Content area
This study investigates code-switching practices, perceived ease, and attitudes within the Armenian community of Kuwait, focusing on how these dimensions vary by age, gender, and educational background. Drawing on data from 225 participants, the research employs the Bilingual Code-Switching Profile (BCSP) as the principal instrument, complemented by linear and logistic regression analyses. The results show that younger participants tend to switch languages more frequently and with greater ease, particularly in informal and self-directed contexts. On the other hand, older participants emphasise the naturalness of switching and exhibit greater ease in understanding multilingual written texts. Gender differences were also observed: female respondents were more concerned with how naturally their switching was perceived, while male respondents were more likely to identify with communities that code-switch. Education was associated with increased switching in formal settings and greater ease in comprehending bilingual written materials. These findings point to the social conditioning of code-switching behaviour and highlight the influence of demographic variables on shaping bilingual practices in diaspora contexts. The study contributes to the broader understanding of multilingualism in under-researched minority communities and offers insights into how linguistic repertoires are managed across generational, social, and institutional settings.
Abstract-This study investigates code-switching practices, perceived ease, and attitudes within the Armenian community of Kuwait, focusing on how these dimensions vary by age, gender, and educational background. Drawing on data from 225 participants, the research employs the Bilingual Code-Switching Profile (BCSP) as the principal instrument, complemented by linear and logistic regression analyses. The results show that younger participants tend to switch languages more frequently and with greater ease, particularly in informal and self-directed contexts. On the other hand, older participants emphasise the naturalness of switching and exhibit greater ease in understanding multilingual written texts. Gender differences were also observed: female respondents were more concerned with how naturally their switching was perceived, while male respondents were more likely to identify with communities that code-switch. Education was associated with increased switching in formal settings and greater ease in comprehending bilingual written materials. These findings point to the social conditioning of code-switching behaviour and highlight the influence of demographic variables on shaping bilingual practices in diaspora contexts. The study contributes to the broader understanding of multilingualism in under-researched minority communities and offers insights into how linguistic repertoires are managed across generational, social, and institutional settings.
Index Terms-code-switching, bilingualism, Armenian diaspora, Kuwait, sociolinguistics
1. INTRODUCTION
Code-switching refers to alternating between two or more languages within a conversation, occurring within or between sentences (Bullock & Toribio, 2009). It can occur for reasons such as social solidarity, topic shifts, or emotional expression (Abdel Tawwab, 2014). Research by Gosselin and Sabourin (2023) indicates that bilingual individuals who frequently engage in code-switching might exhibit cognitive advantages, particularly in inhibitory control, thereby establishing a connection between code-switching behaviour and cognitive capacities.
A massive number of people around the world are bilingual, especially in countries with myriads of multilingual labourers. In Kuwait, a country where expatriates account for 70% of the population (World Population Review, 2024), those who are in command of Arabic, Bengali, Hindi, Urdu, and Tagalog, to name a few of the many languages spoken in the country, tend to code-switch frequently during their daily interactions. Hence, in a multilingual society such as Kuwait, using more than one language is common. However, code-switching has not been widely studied in the Kuwaiti context. Numerous studies have explored the role of code-switching in education (Al-Qaysi, 2019; Mahsain, 2014); however, there 15 a gap in the literature exploring the phenomenon of code-switching among the linguistically diverse communities in the Gulf generally and in Kuwait specifically. There is a need to investigate code-switching in Kuwait, especially concerning code-switching among the Armenian community, one of the oldest communities in the country, whose mother language, western Armenian, in 2010 featured on UNESCO's list of endangered languages.
Armenians came to Kuwait from Iran, Iraq, Lebanon, and Syria in the early 1950s after the discovery of oil and the subsequent economic boom. Their number increased in the 1970s following the Lebanese Civil War and the Iranian Revolution (Vosgemadyan, 2011). With a strong emphasis on preserving their cultural identity, Armenians established a school and church in 1960 and pan-Armenian diaspora educational, cultural, scouts, and sports organisations, such as the Armenian Relief Society, Homenetmen, and Hamazkayin (Aprahamian, 1967). The Armenian Apostolic Church plays a central role in preserving linguistic and cultural traditions, serving as a hub for religious observance and social interaction (Pattie, 2005; Jebejian, 2022). Early Armenian settlers were craftsmen, including carpenters, car technicians, electricians, and jewelers (Bournoutian, 2003). Subsequent generations pursued higher education, attending colleges and universities to obtain advanced degrees. As a result, many transitioned from traditional trades to professional fields, becoming engineers, bankers, and doctors (Jebejian, 2022). They are also in close contact with relatives and friends who still live in Armenia, Iran, Iraq, Lebanon, and Syria, and with other Armenians worldwide through social media.
This study aims to examine code-switching patterns among Armenians in Kuwait. 225 members of the Armenian community participated in the questionnaire. The participants were selected using a convenience sampling method to capture a broad representation of this bilingual community. The study employed the Bilingual Code-Switching Profile (BCSP) as the data collection instrument. The BCSP is a validated and reliable self-assessment tool to measure bilinguals' code-switching experiences, proficiency, and attitudes. It was selected due to its comprehensive approach, assessing three key dimensions: Code-switching Use (frequency of language switching in various contexts), Code-switching Ease (ease of producing and understanding language switches), and Code-switching Attitudes (identity and perceptions related to language switching) (Olson, 2022). This study seeks to answer the following research questions:
1. How do different generations within the Armenian community in Kuwait differ in their code-switching behaviors and attitudes?
2. What role does educational background play in shaping these code-switching dimensions?
3. Are there gender-based patterns in code-switching experiences and perceptions?
By applying the BCSP to an underexplored ethnolinguistic group within a multilingual Gulf state, this research contributes new empirical data to the study of bilingualism and diaspora language practices. It offers insight into how intersecting social variables mediate language behavior in a small, diasporic community and extends the methodological applicability of the BCSP to heritage language speakers in a non-Western, expatriate-rich environment.
II. LITERATURE REVIEW
A. Generational Divergence in Code-Switching Practices
The observed divergence in code-switching practices across generations is linked to a complex interplay of sociolinguistic factors, including language proficiency, attitudinal dispositions, and the communicative functions of languages within specific contexts (Yim & Clément, 2021). For example, a comparative analysis of language selection and discoursal characteristics between elderly and youthful members of the Hokkien-speaking Chinese community in Tangkak, Johor, Malaysia, reveals a discernible divergence in linguistic preferences across different generations: older informants predominantly utilised the Hokkien dialect, whereas younger informants exhibited a predilection for Mandarin (David et al., 2009). Furthermore, the elderly displayed limited code-switching, primarily between Hokkien and Mandarin, while the youth demonstrated a wider range of language mixing, incorporating Mandarin, Malay, and/or English.
Jebejian's (2004) research within the Armenian community in Beirut reveals a clear generational divide. Older Armenians exhibit a marked reluctance towards code-switching, associating it with linguistic deficiency. Their reported code-switching instances are infrequent and include Armenian, English, and Turkish, accompanied by feelings of guilt and shame, reflecting a concern for language erosion and endangerment. This perspective contrasts sharply with younger Armenians, who readily employ code-switching involving Arabic, Armenian, English, and French, viewing it as a natural and effective communicative tool. This attitudinal disparity highlights the evolving socio-symbolic values attached to language use across generations.
Zentella's (2017) work provides an essential examination of bilingualism in the context of the Hispanic diaspora in the United States. Her research highlights how bilingual individuals negotiate identities and language choices within a sociocultural landscape shaped by immigration, socio-economic factors, and the pressures of assimilation into Englishspeaking environments. Zentella identifies the complexities surrounding language use, which can include a mix of Spanish and English, often referred to as "Spanglish," to express cultural identity among Latinos in diasporic contexts. She explains that the higher frequency of code-switching among some older bilinguals is not arbitrary but can be linked sociolinguistically to their life histories and the nature of the community networks they were embedded in, which fostered and normalized bilingual language practices over extended periods.
The proliferation of digital technologies and social media has exacerbated generational differences in code-switching. Younger generations, characterised by their digital fluency (McCrindle, 2018), readily integrate code-switching into their digital communication repertoires to express identity, connect with peers, and enhance content relatability (Hukara, 2022). In contrast, older generations, often exhibiting a more cautious approach to digital integration, tend to employ codeswitching less frequently and may hold more conservative views on its appropriateness, driven by differing cultural and social norms rather than linguistic incompetence.
B. The Interplay of Attitudes and Code-Switching
While code-switching may be a common linguistic feature within bilingual populations, attitudes towards it are not monolithic (Jebejian, 2004; Yim & Clément, 2021). Instances perceived as indicative of linguistic incompetence are often met with more severe judgment than those attributed to sociocultural factors (Gardner-Chloros, 1991; Jebejian, 2007). For example, Baker (2003) and Dashti (2015) argue that attitudes and behaviours surrounding code-switching are profoundly context-dependent, challenging simplistic views of language use. Baker highlights the inherent complexity and contradiction within individual attitudes, emphasising the influence of social settings and personal experiences, such as bilingual proficiency and discriminatory encounters, on language evaluation. Similarly, Dashti underscores the contextual specificity of code-switching, demonstrating that language choice shifts significantly across different social domains, driven by communicative purposes and situational factors. Baker's observations of varying valuations in formal versus informal settings are corroborated by Dashti's findings of professional environments favouring code-switching for pragmatic reasons, while intimate gatherings prioritise language loyalty for sociocultural affirmation.
Blommaert and Rampton's (2011) work asserts that the multiplicity of languages and communication strategies present in contemporary society reflects a dynamic and intertwined relationship with identity, cultural practices, and social interactions. Their concept of superdiversity relates directly to the transformation of sociolinguistic practices in increasingly multicultural and multiethnic societies. Globalisation has heightened the complexities of language use, where individuals often draw on diverse linguistic resources, not confined to discrete languages but encompassing an array of sociolects employed situationally, which they term "translanguaging". Garcia and Wei (2014) advance a compelling conceptualization of translanguaging, defining it as a process where bilingual individuals utilise their entire linguistic repertoire to communicate effectively across various contexts. This concept challenges the traditional view of bilingualism as simply alternating between languages and promotes the idea that bilinguals fluidly navigate languages based on situational demands and communication needs. Empirical evidence derived from numerous investigations within educational settings robustly demonstrates that the adoption of a translanguaging pedagogical stance yields demonstrable enhancements in bilingual students' learning outcomes and the positive affirmation of their multifaceted identities.
The relationship between attitudes and code-switching is complex and context-dependent. This study is predicated on recognising that universal generalisations regarding attitudes towards code-switching are elusive, thus providing a compelling rationale for the present context-specific investigation.
C. Gender as a Variable in Code-Switching Practices
Comparative analyses of language choice across diverse ethnic communities reveal distinct, community -specific gender patterns. For instance, within the Vietnamese community in Australia, men and boys exhibited a higher tendency to use Vietnamese in the 'neighbourhood' domain, while women predominantly employed the minority language in the 'transaction' domain (Pauwels, 1995). Conversely, Cheshire and Gardner-Chloros (1998) reported no significant gender differences in code-switching patterns among Greek Cypriot communities in Britain, underscoring the variability of gendered linguistic practices. Wong's (2006) study in Hong Kong revealed a statistically significant gender disparity, with females exhibiting nearly twice the frequency of code-switching compared to males during interviews and higher use of English within their code-switched utterances. Wong interpreted these findings within a sociolinguistic framework, claiming that code-switching is an indexical marker of educational attainment.
Holmes (2006) who analysed real-life conversations from various New Zealand workplaces, discusses the concept of code-switching as a frequent linguistic phenomenon that is not only prevalent in multilingual societies but also shaped by the social roles and identities of speakers. According to her work, various sociolinguistic studies suggest that women are generally more adept at navigating linguistic codes to maintain social relationships. This tendency may be attributed to their societal roles as caregivers and communicators who prioritize relational aspects of conversations. As such, women's code-switching could be viewed as a strategic maneuver for achieving social harmony and rapport, thereby reinforcing their social identities.
Pavlenko and Blackledge (2004) argue that social identities, including gender, are not static attributes but rather fluid, multiple constructs dynamically negotiated through linguistic and social interaction. Central to their framework is the contention that code-switching functions as a crucial strategic resource deployed by speakers for the active performance, contestation, and negotiation of identities. They maintain that prevalent attitudes towards code-switching are deeply embedded within broader social structures and ideological underpinnings. Consequently, these attitudes exert significant influence on individuals' linguistic choices and their interpretation of these choices, thereby delimiting the range of identities deemed legitimate or achievable within a particular context.
Ahmed et al. (2015) examined code-switching patterns in SMS messages among bilingual university students in Pakistan, revealing a gendered distribution of code-switching functions. Females exhibited a higher frequency of codeswitching, attributed to heightened self-consciousness, and thematic domains of code-switching differed between genders, with males predominantly code-switching in discussions pertaining to social life while females focused on personal matters.
D. The Influence of Education on Code-Switching
In educational environments, code-switching is frequently employed as a pedagogical tool, particularly for learners with varying levels of language proficiency (Murtiningsih et al., 2022; Villarin & Emperador, 2023). It facilitates concept clarification, classroom management, and the creation of inclusive learning environments (Aldalbahy, 2022; Alenezi, 2016; Al-Qaysi, 2019). While the efficacy of code-switching as a pedagogical strategy is widely acknowledged, its potential drawbacks, such as limiting exposure to the target language, are also subjects of ongoing scholarly debate.
Beyond educational institutions, higher education is associated with a wider range of code-switching functions, reflecting a refined understanding of the socio-cultural implications of language choice. Educated adults demonstrate a greater capacity to adapt their language use to diverse social contexts, employing code-switching to signal group affiliation, negotiate social identities, and manage interpersonal relationships. For example, Bautista (2004) found that higher education levels in the Philippines correlated with increased "Taglish" use, signifying linguistic proficiency and social identity among educated, urban Filipinos. Similarly, Yim and Clément (2021) observed that Cantonese-English bilinguals in Canada with higher proficiency in both languages exhibited more fluid code-switching. However, the interpretation of code-switching as a marker of linguistic deficiency, as suggested by Hughes et al. (2006), is contested by scholars who view code-switching as a rule-governed phenomenon reflecting intellectual advantage and unique cognitive ability (Parama et al., 2017).
Dewaele and Wei (2014) suggest that higher education fosters sociolinguistic awareness and the ability to navigate diverse communicative landscapes. This proficiency 1s cultivated through exposure to varied linguistic environments and the development of refined communication skills. Chan (2023) corroborated these findings, demonstrating a positive correlation between higher education and both the frequency and complexity of code-switching among homeland bilinguals in Hong Kong, heritage bilinguals in the United States, and immersed bilinguals who migrated from Hong Kong to the United States. Bilinguals with more advanced education exhibited a greater tendency to code-switch and a more sophisticated and flexible code-switching repertoire, highlighting the role of formal learning in shaping bilingual linguistic behaviour. Chan concluded that education equips individuals with the cognitive and linguistic tools essential for analysing and adapting to complex communicative situations, comprehending the social implications of language choices, and developing the flexibility necessary for effective code-switching.
Dashti (2015) reveals a compelling sociolinguistic dynamic among Kuwaitis, demonstrating that Arabic-English codeswitching in social contexts is a marker of elevated educational attainment and social standing. Within informal social settings, code-switching establishes or reinforces social hierarchies, highlighting the intricate relationship between language, social status, and identity in Kuwait and underscoring English as a valuable social and economic asset. This perception, akin to that prevalent in other Gulf nations (AlRumaihi, 2021; Omar & Ilyas, 2018), stems from the prestige accorded to the English language, which is closely associated with globalisation, international business, and advanced education. Consequently, individuals who adeptly navigate between Arabic and English are often perceived as belonging to a more educated and cosmopolitan social stratum.
III. METHODOLOGY
A. Participants
This study examined code-switching patterns among 225 members of the Armenian community in Kuwait. The participants were selected using a convenience sampling method to capture a broad representation of this bilingual community. The sample (Table 1) included 107 males (47.56%), 115 females (51.11%), and 3 respondents who preferred not to disclose their gender (1.33%). Participants ranged in age from below 20 to over 61, with the largest age group being 41-50 years (24%), followed by 21-30 years (23.56%) and 31-40 years (20%).
Regarding educational background (Table 2), 3.11% had some schooling, 41.78% completed high school, 5.78% received vocational training, 38.22% held university degrees, and 11.11% had postgraduate qualifications. Notably, a greater proportion of females held higher educational qualifications, with 14 females holding postgraduate degrees compared to 11 males. The diverse demographic composition allowed a subtle exploration of language-switching patterns across gender, age, and educational background.
B. Instrumentation
The study employed the BCSP as the data collection instrument. The latter is a validated and reliable tool for measuring bilinguals' code-switching experiences, proficiency, and attitudes. In validating the BCSP, Olson (2022) demonstrated that the BCSP is a reliable framework that effectively captures the subtle distinctions of bilingual experiences, including differential engagement with code-switching in various contexts (Zhong & Fan, 2023). The BCSP has been shown to yield consistent and valid results across varied bilingual populations, making it an appropriate tool for exploring the complex sociolinguistic dynamics of the Armenian community in Kuwait. The questionnaire was administered in Armenian to ensure participants' comfort and accuracy in self-reporting. Examples of different types of code-switching were provided to standardise understanding, and participants were required to identify instances of code-switching before proceeding, ensuring a shared conceptualisation of the phenomenon.
C. Data Collection
Data collection was conducted using Google Forms, ensuring accessibility and convenience for participants. The questionnaire was completed in approximately 5-7 minutes. Before participation, respondents were provided with a consent form explaining the study's purpose, procedures, and benefits, as well as assurances of anonymity and confidentiality. Participation was entirely voluntary, and respondents were informed of their right to withdraw at any time without penalty. The BCSP was administered following best practices in bilingualism research to enhance validity and reliability, including clear definitions, context-specific questions, and a culturally sensitive approach tailored to the Armenian community's bilingual experiences. The collected data were anonymised and securely stored, accessible only to the research team, as stipulated in the consent form.
D. Data Analysis
The data were analysed using regression models to explore the relationships between age, educational background, and gender with code-switching behaviour across different contexts (e.g., school or work), internal cognition (e.g., talking to oneself or counting), and digital communication (e.g., writing on social media). The analysis focused on three main dimensions: frequency, ease, and attitudes towards code-switching. The statistical models examined the occurrence of code-switching and the socio-cognitive factors influencing these patterns. By incorporating a multifaceted analytical approach, the study provided a comprehensive understanding of the multilingual dynamics within the Armenian community in Kuwait.
E. Ethical Considerations
Ethical approval was obtained from the American University of the Middle East, Kuwait, ensuring the study adhered to international research ethics standards. Informed consent was obtained from all participants, and the study maintained strict confidentiality and anonymity throughout the data collection and analysis processes. As outlined in the consent form, the collected information was used solely for research purposes and securely stored to protect participants' privacy. Using the Armenian language for the consent form and the questionnaire facilitated accurate self-reporting while respecting cultural and linguistic sensitivities. Participants were fully informed about the study's objectives, procedures, and rights, thereby contributing to an ethical and transparent research process.
IV. RESULTS
This chapter presents the results of an empirical investigation into code-switching patterns among 225 members of the Armenian community in Kuwait. The findings provide insights into generational differences, educational influences, and gender-based variations, contributing to an understanding of bilingualism within this cultural context.
A. Relationship Between Age and Code-Switching
The regression model examined code-switching in three specific contexts (Table 3):
* Code-switching at School or Work (Statement 16):
A negative relationship was observed between age and the frequency of code-switching in educational or professional settings (Coefficient = -0.160, Standard Error = 0.054, р < 0.05). This suggests that younger respondents are more likely to code-switch at school or work than older respondents. The negative coefficient indicates that there is an associated decrease in age for each unit increase in the frequency of code-switching in these settings. This pattern could reflect a generational shift in bilingual communication practices, possibly influenced by increased exposure to multilingual environments among younger generations.
* Code-switching When Talking to Oneself (Statement 18):
A significant negative relationship was also identified for code-switching when talking to oneself (Coefficient = -0.265, Standard Error = 0.062, p < 0.05). This suggests that younger individuals are more inclined to engage in internal codeswitching. The relatively larger negative coefficient than other contexts implies a stronger age-related difference in this personal and cognitive domain. This could indicate a higher cognitive flexibility or comfort with bilingualism among younger respondents, reflecting changes in language socialisation and identity formation within the community.
* Code-switching When Counting (Statement 19):
In contrast, a positive relationship was found between age and code-switching when counting (Coefficient = 0.190, Standard Error = 0.049, p < 0.05). This indicates that older respondents are more likely to code-switch during numerical tasks. The positive coefficient suggests that as age increases, the likelihood of code-switching when counting also increases. This may reflect a reliance on a primary or more dominant language for cognitive tasks among older individuals, possibly linked to their language acquisition patterns or educational experiences.
B. Relationship Between Age and Ease of Code-Switching
The regression model examined the ease of code-switching in the following contexts (Table 4):
* Ease of Code-switching While Writing (Statement 22):
A negative relationship was observed between age and the ease of code-switching while writing (Coefficient = -0.224, Standard Error = 0.068, р < 0.05). This suggests that younger respondents find it easier to code-switch when writing, such as in text messages or social media, than older respondents. The negative coefficient indicates that age decreases as the ease of code-switching in writing increases. This may reflect greater digital literacy or more fluid bilingual communication among younger individuals who are more accustomed to code-switching in written, informal contexts.
* Ease of Understanding Code-switching in Written Text (Statement 23):
A positive relationship was found between age and the ease of understanding code-switching in written text (Coefficient = 0.176, Standard Error = 0.082, p < 0.05). This indicates that older respondents find it easier to understand code-switching in written communication than younger respondents. The positive coefficient suggests that as age increases, so does the ease of comprehending bilingual text. This may reflect more extensive language exposure or greater linguistic experience accumulated over time, enhancing the ability to process complex language patterns.
C. Relationship Between Age and Attitudes Towards Code-Switching
The regression model examined attitudes towards Code-switching in the following contexts (Table 5):
* Feeling Like Oneself When Code-switching (Statement 24):
A negative relationship was observed between age and the attitude of feeling like oneself when code-switching (Coefficient = -0.192, Standard Error = 0.064, р < 0.05). This indicates that younger respondents are more likely to feel like themselves when code-switching than older respondents. The negative coefficient suggests that the likelihood of associating code-switching with authentic self-expression decreases as age increases. This could reflect a generational difference in bilingual identity, with younger respondents possibly perceiving bilingualism as a natural and integral part of their cultural identity.
* Importance of Code-switching Naturally (Statement 26):
Conversely, a positive relationship was found between age and the importance of code-switching naturally (Coefficient = 0.139, Standard Error = 0.072, p < 0.05). This suggests that older respondents place greater importance on codeswitching naturally than younger respondents. The positive coefficient indicates that as age increases, so does the value placed on natural code-switching. This may reflect a more structured approach to bilingualism among older individuals, possibly linked to traditional language norms or cultural expectations.
D. Relationship Between Educational Background and Code-Switching
The regression model examined code-switching in the following contexts (Table 6):
* Code-switching with Family (Statement 15):
A negative relationship was observed between educational background and the frequency of code-switching with family (Coefficient = -0.126, Standard Error = 0.052, p < 0.05). This suggests that respondents with higher educational backgrounds are less likely to code-switch when interacting with family members. The negative coefficient indicates that the frequency of code-switching in family settings decreases as education level increases. This may reflect a stronger preference for language consistency or the maintenance of a heritage language in familial contexts among more educated individuals.
* Code-switching at School or Work (Statement 16):
A positive relationship was found between educational background and code-switching at school or work (Coefficient = 0.144, Standard Error = 0.047, р < 0.05). This indicates that respondents with higher educational levels are more likely to engage in code-switching in formal environments. The positive coefficient suggests that the frequency of codeswitching in academic and professional settings increases as education level increases. This may reflect greater bilingual proficiency or pragmatic adaptability, allowing educated individuals to navigate multilingual environments more effectively.
E. Relationship Between Educational Background and Ease of Understanding Code-Switching
The regression model examined the ease of understanding code-switching in the following context (Table 7):
* Ease of Understanding Code-switching in Written Text (Statement 23):
A positive relationship was observed between educational background and the ease of understanding code-switching in written text (Coefficient = 0.186, Standard Error = 0.067, p < 0.05). This suggests that respondents with higher educational levels find it easier to comprehend code-switching in written communication. The positive coefficient indicates that as the level of education increases, so does the ease of understanding bilingual or multilingual text. This may reflect advanced literacy skills, greater exposure to diverse linguistic inputs, or enhanced cognitive processing abilities associated with higher education levels.
F. Relationship Between Educational Background and Attitudes Towards Code-Switching
The regression model examined attitudes towards code-switching in the following context (Table 8):
* Importance of Code-switching Naturally (Statement 26):
A positive relationship was observed between educational background and the importance of code-switching naturally (Coefficient = 0.149, Standard Error = 0.059, p < 0.05). This suggests that respondents with higher educational levels place greater importance on code-switching in a natural and seamless manner. The positive coefficient indicates that as the level of education increases, so does the value placed on natural language switching. This pattern may reflect a heightened awareness of bilingual identity or social norms associated with language proficiency among more educated individuals.
G. Relationship Between Gender and Code-Switching
The logistic regression model examined code-switching in the following contexts (Table 9):
* Code-switching at School or Work (Statement 16):
A positive relationship was observed between gender and the frequency of code-switching at school or work. This suggests that female respondents are 14% more likely to code-switch at school or work than male respondents, as indicated by the positive change in odds. This pattern may reflect gendered communication norms or greater social adaptability in educational and professional settings.
* Code-switching When Talking to Oneself (Statement 18):
A negative relationship was found between gender and code-switching when talking to oneself. Female respondents are 16% less likely to code-switch when talking to themselves than male respondents. This finding could indicate differences in cognitive processing or socialisation patterns influencing internal language use.
* Code-switching When Counting (Statement 19):
A significant negative relationship was also identified for code-switching when counting. Female respondents are 20% less likely to code-switch when counting compared to male respondents. This may reflect a stronger attachment to a dominant language for numerical cognition or differing educational experiences.
H. Relationship Between Gender and Ease of Code-Switching
The logistic regression model examined the ease of code-switching in the following contexts (Table 10):
* Ease of Code-switching While Speaking (Statement 20):
A negative relationship was observed between gender and the ease of code-switching while speaking. This suggests that female respondents are 13% less likely to find it easy to code-switch while speaking compared to male respondents, as indicated by the negative change in odds. This pattern may reflect differences in social norms or communication styles influencing spoken language use.
* Ease of Understanding Code-switching in Spoken Conversation (21):
A positive relationship was found between gender and the ease of understanding code-switching in spoken conversation. Female respondents are 50% more likely to find it easy to understand when others code-switch in spoken conversation compared to male respondents. This may indicate greater linguistic sensitivity or social awareness among female respondents, possibly due to socialisation patterns, emphasising attentive listening and contextual interpretation.
I. Relationship Between Gender and Attitudes Towards Code-Switching
The logistic regression model examined attitudes towards code-switching in the following contexts (Table 11):
* Importance of Code-switching Naturally (Statement 26):
A negative relationship was observed between gender and the importance of code-switching in a natural way. This suggests that female respondents are 18% less likely than male respondents to consider it important to code-switch in a natural and seamless manner. This could reflect gendered socialisation patterns where men might place more emphasis on maintaining a natural flow in bilingual interactions, possibly due to differing communication expectations or identity expressions.
* Desire to be Perceived as Code-switching Naturally (Statement 27):
A positive relationship was found between gender and the desire to be perceived as code-switching naturally. Female respondents are 15% more likely to care about being perceived as natural in their code-switching compared to male respondents. This pattern could reflect heightened social sensitivity or a greater awareness of linguistic norms among female respondents.
J. Summary of Results
The results indicate significant age-related differences in code-switching patterns. Younger respondents were more likely to code-switch at school or work and when talking to themselves, reflecting a generational shifttowards more fluid bilingualism. In contrast, older respondents tended to code-switch when counting, possibly due to cognitive reliance on a dominant language for numerical tasks. Younger individuals found it easier to code-switch while writing, whereas older respondents showed greater ease in understanding bilingual text. Additionally, younger participants associated codeswitching more with authentic self-expression, while older respondents placed greater importance on maintaining naturalness in bilingual communication.
Educational background also played a key role in shaping code-switching behaviour. Respondents with higher educational attainment were less likely to code-switch within family settings but more likely to do so in formal environments such as school or work, suggesting enhanced bilingual proficiency and pragmatic adaptability. Higher education was associated with a better understanding of bilingual text and a greater emphasis on natural code-switching, reflecting advanced literacy skills and heightened awareness of social norms. These findings suggest that educational experiences significantly influence bilingual communication strategies and identity expressions.
Gender differences were evident in both the frequency and ease of code-switching. Female respondents were more likely to code-switch at school or work but less likely to do so when talking to themselves or counting, highlighting gender-specific cognitive and social patterns. Women found it more challenging to code-switch while speaking but were more adept at understanding bilingual conversation, suggesting greater social awareness and contextual interpretation. Additionally, men placed more emphasis on maintaining a natural flow in bilingual interactions, while women showed a stronger desire to be perceived as natural in their code-switching. These gendered patterns underscore the complex social dynamics influencing bilingual communication within the Armenian community in Kuwait.
K. Discussion
This study examined code-switching practices, perceived ease, and attitudes among members of the Armenian community in Kuwait, with a focus on how these behaviours vary in relation to age, gender, and education. The findings contribute to the body of sociolinguistic research on bilingualism in diasporic settings, especially those shaped by complex historical and linguistic trajectories.
Age was found to influence all three dimensions of code-switching. Younger respondents reported greater ease and frequency of switching in contexts such as writing, speaking, and self-directed speech, consistent with studies showing younger bilinguals as more flexible and context-adaptive (Dewaele & Wei, 2014). In contrast, older participants valued naturalness in switching and reported greater ease in comprehending bilingual written texts. This aligns with findings from bilingual literacy research suggesting that older individuals often rely on metalinguistic awareness shaped by traditional schooling. However, some studies, such as Zentella (2017), suggest that older bilinguals in diaspora communities may be more frequent switchers due to greater immersion in heritage language networks, a finding not reflected in the current data.
Gender differences were also observed. Female participants were more concerned with how naturally they were perceived when switching languages, while male participants were more likely to identify with multilingual communities. These observations are broadly consistent with sociolinguistic work indicating that gender can mediate language attitudes and the performance of bilingual identity (Pavlenko & Blackledge, 2004). However, unlike Holmes (2006), who found that women in multilingual societies tend to switch more often than men to manage interpersonal relationships, our data do not show higher switching frequency among female respondents.
Educational background was positively associated with switching in formal contexts (e.g., at work or school) and with the perceived ease of understanding multilingual texts. These results support studies linking higher education to broader communicative repertoires and greater control over multilingual practices (Blommaert & Rampton, 2011). However, the finding that more educated participants reported lower switching frequency with family diverges from research by García and Wei (2014), which suggests that educated bilinguals often carry over their multilingual competence into home settings as part of translingual identity construction.
Taken together, the findings suggest that both individual characteristics and broader social conditions shape codeswitching among Armenians in Kuwait. While many patterns align with the literature on bilingual identity and language socialisation, some divergences-particularly those relating to age and family use-point to the importance of contextspecific language ideologies and community norms. Future research might explore these findings further through ethnographic methods or cross-comparative designs with other diaspora groups.
L. Limitations
This study presents certain limitations that should be acknowledged. Firstly, the exclusive use of quantitative methods restricts the depth of insights into participants' personal experiences and interpretations of code-switching. Employing a mixed-methods design, incorporating qualitative approaches such as interviews or observations, would have provided richer, more nuanced perspectives on why certain demographic groups reported particular code-switching behaviours or attitudes. Additionally, the convenience sampling method utilised may limit the generalisability of the findings, as the respondents might not fully represent the entire Armenian community in Kuwait. Self-report data collection through surveys can also introduce biases, such as social desirability or recall inaccuracies, potentially affecting the reliability of responses. These limitations suggest the need for future research to include qualitative methods and more representative sampling strategies to deepen the understanding of bilingual practices within diasporic communities.
V. CONCLUSION AND RECOMMENDATIONS
This study examined code-switching behaviours, ease, and attitudes within the Armenian community in Kuwait, focusing on how these dimensions are shaped by age, gender, and educational background. Drawing on data from 225 respondents and guided by the Bilingual Code-Switching Profile framework, the study revealed that demographic variables influence how individuals perceive and practise code-switching across various social and cognitive contexts.
Age emerged as a key factor, with younger participants reporting greater frequency and comfort with switching in informal and self-directed settings, while older respondents emphasised comprehension and naturalness. Gender was associated with different orientations towards social perception and group affiliation, and education correlated with increased switching in formal domains and greater ease in processing multilingual texts. These findings suggest that codeswitching is not a uniform behaviour but one conditioned by personal experience, social roles, and context-specific communicative norms.
By highlighting these patterns in a diasporic, multilingual community, the study adds to the understanding of bilingual language use in under-researched sociolinguistic environments such as the Arab Gulf. It also demonstrates the relevance of linking demographic factors to language practices in communities where multilingualism plays a central role in identity and interaction. Further research is encouraged to explore how such patterns evolve over time and across different generations or diaspora settings.
Given the limitations of this study, future research should adopt a mixed-methods approach, integrating qualitative tools such as interviews or ethnographic observations to explore the motivations, functions, and personal meanings of code-switching more deeply. Comparative studies across different Armenian diaspora communities-both within and beyond the Arab Gulf-could also provide insight into the influence of local sociolinguistic ecologies. Additionally, educational stakeholders and community organisations may benefit from greater awareness of how generational and educational differences shape bilingual practices, which could inform culturally responsive language policies and heritage language maintenance efforts.
REFERENCES
[1] Abdel Tawwab, A. (2014). Socio linguistic study of code-switching of the Arabic language speakers on social networking. International Journal of English Linguistics, 4(6), 78-86. Retrieved March 25, 2025, from http://dx.doi.org/10.5539/ijel.v4n6p78
[2] Ahmed, K., Ali, I., Xianfa, D.H. (2015). Code‐Mixing as a Marker of Gender Identity in SMS language in Pakistan. Journal of Humanities and Social Science, 20, 58-65. DOI: 10.9790/0837-20155865
[3] Aldalbahy, K. K. (2022). The Relationship between Code-switching and the Level of Language Proficiency in the Second Language of Saudi Female Students. Arab World English Journal, 13(4), 309-324. Retrieved March 29, 2025, from https://dx.doi.org/10.24093/awej/vol13no4.20
[4] Alenezi, M. Q. (2016). Gender and Students' Attitude toward Code-Switching: A Correlational Study with Reference to Saudi Arabian Medical Students at Northern Boarder University. International Journal of Englis Language & Translation Studies, 4(3), 154-166.
[5] Al-Qaysi, N. (2019). Students and educator's attitudes towards code-switching: A longitudinal study. International Journal of Information Technology and Language Studies, 3(2), 61-72.
[6] AlRumaihi, A. (2021). Code-Switching Among Kuwaitis in the Social Context: Attitudes and Practices. English Language and Literature Studies, 11(2), 42-53. Retrieved March 26, 2025, from https://doi.org/10.5539/ells.v11n2p42
[7] Aprahamian, A. (1967). Hamarod Urvakits Hay Kaghtavayreri Badmoutyoun (Concise Outline of the History of Armenian Colonies) (Vol. 2). Yerevan.
[8] Baker, C. (2003). Foundations of bilingual education and bilingualism (3rd ed.). Multilingual Matters.
[9] Bautista, M. L. S. (2004). Tagalog-English code-switching as a mode of discourse. In A. Butler, & J. Y. T. Zuo (Eds.), Asian Englishes: Innovation, diffusion and identity (pp. 165-180). John Benjamins Publishing Company.
[10] Blommaert, J., & Rampton, B. (2011). Language and superdiversity. In J. Blommaert, B. Rampton, & M. Spotti (Eds.) Diversities, 13(2), 1-21. Retrieved April 14, 2025, from https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=4406ba46a37074ea22a030d8a74f98f1e810570d
[11] Bournoutian, G. A. (2003). A Concise History of the Armenian People. Mazda Publishers.
[12] Bullock, B. E., & Toribio, A. J. (Eds.). (2009). The Cambridge handbook of linguistic code-switching. Cambridge University Press.
[13] Chan, A. S. L. (2023). The Diaspora of Bilinguals: Code-Switching in Three Groups of Highly Proficient Cantonese-English Bilinguals. [Unpublished doctoral dissertation]. University of California.
[14] Cheshire, J., & Gardner-Chloros, P. (1998). Code-switching and the sociolinguistic gender pattern. International Journal of the Sociology of Language, 129, 5-34. Retrieved March 4, 2025, from https://doi.org/10.1515/ijsl.1998.129.5
[15] Dashti, A. (2015). The role and status of the English language in Kuwait: How is English used as an additional language in the Middle East? English Today, 31(3), 28-33. Retrieved March 25, 2025, from https://doi.org/10.1017/S026607841500022X
[16] David, M. K., Yee, W. M. T., Ngeow, Y. M., & Hui, G. K. (2009). Language choice and code switching of the elderly and the youth. International Journal of the Sociology of Language, 200, 49-74. DOI:10.1515/IJSL.2009.044
[17] Dewaele, J.M., & Wei, L. (2014). Attitudes towards code-switching among adult mono-and multilingual language users. Journal of Multilingual and Multicultural Development, 35(3), 235-251. Retrieved March 25, 2025, from http://doi.org/10.1080/01434632.2013.859687
[18] Garcı́a, O. and Wei, L. (2014). Translanguaging: Language, bilingualism and education. Palgrave. Retrieved April 14, 2025, from https://doi.org/10.1057/9781137385765
[19] Gardner-Chloros, P. (1991). Language selection and switching in Strasbourg. Clarendon Press.
[20] Gosselin, L. and Sabourin, L. (2023). Language athletes: dual-language code-switchers exhibit inhibitory control advantages. Frontiers in Psychology, 14. Retrieved March 4, 2025, from https://doi.org/10.3389/fpsyg.2023.1150159
[21] Holmes, J. (2006). Gendered talk at work: Constructing social identity through workplace interaction. Wiley.
[22] Hughes, C.E., Shaunessy, E.S., Brice, A.R., Ratcliff, M.A., & McHatton, P.A. (2006). Code switching among bilingual and limited English proficient students: Possible indicators of giftedness. Journal for the Education of the Gifted, 30(1), 7-28. Retrieved March 25, 2025, from https://doi.org/10.1177/016235320603000102
[23] Hukara, E. (2022). Code-Switching in Social Media - Generational Differences. [Unpublished final paper]. University of Sarajevo.
[24] Jebejian, A. (2004). Code-switching Among Armenians in Lebanon. Changing Agendas in Educational Research: National and International Perspectives, 1, 101-126. University of Leicester.
[25] Jebejian, A. (2007). Changing ideologies and extralinguistic determinants in language maintenance and shiftamong ethnic diaspora Armenians in Beirut. [Unpublished doctoral dissertation]. University of Leicester.
[26] Jebejian, A. (2022). Intergenerational Construction of Identity among the Armenian Community in Kuwait. Armenian Diaspora in Dynamics, 1, 117-143.
[27] Mahsain, F. (2014). Motivations Behind Code-switching Among Kuwaiti Bilingual Schools. [Unpublished doctoral dissertation]. University of Manchester.
[28] McCrindle, M. (2018). Understanding Generation Z. McCrindle Research Centre.
[29] Murtiningsih, S., Munawaroh, M., & Hidayatulloh, S. (2022). Code-switching in EFL classrooms: factors influencing teachers to use code-switching and its types used in the classrooms. Journal on English as a Foreign Language, 12(2), 318-338. DOI: 10.23971/jefl.v12i2.3941
[30] Olson, D. J. (2022). The bilingual code-switching profile (BCSP): Assessing the reliability and validity of the BCSP questionnaire. Linguistic Approaches to Bilingualism. 1-29. Retrieved November 6, 2024, from https://doi.org/10.1075/lab.21039.ols
[31] Omar, A., & Ilyas, M. (2018). The Sociolinguistic Significance of the Attitudes towards Code-Switching in Saudi Arabia Academia. International Journal of English Linguistics, 8(3), 79-91. Retrieved March 17, 2025, from http://doi.org/10.5539/ijel.v8n3p79
[32] Parama, K. S., Kreiner, D. S., Stark, K. S. & Schuetz, S. A. (2017). Monolingual and Bilingual Perceptions of Code-Switching: A Difference in Cognition but Not Competence. North American Journal of Psychology, 19(1), 87-101.
[33] Pattie, S. (2005). "Armenians in Diaspora." In E. Herzig, & M. Kurkchyan (eds.). Past and Present in the Making of National Identity (126-146). New York: Routledge.
[34] Pauwels, A. (1995). Immigrant women and language maintenance in Australia: The case of Vietnamese. Journal of Multilingual and Multicultural Development, 16(1-2), 17-32.
[35] Pavlenko, A., & Blackledge, A. (Eds.). (2004). Negotiation of identities in multilingual contexts. Multilingual Matters.
[36] Villarin, S. J., & Emperador, E. A. (2023). Factors, Forms, and Functions of Code Switching on English Achievement among Grade 11 Students. International Journal of Science and Management studies, 6(2), 99-109. DOI: 10.51386/25815946/ijsmsv6i2p114
[37] Vosgemadyan: Armenian National School of Kuwait 1960-2010. (2011). Kuwait Armenian School Council.
[38] Wong, K. L. J. (2006). Gender and Codemixing in Hong Kong. The University of Sydney.
[39] worldpopulationreview.com/countries/kuwait (4,973,670 million)
[40] Yim, O., & Clément, R. (2021). Acculturation and attitudes toward code-switching: A bidimensional framework. The International Journal of Bilingualism, 25(5), 1369-1388. Retrieved March 16, 2025, from https://doi.org/10.1177/13670069211019466
[41] Zentella, A.C. (2017). "Dime con quién hablas, y te diré quién eres": Linguistic (in)security and Latina/o Unity. In J. Flores & R. Rosaldo (Eds.) A Companion to Latina/o Studies, pp. 25-38. Retrieved April 14, 2025, from https://doi.org/10.1002/9781405177603.ch3
[42] Zhong, Z. and Fan, L. (2023). Worldwide trend analysis of psycholinguistic research on code switching using bibliometrix r - tool. Sage Open, 13(4). Retrieved April 12, 2025, from https://doi.org/10.1177/21582440231211657
Copyright Academy Publication Co., Ltd. 2025