Request Modifications by Malay Speakers of English in the Workplace: A Contrastive Pragmatic Analysis

A request is one of the most frequently used speech acts in a person ’ s daily life. Many studies have been conducted on the act of request and several researchers have developed strategies and modifications that are used in delivering requests to mitigate the imposition of the FTA. Thus, this study aims to investigate the internal and external modifications applied in requests by Malay speakers of English and Malay to specific requestees in the workplace context following Blum-Kulka et al. ’ s (1989) Request Modification framework. To achieve the purpose of this study, a qualitative approach was employed. A number of 30 Malay workers were asked to complete a Written Discourse Completion Task (WDCT) which involves eliciting requests. The findings revealed that the respondents used more internal modifications in their requests with people of equal relative power compared to high and low relative power, and used more external modifications in requests with a higher degree of imposition in certain contexts. This implies the act of request from the viewpoint of the Malaysian workplace context. It offers meaningful insight into the preferences of working Malay speakers of English in modifying their requests, and shows the level of pragmatic competence of Malay speakers of English and Malay in the workplace environment.


INTRODUCTION
It is inadequate for a person to achieve only accurate linguistic knowledge that follows the rules of grammar but leaves out the lexical aspects inadequate. Particularly, a person must also know how to use knowledge in communication by attaining pragmalinguistic and sociopragmatic understandings (Chang & Ren, 2020). If one is from a different culture with a different social understanding and lacks pragmatic and linguistic knowledge when interacting, failures in verbal communication can occur (Maros & Halim, 2018). Therefore, it is imperative for a person, especially L2 speakers, to develop their communicative competence.
'Communicative competence' was introduced by Hymes (1972) who stated that speakers need to be able to communicate successfully through the knowledge they gained from acquiring a language based on certain contexts and hearers. This concept is also known as 'pragmatic competence'. Halupka-Rešetar (2014) stated that pragmatic competence is classified under communicative competence which, based on Bachman's (1990) model, encompasses illocutionary competence (the knowledge of using speech acts) and sociolinguistic competence (the ability to utilize language appropriately according to the context of the conversation). Hence, pragmatic competence can be defined as the ability to use a language by uttering socially appropriate speeches based on various contexts (Daskalovska et al., 2016).
According to Zhu (2012), a person's pragmatic competence can be evaluated based on his/her speech act behaviour. Speech acts are actions made through utterances that speakers use to convey intended actions and where hearers interpret the intended meaning-apologies, requests, refusals, complaints, etc. (Halupka-Rešetar, 2014;Khalib & Tayeh, 2014). Request acts are one of the most used speech acts in everyday interactions (Halupka-Rešetar, 2014;Shafran, 2019), and being pragmatically competent when making requests is a crucial aspect (Alsout & Khedri, 2019). For years, the study of the request speech act has taken the interest of many researchers around the world (Al Masaeed, 2017;Su & Ren, 2017). This is because a request is considered a Face-Threatening Act (FTA) which, if not delivered appropriately through modifications and strategies, can cause communication breakdowns (Alsout & Khedri, 2019;Brown & Levinson, 1987;Halupka-Rešetar, 2014).
Therefore, for a person to become pragmatically competent when requesting in both their L1 and L2, one's linguistics knowledge needs to be closely associated with sociopragmatic knowledge (the relative degree of imposition in the culture of the L2) and pragmalinguistic knowledge (the degree of politeness of utterance in the culture of the L2) (Halupka-Rešetar, 2014;Maros & Halim, 2018). However, forming the appropriate and effective form of request in one's L2 has still become a challenge to non-native speakers, especially if their mastery of that language is low (Shafran, 2019).
Additionally, studies have shown that, in some cultures, making a direct request to someone that is not close to the requester or has a higher status may make the requestee feel offended or imposed on (Halupka-Rešetar, 2014). In terms of using a second or foreign language, even though most people know how to make requests appropriately in their first language (L1), this does not necessarily mean that they know how to do it in a second language (L2) (Maros & Halim, 2018), especially when speaking to people from different cultural backgrounds. Cross-cultural differences may affect the realization of requests for native speakers and non-native speakers of a language (Barron, 2016).
Previous studies have also discussed the types of strategies and modifications used in delivering requests within the academic setting (e.g., Alsout & Khedri, 2019;Daskalovska et al., 2016;Halenko & Jones, 2017;Karatepe, 2016;Khalib & Tayeh, 2014), however, there are still few studies on the use of these request strategies and modifications in the workplace environment. Moreover, the number of studies on the comparison between L1 Malay and L2 English speakers on request act is still scarce compared to other interlanguage and cross-cultural studies on request speech act (e.g., Al Masaeed, 2017;Ninomiya & Shadayeva, 2020;Shafran, 2019). Although there is a study comparing Malay and English speakers' use of request strategies (e.g., Maros & Halim, 2018), it only focused on one modification among the other request modifications that this study explored.
Therefore, to attend to this gap, this study examined the modifications applied towards specific requestees made in L1 Malay and L2 English at the workplace following Blum-Kulka et al.'s (1989) Request Modification framework. This study hopefully sheds light on the assumptions and perceptions regarding effective request approaches and how L1 may either interfere with or facilitate the transfer of sociopragmatic and pragmalinguistic knowledge to L2. To fulfil the purpose of the research, the present study seeks to answer this research question as follows: What are the internal and external request modifications employed by the Malay speakers of English when delivering requests in Malay and English based on the social variables of power, social distance, and degree of imposition?

Sociological Variables in Making Requests
A request is one of the Face-Threatening Acts (FTA) as it involves the speaker (requester) impeding the hearer's (requestee) freedom of action (Brown & Levinson, 1978). The requester is asking the requestee to do something that the requestee does not benefit from; thus, complying with such requests can be hard for the requestee as the act is imposing on his/her time. Hence, speakers need to take note of the social status of the involved interlocutors (Yazdanfar & Bonyadi, 2016). Three sociological variables were introduced by Levinson (1978, 1987) to adjust the degree of politeness and determine the severity of the FTA before executing the request act (Ninomiya & Shadayeva, 2020;Shafran, 2019). These variables consist of 'relative power' (P), 'social distance' (S), and 'rank of imposition' (R).
According to Brown and Levinson (1978), relative power refers to the authority of a person over another, such as a professor to a student. Under this circumstance, those with higher power have authority over those inferior to them. Then, social distance is the frequency of interactions between the interlocutors where friends and relatives are considered close social distance. Lastly, the degree of imposition refers to the degree of interference with the hearer's negative or positive face wants. 'Face wants' refers to the social desire to maintain or enhance their social image and identity A. A. M. Idris & I. N. Ismail, Request modifications  in interaction. This concept is closely related to the idea of 'face-threatening acts' (FTAs), which are speech acts that can potentially damage or challenge the face of the interlocutor (Brown & Levinson, 1987;Ting-Toomey, 1988, 2017. These variablespower, social distance, and degree of imposition mostly compare the speaker's (requester) and hearer's (requestee) social status and the nature of the imposition (big or small). To simplify, the farther the social distance between the requester and requestee, the more politeness is needed in the request act; the more the relative power of the requester (e.g.: boss) to the requestee (e.g.: employee), the less politeness is required; and the higher the degree of the imposition, the more politeness is recommended.

Request Modifications
According to Faerch and Kasper (1989), request modifications can be implemented internally and externally to the main request act (Halupka-Rešetar, 2014). 'Internal modification' is a strategy through the use of lexical and syntactical components that function to either minimize or intensify the force of the request made (Al Masaeed, 2017;Halupka-Rešetar, 2014;Ninomiya & Shadayeva, 2020;Su & Ren, 2017). These modifications, as the term implies, are internally embedded into the head act of the request to modulate the illocutionary force. For example, the word 'please' is an internal modification that is commonly embedded within the head acts. The final taxonomy of internal modifications used in this study is presented in Tables 1 and 2:  (2010)).

Lexical modifications Definition Example Marker 'please'
An optional element added to a request to bid for cooperative behaviour.
"Help me with these bags, please."

Consultative devices
Expressions by means of which the speaker seeks to involve the hearer directly bidding for cooperation.
"Would you mind", "Do you think", "Would it be all right if", "Is it/would it be possible", "Do you think I could", Downtoners Modifiers that are used by the speaker to modulate the impact his or her request is likely to have on the hearer.
'possibly', 'perhaps', 'just', 'rather', 'maybe' Understaters/hedges Adverbial modifiers by means of which the speaker underrepresents the state of affairs denoted in the proposition.

Subjectivizers
Elements in which the speaker explicitly expresses his or her subjective opinion vis-avis the state of affairs referred to in the proposition, thus lowering the assertive force of the request.
'I'm afraid', 'I wonder', 'I think/suppose' Table 1 continued… Cajolers Conventionalized, addressee-oriented modifiers whose function is to make things clearer to the addressee and invite him/her to metaphorically participate in the speech act.
"You know…", "You see…" Appealers Addressee-oriented elements occurring in a syntactically final position. They may signal turn availability and "are used by the speaker whenever he or she wishes to appeal to his or her hearer's benevolent understanding.
'Wash the dishes, dear, will you?

Lexical upgraders
Overstater 'Exaggerated utterances that form part of the request and are employed by the speaker to communicate their need of the request being met.
"I'm in desperate need of your help." Intensifiers An adverbial modifier that stresses specific elements of the request.
"I truly/really need your help"

Time intensifiers
Time intensifiers users to stress the urgency of the request.

Lexical uptoners
Stressing on the lexical to heighten the need for the matter requested by the speaker.

Definition Example
Conditional structure The use of conditional structures. "Could you help me with this?" Conditional clause The use of conditional sentences like 'if' in a clause.
"… if it's possible to set a meeting today."

Interrogative
The use of '?' as a question within a request.
"Will you wash the dishes tomorrow?" Negation of a preparatory condition Sentence structured in a negation form.
"I don't suppose there is any chance of an extension?" Past tense Sentence structured with a verb modified into the past tense.
"Is it alright if I asked for your help with this…" Aspect The use of aspect marker in a clause.
"I was wondering if it's possible to set a meeting today." 'External modification' is a strategy that modifies the request outside of the head act by preceding or following the main request (Maros & Halim, 2018;Ninomiya & Shadayeva, 2020). As a result, this strategy is known as 'external' or 'supportive moves'. In this case, external modifications function to either aggravate or mitigate the speech act of request (Flores-Salgado & Castineira-Benitez, 2018; Halupka-Rešetar,  Table 3 shows the final modified external modification taxonomy used in this research. Table 3. External request modifications.

Adjunct to head act Definition Example
Checking for availability The requester tries to see whether the requestee is available to commit to the request.
"Are you on your way to the supermarket? If so, is it okay if I come with you?" Getting a precommitment The requester checks for a potential refusal before delivering the request by trying to get a pre-committal from the requestee.
"Would you do me a favour? Can you help with my bags?"

Preparator
The requester prepares the hearer for the following request by stating their action.
"I need a favour from you. Is it alright if I borrow your car tonight?" Grounder The requester provides reasons, explanations, or justifications behind the request.
Can I borrow some money? I forgot my wallet."

Disarmer
The requester tries to sway the requestee from raising potential objections.
"I know you are a busy person, but could you help with this paper?" Imposition minimizer The requester tries to mitigate the imposition placed on the requestee by the request.
"May I ask for some of your time? Just for a few minutes."

Apology
The requester acknowledges the imposition created by the request and apologizes for it.
"I apologize for disturbing you, but could you help me with this box?" Gratitude The requester expresses their appreciation for the requestee's compliance with the request.
"Is it possible for you to review my work? Thank you in advance."

Request Modifications across Cultures and Languages
Request modifications are lexical and syntactic elements that are included either internally or externally in the main request act (Flores-Salgado & Castineira-Benitez, 2018) and they function as mitigators to reduce the force of a request. Previous studies have been conducted to see how different people across the world modify their requests. Muthusamy and Farashaiyan's (2016) research showed that a majority of the international postgraduate students from different countries (Iraq, Iran, Jordan, India, and Tunisia), studying in Malaysia use more external modifications (66.6%) compared to internal modification (33.3%) through the frequent use of 'please' marker and grounders. Here, the high frequency of such use of modifiers, according to Blum-Kulka and Olshtain (1984), is because they are easy to produce by non-native speakers (Muthusamy & Farashaiyan, 2016). To put it simply, a 'please' marker is the easiest way to express one's need for assistance while grounders do not require proficient pragmalinguistic knowledge to justify their request.
Such findings are very much consistent with those in Halupka-Rešetar (2014) where the study examined the type and frequency of internal and external modifications used in the request-making among intermediate proficiency level ESP students in Serbia. The request-making of the students showed little variation in terms of the type of modification and the frequency of use. However, among the external modification elements, grounders were used most often and almost exclusively. On the other hand, the use of the politeness marker 'please' and conditional clauses showed the highest percentage of internal modification. Thus, the findings in both studies show significant similarities which demonstrate how 'please' markers and grounders are two of the most convenient types of internal and external modifiers to mitigate requests among non-native speakers of a target language.
However, studies like Woodfield and Economidou-Kogetsidis (2010) discovered contradicting results of internal modification use among English learner participants (Greek, Japanese, and German) where, through the production of written requests, the use of 'please' are quite low in percentage (16.9%) while the use of zero marking (i.e., no use of any internal modifier) shows the highest percentage in both lexical and syntactic downgraders; (38.2%) and (64%). Although these English learners are deemed to have high English proficiency skills with good scores in both TOEFL and IELTS exams, it is surprising to see their lack of internal modifications within their requests. Yet, in terms of external modifications, findings showed similar results with high-frequency zero marking use (21.34%) but with the use of grounders being the highest (71.91%) and other forms of external modifiers as the lowest (from 0% to only 8.98%). Hence, this shows that while they do provide reasons for their requests, the lack of other external modifiers makes their requests sound quite direct and coercive to hearers. Subsequently, the reasoning for this may be that while they are good in terms of grammatical and lexical commands, their pragmalinguistic knowledge is still quite underdeveloped. Thus, this proves how English learners need to gain a better grasp in terms of sociopragmatic and pragmalinguistic knowledge to produce better requests that are mitigated properly and sound more polite.

Research Design
This study aims to investigate the request modifications used by Malay speakers of English in the workplace. Therefore, to fulfil this objective, a qualitative research approach was employed to obtain an in-depth understanding of participants' perspectives on the research phenomenon from the participants' responses. The data was analysed according to Blum-Kulka et al.'s (1989) Request Modifications taxonomy.

Research Sample
A purposive sampling method was applied to select participants who met specific criteria. The criteria for inclusion were Malay individuals who speak English as a second language, and who work at various private organizations in Selangor, Malaysia. By selecting participants based on these criteria, we were able to focus our study on the experiences of a specific population and gather in-depth data about their experiences. Previous studies that employed discourse completion tasks (DCT) regarding request acts in their research (e.g., Al Masaeed, 2017;Halupka-Rešetar, 2014;Khalib & Tayeh, 2014;Shafran, 2019;Woodfield & Economidou-Kogetsidis, 2010), the common number of participants is in the range of 30-176. Hence, for this study, putting into consideration the then pandemic which can affect participants' willingness to participate in the research, and the limited time given to collect and analyse the data, the ideal sample size to obtain the qualitative data is 30.

Research Instruments
The research instrument is a Written Discourse Completion Task (WDCT) administered via Google Forms. The written task adapted and modified from Daskalovska et al. (2016) and Shafran (2019), was designed to collect and elicit requests in written form. Participants were given six different scenarios of real-life situations that require them to form their requests in both Malay and English to people of different relative power (P) and social distance (S) and ask favours of different degrees of imposition (R). The comparison between the two languages was done to examine the similarities and differences in the requests made by anglophones in their mother tongue, as well as in English. In this study, the rate for the P, D, and R is determined through the context of each scenario. According to Brown and Levinson (1978), relative power (P) refers to the level of authority one has over another. For instance, the requestees of Scenarios 1 and 2 have high relative power (+P) as they have the authority/influence over the requester's situation/needs. Scenarios 3's and 6's requestees have equal relative power (=P) since they are those that the requester has no authority over them nor them to the requester. Scenarios 4's and 5's requestees have low relative power (-P) as they are considered junior/subordinate and required to follow the requester's requests.
Then, social distance (D) is the level of familiarity between interlocutors. Here, Scenarios 2, 3, and 4 are high social distance (+D) as the requestees do not have close relationships with the requester, whereas Scenarios 1, 5, and 6 are low social distance (-D) as the requestees have close relationships with the requestee.
Finally, the degree of imposition (R) is the degree of interference to the hearer's (requestee) negative or positive face wants. For example, Scenarios 2, 3, 4, 5, and 6 are a high degree of imposition (+R) as the requests involve big favours that can cause one to feel imposed upon, while Scenario 1 is a low degree of imposition (-R) as the request is quite minor and would not impose too much on the requestees' time or threaten their positive face.
This method is significant for this study as relevant data can be acquired. From here, the study would be able to see how respondents would formulate their requests according to requestees of different sociological variables. Respondents' answers can elicit internal and external request modifications (e.g., conditional structures, intensifiers, grounder, imposition minimizer, etc.) that respondents may use in their requests. Additionally, WDCT has been utilized by other researchers such as Halupka-Rešetar (2014) and Maros & Halim (2018) to collect data on external and internal request modifications through social situations.

Data Collection and Data Analysis
The link to the Google Forms containing WDCT was distributed to 30 Malay speakers of English in the private sector workplace via online instant messaging platforms (i.e., WhatsApp Messenger and Telegram) and social media (i.e., Twitter, Instagram, and Facebook). They were expected to answer the WDCT in both Malay and English languages by forming six requests according to the scenarios given in the survey.
Next, once the responses were selected and tabulated, the total frequency of both external and internal modifications used by participants was calculated. Specifically, a coding process based on Blum-Kulka et al.'s (1989) Request Modification taxonomy was conducted by classifying participants' responses (i.e., requests made for the scenarios in the written task) in a table according to the internal or external modifications used (e.g., appealers, intensifiers, grounder, gratitude, imposition minimizer, etc.) for each scenario. Then, the data was calculated with the degree of imposition (e.g., small to large favours) and the social power (combination of social distance and power) according to the roles in the scenarios. This is to observe the number of modifications used by the respondents in situations that have either a higher or lower degree of imposition involving people of different power and social distances. Using Blum-Kulka et al.'s (1989) Request Modification taxonomy is a crucial part of this study as it helps the researcher to classify and analyse the modifications (internally and/or externally) employed by the respondents in their requests to requestees of different sociological variables within different contexts.

Cohen Kappa's Inter-Rater Agreement
To maintain the data reliability of the WDCT responses, Cohen Kappa's interrater agreement (Dobakhti, 2020) was measured. Two native Malay speakers who are majoring in English and familiar with research analysis and methodology were asked to independently rate their agreement or disagreement with the already coded English and Malay request modifications in participants' WDCT responses. Rau and Shih (2021) commented that this test involves the extent to which raters make the same judgment. High agreement means raters assign the same units to the same category (i.e., nominal, ordinal, and interval) and the agreement value can be calculated and evaluated for its reliability based on Kappa's value scale; <0 (no agreement) to 0.80-1.00 (almost perfect agreement). Li and Raja Suleiman's (2017) study on Chinese EFL learners' L2 proficiency and ability to produce complaints used this data reliability test to see the consistency in the coding scheme.
To illustrate, over the 360 English and Malay responses that have been coded with different request modifications, 20% of them (36 responses; 18 in English and 18 in Malay) were given to the raters separately. There were over 297 internal modifications and 267 external modifications coded in those 36 responses and raters have to either agree or disagree with the request modifications codes labelled by the researcher. The calculation was conducted using the following Cohen Kappa's Interrater agreement formula: Where: P_Rated : the number of coded strategies agreed P_Expected : 50% of the number of coded strategies expected to be agreed upon N : the total number of coded strategies measured for agreement From here, the Kappa values collected from all the raters were gathered and calculated to acquire the mean Kappa value. This mean value is the Kappa value that indicated the reliability of the overall data in this study. According to the calculations done, the Cohen-Kappa Inter-rater agreement showed a Kappa value of 0.98 for internal modifications and 0.96 for external modifications coded. Both values indicate an almost perfect agreement of the request modification codes labelled by the researchers.

Internal Modifications Employed in English and Malay Requests
The results in Table 5 below show very close numbers of internal modifications employed in both English and Malay requests for each scenario. Among all the requests elicited by the respondents, Scenario 6 shows the most internal modifications used in both English and Malay with a total of 114 and 113 internal modifiers employed respectively. This is then followed by Scenario 4 (English,98;Malay,91) and Scenario 3 (English, 97; Malay, 91). On the other hand, Scenario 2 shows the least number of internal modifications used in English and Malay with only a frequency of 39 and 35. This is followed by Scenario 5 (English,46;Malay,46) and Scenario 1 (English,59;Malay,59).
To specify, it can be noted that all requests elicited were delivered to the requestees that have different sociological variables (power, social distance, and degree of imposition) and this seems to affect the number of internal modifiers used by respondents and the reason why the number of internal modifications in English and Malay is close to each other based on the different scenarios. For instance, based on Table 5, the requests for requestees of higher relative power (+P) like in Scenario 1 and Scenario 2 had a little number of request modifiers used (English,59 and 39;Malay,59 and 35). In contrast, requests for requestees of equal power (=P) like in Scenario 3 and Scenario 6 had a larger number of request modifiers used (English,97 and 114;Malay,91 and 113).
However, requests for requestees that had lower power (-P), such as in Scenario 4 and Scenario 5, had a large gap in number. Specifically, Scenario 4 had a total number of 98 request modifications in English and 91 in Malay while Scenario 5 only had 46 internal modifications done in both English and Malay. Therefore, other sociological variables (social distance and degree of imposition) need to be examined. Looking more specifically into the types of internal modifications used based on the different request scenarios, the findings showed that in the context of 'normal worker to boss' in Scenario 1 (high relative power, low social distance, and low degree of imposition), and 'normal worker to potential guest' context in Scenario 2 (high relative power, high social distance, and high degree of imposition), respondents mostly used formal language involving direct speech for both languages in their request to state their intentions and needs clearly.  Table 5 continued…   Syntactic downgraders  Conditional Structure  11  17  0  0  14  27  Conditional Clause  7  3  8  6  15  11  Interrogative  19  19  3  2  28  29  Negation  0  5  0  0  0  6  Past Tense  0  0  1  0  0  0  Aspect  0  0  1  0  2  0  Total  37  44  13  8  59  73  TOTAL  59  59  39  35  97 91 Note: P = Power D = Social Distance R = Degree of Imposition Scenario 1 involved formally requesting a due date extension for a minor project. The way of requesting had no internal modifications since respondents intended to sound clear and not mitigate their request which can cause ambiguity. However, there had been other respondents that employed certain types of internal modifiers in their requests for both scenarios to mitigate or intensify their needs. For instance; (1) "Hi, boss. I'm really (Intensifier) sorry for bothering you at this time but is it possible (Consultive Devices) to get an extension for the project report The requests elicited to show that the internal modifiers used were meant to sound more polite and to give the requestee the freedom to reject the request rather than being insistent which can cause an FTA towards their negative face. Yet, the number of internal modifiers was used less compared to other scenarios.  12  3  0  0  8  3  Consultative devices  7  1  2  0  3  2  Downtoners  5  2  0  0  3  0  Understaters/hedges  3  5  0  1  19  17  Subjectivizers  7  4  0  1  2  2  Cajolers  0  0  0  0  1  0  Appealers  3  6  0  0  1  1  Total  37  21  2  2  37  25  Lexical upgraders  Overstater  0  1  0  0  2  3  Intensifiers  4  4  1  1  14  10  Time intensifiers  2  3  1  0  1  1  Lexical uptoners  0  0  0  0  0  0  Total  6  8  2  1  17  14 993 | Studies in English Language and Education, 10(2), 981-999, 2023 Table 6 continued…   Syntactic downgraders  Conditional structure  19  24  8  7  20  27  Conditional clause  6  1  5  3  12  7  Interrogative  28  27  28  28  27  28  Negation  0  10  0  5  0  12  Past tense  0  0  0  0  0  0  Aspect  2  0  1  0  1  0  Total  55  62  42  43  60  74  Total  98  91  46  46  114 113 Note: P = Power D = Social Distance R = Degree of Imposition In contrast, in the context of 'colleague to new colleague' in Scenario 4 (low relative power, high social distance, and high degree of imposition), and 'friend to friend' context in Scenario 6 (equal relative power, low social distance, and high degree of imposition), the respondents used much more casual language in their requests to the requestees compared to the previous scenarios but employed more internal modifications despite having low and equal relative power respectively. For example: (2) "Anis, I'm not feeling very well and I don't I think (Subjectivizer) I can come to work today.
Could you (Conditional structure) please (Please marker) cover my work just ( Based on these responses, the respondents utilized more internal modifiers in their requests compared to those in Scenarios 1 and 2. The context here (Scenario 4) shows respondents asked their requestees for things that can threaten their negative face since they imposed on their requestee's time and effort. The respondents were shown to have used at least two types of internal modifications; 'lexical downgraders' like subjectivizer, please marker, and downtoner (English), and understater and please marker (Malay), 'syntactic downgraders' like conditional structure, conditional clause, and interrogative (English), and conditional structure, and interrogative (Malay), 'lexical upgraders' like time intensifier (Malay). These modifiers were able to form polite requests that could reduce the impingement and mitigate the FTA in the requests as they did not assert their needs onto the requestee and gave them pressure.
The high number of internal modifiers used in both requests to the requestees of low and equal relative power compared to high relative power suggested that respondents took more into consideration of the context and degree of imposition within the request rather than who the requestee was. Additionally, despite Scenario 2 having a similar context to Scenario 4 that imposed on the requestee's time, the context which involved requesting Mr. Hakimi's presence as a guest speaker in a form of an invitation gave a sense of acknowledgement of the requestee's credibility; hence, A. A. M. Idris & I. N. Ismail, Request modifications  enhancing his positive face value (Bardovi-Harlig, 2019). That said, this way of requesting was seen to be more effective and seemed to not need many internal modifications.

External Modifications Employed in English and Malay Requests
Similar to the internal modification, the findings in Table 7 show very close numbers of external modifications used for each scenario in both English and Malay requests. Among all the requests elicited by the respondents, Scenario 3 shows the highest number of external modifications employed in both English and Malay with a total number of 109 and 102 respectively, followed by Scenario 6 (English, 81; Malay, 83) and Scenario 4 (English,68;Malay,62). Contrastingly, Scenario 5 shows the least number of external modifications employed in English and Malay with only 33 and 30 external modifiers, followed by Scenario 2 (English, 60; Malay, 55) and Scenario 1 (English, 65; Malay, 62). except for the degree of imposition (+R) in Scenarios 3 and 6, and Scenarios 2 and 5. Therefore, the context for each scenario, especially those that influence the degree of imposition, needs to be examined for better interpretations of the results. Looking more specifically into the types of external modifications used based on the different request scenarios, the findings showed that in the context of 'stranger to stranger' in Scenario 3 (equal relative power, high social distance, and high degree of imposition), and 'friend to friend' context in Scenario 6 (equal relative power, low social distance, and high degree of imposition), respondents had shown to use very casual and polite language but with many external modifications employed in both languages. For example: (3) "Hi! So sorry to bother you (Apology In the context of this scenario, it involved asking the requestee's help with something that could threaten the requestees' negative faces. This is because the request was imposing on the requestees' personal space and time since the requestee was facing a stranger that could have ill intentions. Thus, the reason why the degree of imposition was high. Therefore, it can be deduced that in order for the requestee to comply with the requests, respondents had to employ many external modifications like an apology (to admit the imposition and to apologize for it), preparator (to prepare the requestee of an upcoming request), checking for availability (to see if the requestee is available to commit to the request), grounder (to explain the reason behind the request), imposition minimizer (to mitigate the imposition), and gratitude (to express appreciation towards the requestee's compliance). Hence, this showed that the number of external modifications heavily relied on the context and degree of imposition within the requests.
In brief, based on the findings of the internal and external modifications used in English and Malay requests, it can be concluded that the number of internal and external modifications used within a request relied heavily on the requester-requestee's relative power and social distance, the context of the request based on the degree of imposition, and the type of request elicited (i.e., invitations).

DISCUSSION
Based on the findings, it has been discovered that Malay speakers of English in the workplace show high pragmatic competence whereby most of them were able to A. A. M. Idris & I. N. Ismail, Request modifications  use both internal and external modifications in both English and Malay to make polite requests. In the use of internal modifications, respondents had been shown to use more internal modifiers to requestees of equal relative power compared to those of higher or lower power. In fact, it was found that respondents used the least number of internal modifications to requestees of high relative power and had been shown to use more direct speech to them with very few modifications done. One reason for this could be due to the level of formality maintained by the respondents within their requests to their boss and client to provide a clear statement of their needs without hedged and ambiguous statements. This is further supported by Balman and Sangmok (2020) who argue that in some instances in requests with people of a higher relative power, getting to the point of the requests without lengthy and hedged explanations is better than wasting one's time reading/listening to the requests. In addition, based on the analysis, it can be deduced that more internal modifications were used to those of equal to lower relative power because of the degree of imposition within the requests where the context involves asking for one's valuable belongings (a large amount of money), imposing on one's time and effort (cover for work), and imposing on one's time and space (giving a stranger a ride). This is quite similar to the use of external modifications where respondents used more external modifiers in requests that had a higher degree of imposition with specific contexts involving requestees' valuable belongings (a large amount of money) and personal time and space (giving a stranger a ride). Consequently, these contexts caused respondents to use more explanations, promises, apologies, expressions of appreciation, and disarmers in order to increase their chances to have their requests accepted. This result is shown to be consistent with the results in Halupka-Rešetar's (2014) study where despite not having clear evidence of whether relative power influence the number of supportive moves used, the degree of imposition, on the other hand, did affect the number of external modifications used. In other words, the higher the degree of imposition, the more mitigations were needed. This shows that Malay speakers of English in the workplace focused more on the specific context of the requests and the degree of imposition rather than the relative power and social distance of the requestees.

CONCLUSION
Despite the great number of studies done in the field of request speech acts, there is still a limited number of studies done based on the Malaysian context. Since working Malaysian people use English frequently as their L2, investigating the request approach made by them and how they react when receiving them is important. To illustrate, not only would it provide advancement in the field of pragmatics, but it could also help English speakers in Malaysia, especially employees that interact with people daily, to reflect on themselves and see whether their usual request approach is acceptable or not in the eyes of others, especially to native English speakers. Apart from displaying proficient pragmatic competency level among the respondents, the findings in this study also revealed that they used more internal modifications in their requests with people of equal relative power compared to high and low relative power, and used more external modifications in requests that have a higher degree of imposition within certain contexts.
Several areas can be addressed in future research. Firstly, future studies can employ an observation method to obtain more authentic data on delivering requests in natural conversations. This is important since it may help shed light on how a community portrays politeness in their speech within a natural environment. Secondly, future research can also carry out interviews to observe respondents' perceptions of receiving requests at different directness levels. Here, this would help provide an indepth understanding of why they feel directness is ruder to receive than another and how they perceive receiving requests from people of different social statuses.