemic Decoding
The data also seems to partially conform to the linguistic principle of "textese" governed by the principle of "figure and ground" predicting that consonants appearing in unstressed positions are more likely deleted, vowels in unstressed positions might be either reduced or deleted (Kul, 2007).Arabizi, as our data shows in the Kuwaiti context, seems to preserve consonants regardless of their strength, yet weaken unstressed vowels and even more frequently delete them.

Globalization of English in E-Communications
Kuwaiti Arabizi has also shown resorting to English texting abbreviations such as lol, brb,omg and the like.This in fact adheres to Crystal (2001) claims of non-native English speakers deploying similar English abbreviations that might suggest them as progressing to universal practice in online communications.

Reoccurrence of English/ Kuwaiti Code-Switching and/ or/ Code-Mixing
The present research data also reveals a regular appearance of code-switching between English and Kuwaiti dialect at a discourse level, intersentential and interphrasal positions.Frequent reappearance of lexical switching also signals the practice of code-mixing as well.The profile looks very similar to how AbuSa'aleek (2014) described Arabizi as practiced within the Saudi dialect.Future research is recommended to shed the light on the factors that motivate Kuwaiti users of Arabizi to switch/ and/ or/ mix between English and Kuwaiti in their e-communications.

Why the Romanization?
The present research results suggest three main reasons that drive Kuwaiti young people to Romanize their e-communications; Arabizi is fast-paced, easier and more habitual than standard Arabic.To a lesser degree, Arabizi has also been perceived as being more flexible in discussing taboo topics (sex, religion) and cool (reflecting the younger generation style of e-communication).The results however tend to reject the idea of using Arabizi to reflect a high socio-economic status of its users.This in fact may suggest that Kuwaiti youths still perceive Kuwaiti as an identity of high status despite their profound use of English (words and letters) in their e-communications.The effect of such a prominent Arabizi usage by Kuwaiti youngsters would be a hot topic to tackle in future research on the topic.At this point, I'd like to invite my fellow sociolinguists to embark on studies that look deeper into the effects of Arabizi on Kuwaitis' social identity as the aspect has only been briefly touched upon in the current study.

Conclusions and Suggestion for Further Research
The globalization of English-based technology has led the present day Arab generation (including Kuwaitis) to rely on the use of Latin alphabets when communicating digitally (Allehaiby, 2013) in an innovative Latinized Arabic script known as Arabizi.Due to a number of major differences in the phoneme system between English and Arabic, users of the script employ numbers and symbols to represent non-existing phonemes and letters.Kuwaiti users of the script also rely heavily on codeswitching and codemixing between English and Kuwaiti to express themselves across social platforms including WhatsApp.As a matter of fact, the study's statistical analyses establish a set of sociolinguistic cues, namely gender and age of the Arabizi user, that tend to present different linguistic styles of the fusion of Kuwaiti and English in the practice of codeswitching and codemixing.However, I would be attentive to overgeneralize such findings on a larger scale due to the skewed distribution of the sample in relation to age groups.With the script becoming trendy across the Arab World, Arabizi is a phenomenon that should not be overlooked in the realm of computer-mediated contexts (CMC) (Attwa, 2012).From a linguistic point of view, the present research findings views Arabizi as holding features that can neither be associated to speech, nor writing.The finding in this respect tends to conform to Crystal's view of CMC to be portrayed as a "third medium" of communication, where the technical restriction of 160 characters per message has led its users to deploy new linguistic forms such as acronyms and abbreviations (Crystal, 2001).Kuwaiti Arabizi users tend to deploy some additional unique linguistic features such as vowel deletion/ preservation, consonant stabilization, definite article "al" meaning "the" transliteration rather than transcription, dialectical phoneme shift representations (suggesting phonemic decoding), extralinguistic features representation (unconventional orthographic rule representations & extensive use of Emoji images), resorting to English abbreviations while avoiding Arabic counterparts, and an extensive usage of codeswitching/ mixing in writing.
Further studies should investigate the use of codeswitching/mixing in Kuwaiti Arabizi from a sociolinguistic perspective, i.e, who does the switch/mix, to whom, and for what reasons.Unlike Saudi users of Arabizi, a country which is geographically and culturally so close to Kuwait, Kuwaiti Arabizi users tend to use both Advance and Basic Arabizi.An age distinction reflects more of 26+ users applying Basic Arabizi rules as opposed to younger users applying Advance Arabizi rules.The finding might indicate a future transition towards more excessive use of Advance Arabizi.
The present study also signifies a number of sociolinguistic reasons for which Kuwaiti users of Arabizi employ the script in their e-communication across social platforms.Three main reasons were strongly confirmed by the research informants; Arabizi is more efficient (faster, easier, more habitual), less rigid when discussing taboo topics, and more trendy as it reflects the style of the younger generation.Associating Arabizi with prestige however has not been advocated for by the informants.The finding may suggest that Arabic has not yet lost its prestige among the younger Kuwaiti generation.Future research is recommended to look into whether Arabizi poses a threat to young Kuwaitis identity, hence, leads to the Arabic language attrition.
Given those findings, we can strongly assert that Arabizi as conducted by Kuwaiti young generation has presented itself as noticeable rich and rule-governed linguistic phenomenon.Different from Levantine Arabic EMC described by Clara Gordon, its orthography features a number of unique phonetic characteristics in the case of consonant and vowel representations.That is to say, unlike the consistent transcribed consonant system represented in Levantine Arabizi, Kuwaiti Arabizi tends to alternate between transcription and transliteration in its consonant system representations.In the case of its vowel system, Kuwaiti Arabizi seems to be affected by the pace of the conversation, the stress and length of the syllable, as well the age group of the performer.To put it another way, Kuwaiti users of Arabizi conform to the trend of vowel deletion when they are in a hurry (or anxious), when the vowel appears in an unstressed syllable, and most importantly, when the performer's age lies between 18-25 age group.Another key point in that respect is the high sensitivity of Kuwaiti Arabizi to both the phonology of the dialect and the Arabic orthography rules.The linguistic alternation between the definite article "al" orthographic representation and the phoneme dialectical shift representation may be considered as most compelling evidences for transcription and transliteration to be both at play.
Due to the absence of face-to-face interaction, Kuwaiti users of Arabizi resort to extensive extralinguistic cues to make their EMC application less ambiguous.In the light of the findings of the present work, researchers in the field of digital linguistics are prompted to undertake additional work on the issue.
All in all, users of Kuwaiti Arabizi seem to have sensed the need to adapt to the restrictions of digital communication, and as a result, resorted to a complex set of orthography and communicative rules that will enable them to knit an alphabet system in order to represent languages with extremely different phoneme and literary writing profiles.It must be remembered though, similar to other EMC, Arabizi tends to be susceptible "to fads and the influence of the rapidly changing interfaces that facilitate it" (Gordon, 2011).

Figure
Figure 2. The

Ba3ad
(whatever), what can u do.Akeed (Are you sure) he isn't studying?b.Code-mixing (Bulk in Kuwaiti sparkled with a few English lexemes): Dzely il (send me the) receipt number 5l9eny (hurry up) Al7en eb flim by5l9 (I'm now watching a movie, will finish) in a bit.6al3een way dania (we're out with Dania) O (and) the cuzins.c.English/ Kuwaiti code-switching Mama il walad (mom the boy) is doing ok in the house, 3adi yn6r shway (is it ok to wait) cus im discussing things with moe He already talked to moe, 3adi mertaa7 (he's relaxed) The girl hasn't left the house in 2 wks shfeech inty (what's wrong with you) Learn to use taxis law sam7taw (please) Hes a recovering addict who is struggling to make it to work, ma ynf3 chthi (this isn't working) d.Kuwaiti/ English code-switching shes dealing m3a nas g3d yswon s7r bil7osh (with people who are doing black magic outside) ga3d m3a moe t7t (I'm talking to moe downstairs), chill tadrin shloun bachr il9b7 (you know what, tomorrow morning) ill pass by e. Code-switching (Neither English/ Kuwaiti, nor Kuwaiti/ English) Walla maako shay moo (there's nothing that's not) already maktoob (documented) by scholars before us.

Figure 4 .
Figure 4. Reasons for using Arabizi on smartphones by Kuwaiti youths 3.1.1UnstressedVowels Are Rarely PreservedKuwaiti orthography follows Standard Arabic rules and regulations in general.Arabic has mainly been described to have relatively few distinct vowels (a, i, u).Nevertheless, some of them come in different variations depending on the dialect and surrounding consonants.In standard Arabic, the vowel [a] for example may be lengthened or shortened [a:, a], backed or fronted [a, ae] depending on the lexeme as well as the surrounding sounds.The vowel [i] on the other hand has a number of dialectical variations.Kuwaiti Arabic for example deploys the central mid long vowel [ɛ:] as it is in "zɛ:n" meaning "good" as well as [ɔ:] as it is in a word like "Tɔ:fə" meaning "wall" both of which do not exist in SA.Investigation of the vowels used in the corpus represents inconsistent vowel manifestations of short and long vowels in Arabizi.Thus, the vowel [u] is represented by [o] and [u]; [ɪ] is represented by [e] and [i].The long vowel [u:] on the other hand is represented by [oo], [o] and [ou].[i:] is represented by [e], [ee] and [eɪ].[a:] is represented by [aa] or [a].Users do not distinguish between a back [a] or a front one.Kuwaiti specific vowels [ɛ:] and [ɔ:] were also inconsistently represented by [ea], [ai] and [eɪ] for the former, and [oo], [o] and [ou] for the latter.