How do Saudi EFL learners realize the speech act of request?

Abstract The current study aimed to examine the realization of the speech act of request by Saudi EFL learners, in Arabic as their first language and English as their foreign language. To this end, 240 Saudi undergraduates completed a discourse completion task consisting of 12 situations, six of which were presented in Arabic while the other six were presented in English. The participants’ responses were coded for the request strategies in addition to internal/external modifiers and initiators. The results showed a clear dominance for the conventionally indirect strategy of “query preparatory” in both languages. However, other statistically significant differences emerged. For instance, while the participants used more direct requests in Arabic than in English, they employed more nonconventionally indirect requests in English than in Arabic. Additionally, the participants apologized significantly more often in English than in Arabic whereas they promised and offered prayers to the hearer more frequently in Arabic than in English. Moreover, the participants used a significantly higher number of downgraders in English than in Arabic. Regarding social variables, the participants exhibited sensitivity to social dominance in the two languages.


Introduction
In a world where bilingualism has become the norm, the realization of speech acts by bilinguals has become an increasingly intriguing topic (e.g., El-Dakhs, 2017, 2018;Huang & Lu, 2023;On & Meir, 2022).For language learners, performing speech acts appropriately could be a challenging task because the realization of speech acts is culture-specific (Cheng & Lam, 2020;Chentsova-Dutton and Vaughn, 2012).That is, what may fit in one culture may not be acceptable in another culture.For example, speech acts can be carried out differently, can attract different kinds of responses and can vary in directness across languages and cultures (e.g., Stoian, 2020).Hence, bilinguals need to carefully choose the right strategies to realize speech acts in their different languages.
There are several types of speech acts.One popular classification was introduced by Searle (1979) who classified speech acts into assertives which refer to statements that express the speaker's belief about the truth of a proposition, directives which represent attempts by the speaker to get the hearer to do something, commissives which commit the speaker to some future action, expressives which express the speaker's feelings and emotions and declarations which are acts that change the reality in accord with the proposition of the declaration.Under this classification falls a multitude of speech acts, including the speech act of request which is the main focus of the current study.According to Searle (1979), requests, which can be defined as "an attempt on the part of the speaker to get the hearer to perform or to stop performing some kind of action" (Ellis, 1994, p. 167), fall under the category of "directives." The current study aims to explore how bilinguals realize the speech act of request in their different languages.In particular, our study will examine how Saudi learners of English as a foreign language (EFL) realize the speech act of request in their first language (L1), Arabic, and foreign language (FL), English.Our work is motivated by the special difficulty EFL learners face when they try to express themselves appropriately in their FL.This difficulty often stems from the learners' lack of sufficient pragmatic knowledge, whether in the form of "functional knowledge," which involves the use of linguistic forms to realize pragmatic functions, or "sociolinguistic knowledge," which requires the appropriate use of linguistic forms according to relevant situational variables, such as social distance, social dominance and familiarity of interlocutors (Backman & Palmer, 2010).More specifically, our current study addresses the following research questions: (1) What type of requests do Saudi EFL learners use in their L1 Arabic?
(2) What type of requests do Saudi EFL learners use in their L2 English?
(3) Are there significant differences between the requests produced by Saudi EFL learners in Arabic and English?
(4) Does social dominance have a significant influence on the production of requests by Saudi EFL learners in Arabic and English?
To situate the current study, the following sections will survey the relevant theoretical models and earlier studies on the speech act of request.This will be followed by listing the research questions and explaining the research methodology.Then, the results will be presented and discussed.Finally, relevant conclusions, including directions for future research, will be drawn.

Theoretical Framework
Three models are relevant to the study of the speech act of request.The first model is Brown and Levinson's (1978;1987), politeness theory which views the speech act of request as a facethreatening act to the speaker's and hearer's face.According to this model, interlocutors cooperate to maintain one another's public image (i.e., face) through avoiding the use of facethreatening acts or mitigating their negative influence.In this respect, face can be positive (the desire to be liked and admired) or negative (the desire to be independent and free from imposition).The speech act of request is threatening to both the speaker and the hearer since it involves intrusion and imposition.Thus, it is predicted that speakers will attempt to soften the influence of requests by using more indirect strategies or employing several modifiers.It is also predicted that the realization of speech acts will be modified based on relevant situational factors, including social distance and dominance.However, it must be noted that the politeness theory has been criticized for not accounting sufficiently for cultural differences.The model has been accused of being Eurocentric (Cook, 2022).This means that it is more prone to the Western civilization in which individualism, and thus independence, is highly valued.In fact, different cultures may react differently to speech acts (Spencer-Oatey, 2008).In the current study, Brown andLevinson's (1978, 1987), politeness theory will be used in the discussion section to explain and comment on the types of request strategies employed by Saudi EFL learners in their two languages (i.e., Arabic and English).
The second relevant model is the individualism-collectivism index by Hofstede (2001).The model proposes that developed and Western countries are generally dominated by concepts of individualism, including freedom of imposition, while less developed and Eastern countries are more characterized by collectivism in which solidarity relationships among the community members are prioritized.Since English is mainly spoken in Western countries and Arabic in Eastern countries, it is expected that the participants in the current study will employ more indirect request strategies and mitigation devices in English than in Arabic if they were pragmatically competent because English-speaking native speakers prefer to avoid imposition.Using the same features and modifications across the two languages would reveal L1 interference.It must be noted that the additional concern for independence in English-speaking versus Arabic-speaking countries was proved valid in several earlier studies that compared the realization of speech acts across the two languages (e.g., El-Dakhs & Ahmed, 2021;Hosni, 2020).In the current study, Hofstede's (2001) individualism-collectivism index will be used in the discussion section to explain and comment on the Saudi EFL learners' realization of the speech act of requests in Arabic and English.The model will help explain the choices the participants will make among the available realization strategies of the speech act of request.
The third relevant model was drawn by Blum-Kulka and Olshtain (1984) in their seminal project named the Cross-Cultural Speech Act Realization Project (CCSARP).The project represented a major empirical study that compared the realization of the speech acts of requests and apologies across different languages/dialects (Australian English, American English, British English, Canadian French, Danish, German, Hebrew, and Russian).The study compared the realization of these speech acts by both native and non-native speakers, and proposed important classifications.Of relevance to requests, the model consists of three levels of directness and nine types of strategies.The most explicit level is named the "direct level" and is marked syntactically through the use of imperatives, performatives and hedged performatives.The less explicit level is named "conventionally indirect" and requires the use of specific formulas and frequent patterns, such as "can you . . .", "how about . . ..?" and "why don't you . . .?"The least explicit level, which is referred to as "nonconventionally indirect," involves strategies that require the hearer to interpret the request, such as strong hints.In the current study, this classification of the realization strategies of the speech act of request will be used in the data coding stage to analyze the types of strategies used to make requests by Saudi EFL learners.See Appendix A for the complete model along with relevant examples.
An example of cross-cultural studies is Yazdanfar and Bonyadi (2016) who conducted a comparative study of the request speech act in Persian and English.To this end, the requestive utterances in selected English and Persian TV series were transcribed and categorized as per their level of directness.The results showed that Americans' and Persians' requests were generally characterized with directness.However, it was also noted that the Americans produced significantly more indirect strategies than the Persians while the Persians showed a significantly stronger tendency to use direct strategies.In the same vein, Chintawidy and Sartini (2022) compared the request strategies of two groups of Indonesia's ethnic groups, i.e., Javanese and Sudanese.The Discourse Completion Task (DCT) of 60 participants revealed that the strategies of "mood derivable" and "query preparatory" were the most frequently used.In fact, the request strategies used by both groups were similar and reflected the Indonesian culture.
In the Arab World, two interesting cross-cultural studies are Alshammari (2015) and AlMujaibel and Gomaa (2022).Alshammari (2015) compared the directness and indirectness of the speech act of request among American native speakers of English and Saudi native speakers of Arabic.The results revealed that Americans exhibited a significantly stronger tendency than Saudis to use indirect requests, particularly when addressing unfamiliar friends and when inferiors address their superiors.AlMujaibel and Gomaa (2022) compared the realization of the speech act of request in Kuwaiti Arabic and British English.Using a mixed methods approach, data were collected from 500 participants.The results showed that the strategy of "query preparatory" was the most frequently used by both the Kuwaiti and the British participants.Furthermore, power as a social variable affected the realization of requests by both groups.The participants preferred direct request strategies when the power of the speaker was high.
In addition to the cross-cultural studies, several interlingual studies were conducted.For example, Montenegro (2017) explored the most appropriate types of English requests by Costa Rican EFL learners.The results showed that questions including modal auxiliaries, such as "would you mind" or "could you", are thought as more polite than the other forms.The students also correctly judged the level of appropriateness of other structures, including expressions of wish, interrogatives and imperatives.The results thus showed that EFL learners can develop an adequate level of pragmatic competence based on foreign language exposure.Likewise, Akmal et al. (2022) investigated the differences in the realization of request speech act between the Indonesian English learners and the Australian English native speakers.Data were collected through DCTs and observation.The results showed that the two groups used direct and indirect requests.However, the EFL Learners tended to use more interrogatives while the native speakers preferred to use more declarative sentences.Additionally, EFL learners tended to use more attractors and honorifics than the native speakers.
In the Arab context, a number of studies examined the speech act of request (e.g., Alfghe & Mohammadzadeh, 2021;Deveci & Ben Hmida, 2017;Mohamed, 2019), but not based on the level of directness as is the case in the current study.Of relevance to the present study are Qari (2017), Abdelfattah (2022) and Qari (2021).Qari (2017) investigated the production of requests by Saudi Arabic native speakers, Saudi EFL learners and British native speakers.Data were collected from 160 participants using DCTs.The results showed that Saudis tended to use more direct requests than the EFL learners and the British group.However, the latter two groups showed a stronger preference for indirect strategies.As for mitigation, while Saudis preferred to use semantic softeners, the British preferred the use of syntactic and linguistic devices.In the same vein, Abdelfattah (2022) investigated the production of requests by Moroccan EFL learners and American native speakers of English.Data were again collected through DCTs from 60 participants.The results revealed that the Moroccans used more direct request strategies than the Americans.However, the two groups preferred to use the conventionally indirect request strategies over the other strategies.This was particularly the case with the frequent use of the query preparatory strategy.
As for Qari (2021), her study was instructional in nature.Qari (2021) examined whether explicit instruction of requests could improve Saudi EFL learners' knowledge and use of the appropriate request strategies in English.To this end, 30 Saudi EFL learners were explicitly trained on the use of request strategies in English and data were collected using pre-test and post-test written questionnaires.The results showed that explicit instruction led to a significant improvement in the learners' understanding and use of the request strategies in English since students tended to employ more appropriate strategies in their production and exhibited significant progress in recognizing the functions of the relevant strategies.This finding comes in line with earlier research on instructional pragmatics that supports the role of explicit instruction in improving EFL learners' pragmatic knowledge (Roever, 2022).
Based on the above studies, several observations can be made.First, the strategy of "query preparatory" seems to be commonly used among participants as is the case in Chintawidy and Sartini (2022) and AlMujaibel and Gomaa (2022).Second, the level of directness of requests seems to greatly vary across cultures.This is particularly the case when Western cultures are compared with Eastern cultures.Prior studies (i.e., Abdelfattah, 2022;Alshammari, 2015;Qari, 2017;Yazdanfar & Bonyadi, 2016) have showed that Westerners prefer to realize the speech act of request in a more indirect manner than Easterners, probably because they prefer to avoid intrusion and imposition.Third, social variables may influence the realization of the speech act of request as was the case in AlMujaibel and Gomaa (2022).Fourth, explicit instruction of speech acts seems to be effective for language learners (Roever, 2022).This is particularly the case for EFL learners who receive minimal exposure to the authentic use of the target language (Qari, 2021).Finally, several studies have examined the realization of the speech act of requests globally and in the Arab World.However, further research is still needed on how bilinguals realize the speech act of request with different levels of directness among Arab speakers.This is particularly the case with Saudi EFL learners when they express themselves in their L1 and FL.The current study will address this gap through shedding light on the production of the speech act of request by Saudi EFL learners.
In this context, the current study is significant for three reasons.First, it contributes to research on how bilinguals realize speech acts in their two languages.This is particularly significant because it is important for EFL learners to select the right linguistic forms to realize the target speech acts while ensuring that they are accommodating the requirements of the cultural context and the situational variables.Second, our study compares the realization of the speech act of request in Arabic versus English, which are two distant languages that vary typographically, etymologically and culturally.Hence, the findings will show how bilinguals speaking two distant languages manage their pragmatic choices.Finally, our study contributes to the research conducted on the Arabic language which is still considered a relatively underrepresented language in the pragmatic literature.

Participants
A total of 240 undergraduates from a private Saudi university were randomly selected to voluntarily participate in the current study.They were freshmen at non-English majors, including Law, Business Administration, Computer Sciences and Engineering.They were all native speakers of Najdi Arabic (i.e., the Arabic dialect spoken in Riyadh and central Saudi Arabia) and foreign learners of English as they had studied English at school prior to university admission and also pursued their English study at university.Half the participants were male and the other half were female, and they were aged between 18 and 22.It must be noted that the number of participants was found appropriate based on earlier studies that used DCTs for data collection (e.g., Fareh et al., 2023;Karimkhanlooei & Vaezi (2017); Moafian et al., (2022) Regarding language exposure, two important factors are relevant.First, the university in which the study took place is characterized by its multicultural teaching staff.The faculty members come from different parts of the world and students need to communicate with them in English inside the classroom since English is the medium of instruction.The students also need to speak with non-Arabic speaking faculty members in English outside the classroom.Second, all the participants were enrolled in a Preparatory Year Program (PYP) to improve their English before studying their majors.They were placed at the third level of a three-level program.This level is estimated at B1 level of the Common European Framework.

Instrument
A discourse completion task (DCT) (see Appendix B) was used to collect data in the current study.The DCT consisted of 12 situations, half of which were in Arabic and the other half were in English.The situations were selected based on four focus groups with 30 students to identify the regular situations the students make requests to the target addressees.The aim of the DCT was to compare the participants' realization of the speech act of request in Arabic as their L1 and English as their FL.Each situation included a description of a scenario and the participant was asked to respond to the situation as he/she would in real life.Thus, the participants responded to half the situations in Najdi Arabic and to the other half in English.It must be noted that the discourse completion task was reviewed by five faculty members at the target university who are specialized in Language Studies prior to its administration to ensure the validity of the task.The task was also piloted with a group of 20 students who performed the task well and did not face any comprehension difficulty.
A number of variables were considered while developing the DCT.First, the relation between the participant and the addressee in all the situations was constantly distant.The participants were never asked to communicate with an intimate relation.It would have been unnatural to ask our Saudi participants to speak in English to their parents or friends, for example.However, it was natural for them to speak to their professors, classmates and janitors in English because their university is multicultural with students and employees (both professors and janitors) from several nationalities.Additionally, English is the medium of instruction, and hence students are stipulated to speak in English in class all the time.Second, the situations varied in relation to the variable of social dominance.While the addressee was in a higher position in two situations (i.e., professor), he/she was in an equal position in two situations (i.e., classmate) and in a lower position in two situations (i.e., janitor).This variation allowed us to examine the influence of social dominance on the realization of requests.Finally, the addressee's gender was female for female participants and male for male participants.This is to simulate the situation in Saudi Arabia in which most undergraduates still study in a separate campus for female students and a separate campus for male students.
We chose to use DCTs for data collection to exercise control over the study variables, to collect data from a large sample and to allow for comparisons among participants and across studies (Turnbul, 2001).It must be noted that in the case of second/foreign language research, as is the case in the current study, DCTs are generally viewed as a type of language tests that are used to sample the participants' competence in performing certain tasks (e.g., Labben 2016;Dörnyei 2003).In fact, DCTs are widely used in the literature and Labben (2016) predicts that it will remain the most used instrument to collect data, especially in interlingual pragmatics.However, we need to acknowledge that DCTs are sometimes criticized in the pragmatic literature because the participants' responses in DCTs may represent their intuitions for how they should speak, rather than what they should actually say (Beebe and Cummings, 1996;Turnbul, 2001).

Data collection
Upon receiving the ethical clearance from the Institutional Review Board at the private university from which the data were collected, the researchers sought permission from the university authorities to visit classes and ask the students to complete the DCTs.In every class visit, the researchers explained the purpose and procedure of the study to the participants.Those who agreed to participate completed an informed consent form and the DCT while those who objected were given the freedom to use this time as per their convenience.No time limit was set for the completion of the DCT, but it was completed in 10-15 minutes.Meanwhile, the students were encouraged to seek clarification from the researchers as needed.

Data coding
The data were coded for four components: (1) the head act of requests, (2) external modifiers, (3) internal modifiers and (4) initiators.The head act of requests was coded based on Blum-Kulka and Olshtain's model (Appendix A).According to the model, direct strategies included "mood derivable" (e.g., Give me the remote), "performative" (e.g., I'm asking you to give me the remote), "hedged performative" (e.g., I would like to ask you for the remote), "obligation statement" (e.g., You will have to give me the remote) and "want statement" (e.g., I would like you to give me the remote).The indirect strategies were classified into two levels.The first level, the conventionally indirect level, included "suggestive formula" (e.g., Why don't you hand me the remote?" and "query preparatory" (e.g., Could you give me the remote?).The second level of indirectness, the nonconventionally indirect level, comprised of "strong hints" (e.g., My favourite show will begin soon.)and "mild hints" (e.g., It has been a long time since I watched my favourite show).
As for the modifiers, six external types emerged from the data as shown in Appendix C. The participants gave reasons (e.g., because I want to watch the TV), made promises (e.g., I will give it back soon), offered apologies (e.g., Sorry to bother you.), provided thanks (e.g., Thank you!), posed questions (e.g., What is the time now?) and, only in Arabic cited a prayer for the hearer (e.g., May God always make you happy!).Regarding the internal modifiers, the data were analyzed as per Trosborg's (1995) model (see Appendix D).The model consists of downgraders, such as polite markers (e.g., Please) and hedges (e.g., kind of) and upgraders, including intensifiers (e.g., very) and plus committors (e.g., of course).The data were also quoted for a few initiators that emerged in the coding process.These included greetings (e.g., Hello!), address terms (Mr.), terms of endearment (e.g., dear), names (e.g., Ahmed) and attention grabbers (e.g., Look!).
Here are some examples of coded data:

Results
This section is divided as per the research questions.

What type of requests do Saudi EFL learners use in their L1 Arabic?
As shown in Table 1, the majority of requests in Arabic fall under the category of "conventionally indirect requests" which represents 71.7% of the total requests.This category covers the strategy of "suggestive formula" (e.g., Why don't you give me your charger?), which represents 15.4%, and the strategy of "query preparatory" (e.g., Could you postpone the exam?), which represents 56.3%.The following category in terms of frequency of use is the "nonconventionally indirect requests" which stands for 15.6% of total frequency of use.This category is represented solely by the strategy of "strong hints" (e.g., I have a mid-term exam tomorrow.).Finally, "direct requests" stands for 12% of the total requests, represented by "mood derivable" (e.g., Give me the charger) (7.7%) and "want statement" (e.g., I would like you to postpone the exam.)(3.9%).

What type of requests do Saudi EFL learners use in their L2 English?
As shown in Table 5, the participants' requests in English exhibited different frequencies of use.The participants' requests were dominated by the conventionally indirect requests, which represented 69.7% of the total requests.The conventionally indirect strategies stood at 1.2% and 68.5% respectively for "suggestive formulas" (e.g., How about postponing the exam?) and "query preparatory" (e.g., Can you give me a cup?).The second category of requests was "noncovenetionally indirect requests," which stood at 24.5%.Nonconventionlly indirect requests were represented by "strong hints" (e.g., I cannot drink like that).The direct request strategies came last in the order of frequency of use as they stood at only 5.7%.The direct request strategies were mainly represented by "mood derivable" (e.g., Give me a phone.)(3.3%) and "want statement" (e.g., I would like to visit my aunt.)(2.2%).
In terms of external modifiers, the participants produced a total of 164 instances as shown in Table 6.These were dominated by "giving reasons" (65.9%) and "apologizing" (26.2%).
As for internal modifiers, the participants produced a total of 399 instances as shown in Table 7.These were mostly dominated by downgraders which represented 97.2% of the total modifiers while upgraders represented only 2.8%.The most prominent downgrader was "polite markers" (92.5%) whereas the most prominent upgrader was "intensifiers" (2.5%).
As for initiators, the participants produced 387 instances in English as shown in Table 8.These were mainly represented by address terms (59.2%), greetings (27.1%) and attention grabbers (9.8%).

Are there significant differences between the requests produced by Saudi EFL learners in Arabic and English?
As shown in Table 9, several Chi square comparisons reached significance between Arabic and English in terms of request levels and strategies.The participants produced significantly more direct strategies in Arabic than in English.This was particularly the case with "mood derivable" (e.g., Hand me an eraser.)and "want statement" (e.g., I would to have a pen).On the contrary, the participants produced significantly more nonconventionally indirect strategies in English than in Arabic, as mainly represented by "strong hints" (e.g., I wasn't able to study for the exam.).As for the category of conventionally indirect requests, no overall statistical difference was noted.However, the participants produced a significantly higher number of "suggestive formula" (e.g., How about going out tomorrow instead?) in Arabic than in English while the number of "query preparatory" (e.g., Could you please extend the deadline?)was significantly higher in English than in Arabic.
In terms of external modifiers, Table 10 shows that the participants produced a significantly higher number of "apology" in English than in Arabic while they produced significantly more "promise" and "prayer" in Arabic than in English.
As for internal modifiers, Table 11 shows that the participants produced a significantly higher number of downgraders in English than in Arabic.This was mostly in the form of polite markers.Regarding upgraders, the participants produced a significantly higher number of "intensifier" and "swear by God" in Arabic than in English.Although the participants produced significantly more downgraders in English than in Arabic, they used significantly more "understaters" and "hedges" in Arabic than in English.

Does social dominance have a significant influence on the production of requests by Saudi EFL learners in Arabic and English?
In Arabic, social dominance seems to have a strong influence on the production of requests as shown in Table 12.The participants produced significantly more direct strategies with the addressees who are lower in dominance, as represented by "mood derivable" (e.g., Buy a new pen.).As for conventionally indirect requests, the participants produced significantly more "suggestive formula" (e.g., Why don't you review your paper?) with equal addressees and significantly more "query preparatory" (e.g., Could you please remove your car from here?) with addressees with a higher dominance.Regarding non-conventionally indirect requests, the participants mainly produced a higher number of "strong hints" (e.g., I'm feeling a little cold.) with the addressees of lower and equal dominance.
In terms of external modifiers, only one instance of significant difference was noted as shown in Table 13.The participants produced significantly more "promise" with addressees of higher dominance.
As for internal modifiers, a few instances of statistical difference were noted as shown in Table 14.While the participants produced a significantly higher number of "polite marker" with    lower and higher participants, they used a significantly higher number of "understater" and hedge" with equal addressees.
As shown in Table 15, social dominance had a strong influence on the production of requests in English.While the participants produced significantly more "want statement" and "hint" to equal addressees, they produced significantly more "suggestive formula" and "query preparatory" with addressees of higher dominance.
In terms of external modifiers, two instances of statistical difference were noted as shown in Table 16.While the participants produced significantly more "reason" with the addressees with higher and equal dominance, they produced significantly more "apology" with the addressees with lower dominance.
As for internal modifiers, a few instances of statistical significance were noted as shown in Table 17.While the participants used significantly more "polite marker" with the addressees of higher and lower dominance, the participants employed significantly more "understater," "hedge," and "subjectivizer" with the addressees of equal dominance.

Discussion
The current study aimed to examine the realization of the speech act of request by Saudi EFL learners.The first two study questions aimed to describe how this speech act is realized in the participants' two languages.An important observation is that the conventionally indirect strategy of "query preparatory" was the most frequently used strategy in both Arabic and English.This finding comes in line with the results of AlMujaibel and Gomaa (2022) that the "query preparatory" was the most frequently used strategy in Kuwaiti Arabic and British English.This finding is also aligned with the results of Qari (2017) and Abdelfattah (2022) who found that Saudis, Moroccans, British and Americans used "query preparatory" most frequently in their responses.This could be the result of the basic linguistic structure used in this strategy as in "could you . . .?" and "would you mind . . .?".Additionally, the participants may have preferred the use of this strategy because it is an indirect strategy that can somehow mitigate the negative effect of requests as a facethreatening act (Brown & Levinson, 1978;1987).
The findings in relation to the first two questions also showed that "strong hint" was the most frequently used strategy at the nonconventionally indirect level and "mood derivable" was the most frequently used strategy at the direct level.As for modifiers, the participants used a variety of external modifiers in Arabic while they used only "giving reason" and "apologizing" in English.This could reflect the fact that Arabic is the participants' L1, and, hence, they found it easier to diversify their external modifiers.A final important observation is related to the use of downgraders, which dominated the scene in comparison with upgraders.While the former represented the majority of internal modifiers produced by the participants, the latter were produced to a much lesser extent, especially in English.This reflects the participants' preference to mitigate the negative effect of request as a potentially face-threatening act that represents imposition on the addressees (Brown & Levinson, 1978;1987).
The third research question aimed to compare the realization of the speech act of request between the two languages.In this regard, several statistically significant differences were noted.First, while the participants produced significantly more direct strategies in Arabic than in English, particularly in the form of "mood derivable" and "want statement", the participants produced a significantly higher number of the nonconventionally indirect strategy "strong hints" in English than in Arabic.Additionally, while the Arabic responses included significantly more "suggestive formula", the English responses included a significantly higher number of "query preparatory".This finding comes in line with several earlier studies (e.g., Abdelfattah, 2022;Alshammari, 2015;Qari, 2017;Yazdanfar & Bonyadi, 2016) in which the participants from Eastern countries (i.e., Saudi Arabia, Iran and Morocco) employed direct request strategies more often than the participants from Western countries (i.e., the USA and Britain).This finding is aligned with the prediction of Hofstede (2001) that individualistic cultures exhibit more sensitivity to the hearer's negative face and attempt to use negative strategies to mitigate imposition.
In terms of modifiers, the findings showed that the participants used significantly more apologies in English than in Arabic while they promised and offered prayers to the hearer more frequently in Arabic than in English.Again, this seems in line with Hofstede's (2001) prediction that the English language would exhibit greater sensitivity to the negative face and would try by all means to soften the effect of the imposition.As for the Arabic language, promises and prayers represent a high value on religious grounds, and it is well-known that Arabic language frequently employs religious discourse, especially in conversations.As for internal modifiers, it was noted that the participants used a significantly higher number of downgraders, particularly in the form of "polite marker," in English than in Arabic.This finding is aligned with Qari's (2017) observation that English speakers prefer to mitigate using linguistic devices while Saudis prefer to mitigate using semantic softeners.This finding further contributes to the participants' sensitivity to the negative face in English more than in Arabic.The participants thus seem to switch between the two languages with efficiency and appropriacy in terms of cultural differences.
The last research question aimed to explore the influence of social dominance on the realization of the speech act of request.Interestingly, this social variable seemed influential in the participants' two languages.In Arabic, "mood derivable" was more frequently used with janitors (lower dominance) while "suggestive formula" and "query preparatory" were used respectively more often with classmates (equal dominance) and professors (high dominance).Additionally, hints were less frequently used with professors than classmates and janitors.Likewise, in English, the participants used more "want statements'" and "strong hints" with classmates while they employed more "suggestive formula" and "query preparatory" with professors.Thus, as Brown & Levinson (1978, 1987) predicted, social variables do influence the realization of speech acts.In Arabic, our participants preferred to be more direct (e.g., the use of mood derivables) with addressees of a lower status (i.e., janitors) than with those of equal or higher status (i.e., classmates and professors).As for their choices in English, they tended to select more polite formulas with addressees of a higher status.
Further evidence in this direction comes from an examination of modifiers.In Arabic, more promises were used with professors, more polite markers were produced with professors and janitors while more hedges and understaters were used with classmates.In English, more reasons were given to professors and classmates while apologies were more frequently used with janitors.Additionally, and similar to Arabic, more polite markers were used with professors and janitors than with classmates and more understaters and hedges were used with classmates than with professors and janitors.

Limitations of the study
The results of the current study must be considered with some caution for the following limitations.First, our participants were placed at Level B1 according to the Common European Framework.Hence, students of a lower level of proficiency may not achieve this level of pragmatic competence.Further research is needed to assess the pragmatic competence of EFL learners with lower proficiency.Second, the participants in the current study were enrolled in an intensive pre-university English language training program, and were studying at a multicultural university.This combination offered them special exposure to the foreign language and culture.It is important to conduct similar studies with EFL learners who do not enjoy the same level of enhanced language exposure.Third, the current study was limited to male-male and female-female speech.Future research could examine inter-gender conversation, which could lead to different results.Fourth, our study examined the realization of the speech act of request in Najdi Arabic.Further studies can explore other dialects in Saudi Arabia or other parts of the Arab World.Finally, the current study employed discourse completion tasks.Further research employing other methods, including role-plays and observations, is recommended to examine if the results will remain the same across diverse instruments.

Conclusion
The current study aimed to examine the realization of the speech act of request by Saudi EFL learners.The participants' responses to 12 hypothetical situations in Arabic and English were examined, and the results showed that the participants tended to use more conventionally indirect strategies, in the form of "query preparatory", more frequently than all other strategies.However, some statistically significant differences emerged when the two languages were compared.Among the most notable was that the participants produced significantly more direct request strategies in Arabic than in English and more nonconventionally indirect strategies in English than in Arabic.The results also showed that the participants tended to apologize more frequently in English than in Arabic and to promise and give prayers more in Arabic than in English.Additionally, although the participants used more variety of external modifiers in Arabic than in English, they produced a significantly higher number of downgraders in English than in Arabic.
Based on the findings of the current study, several theoretical and pedagogical implications are proposed.First, the findings in the current study support the face-threatening nature of the speech act of request as described by Brown and Levinson (1978;1987).The findings also lend support to the claim that the realization of speech acts is greatly influenced by relevant social variables, particularly social dominance.Second, the realization of speech acts is culture-specific (Cheng & Lam, 2020;Chentsova-Dutton and Vaughn, 2012), and thus we need to be cautious when generalizing the predictions of Brown and Levinson's (1978;1987), model which seems to lean more towards Western communities.Third, differences between collectivistic and individualistic societies (Hofstede, 2001) seem to have a high impact on the realization of speech acts.In the current study, EFL learners showed sensitivity to the nature of each of their two languages while realizing the speech act of request.Finally, in terms of pedagogical implications, we agree with earlier researchers that EFL learners can develop an adequate level of pragmatic competence with increased foreign language exposure (e.g., Montenegro, 2017).In the current study, the participants seemed to select adequate linguistic forms and successfully accommodate the requirements of the cultural context and social variables in their FL-English.The intensive English language exposure the participants received, whether in terms of class instruction or the multicultural nature of their university, must have greatly contributed to this satisfactory outcome.In other contexts in which this extensive exposure to English is not available, explicit instruction of speech acts will be relevant (Qari, 2021;Roever, 2022) Hello [greeting], Ms. [address term] Moneera [name]!How are you?[greeting] Could you please extend the deadline of our homework [query preparatory] a bit [hedge]?Good morning [greeting], Mr. [address term] Sherif [name]!Would you mind explaining the last section again?[query preparatory] I could not understand it well.[reason] Sheri [name]!Give me a pen [mood derivable], please [polite marker].I need to take notes [reason].
testFE: Fisher Exact p: p value for comparing L1 Arabic and L2 English *Statistically significant at p ≤ 0.05 testMC: Monte Carlo *Statistically significant at p ≤ 0.05 testMC: Monte Carlo *Statistically significant at p ≤ 0.05 square test MC: Monte Carlo Statistically significant at p ≤ 0.05