INTRODUCTION

Politeness as an underlying and overriding constituent of pragmatic competence has often been considered one of the primary interests of many researchers and practitioners within the realm of pragmatics (Holmes, 2006). Politeness serves as a crucial communication tool, essential for maintaining social harmony and avoiding misunderstandings. Politeness markers are language statements used to display courtesy, and they are deeply ingrained in the theory of politeness which encompasses a range of concepts and notions. The misapplication of politeness markers can lead to communication breakdowns which can have significant social and professional consequences (Watts, 2003). While language learners often use politeness markers effectively in their first language (L1), they frequently underuse or misuse these markers in their second language (L2) (Kasper, 1997; Mohammad Hosseinpur & Mousavi, 2021). This discrepancy is particularly pronounced in English as a Foreign Language (EFL) setting, where learners may lack the necessary awareness and ability to appropriately use L2 politeness markers in different contexts. The ability to properly apply politeness markers requires a high degree of processing control and pragmatic competence (Kasper & Rose, 2002).

Despite the extensive research on politeness in various pragmatic studies (e.g., Aini et al., 2023; Bousfield, 2008; Dufon, 2008; Eslami et al., 2023; Knupsky & Nagy-Bell, 2011; Tajeddin & Mohammad Hosseinpur, 2014; van Dorst et al., 2024; Watts, 2003), a significant gap remains in the literature regarding the representation of politeness markers across different measures of pragmatic competence in EFL settings. This gap is particularly problematic because it limits our understanding of how EFL learners process and produce politeness markers in various communicative contexts. Addressing this gap is crucial for developing more effective teaching strategies and materials that can help learners improve their pragmatic competence in L2.

Different measures have been suggested to assess pragmatic competence, which can be broadly categorized into two subcategories: natural discourse and elicited information (Felix-Brasdefer, 2007). Natural discourse involves field observations and recordings of naturally occurring data, while elicited information includes methods such as discourse completion tests (DCTs), questionnaires, and role-plays (Golato, 2003). Each of these measures provides unique insights into learners’ pragmatic abilities, but the extent to which they capture the nuances of politeness marker use remains underexplored.

The present study aims to fill this gap by comparing and contrasting the production of politeness markers among EFL learners across different measures of pragmatic competence, specifically written discourse completion tests (WDCTs), oral discourse completion tests (ODCTs), role--plays, and natural methodology. By examining these different measures, the study seeks to provide a comprehensive understanding of how EFL learners represent and use politeness markers in various communicative contexts. This research is essential for advancing our knowledge of pragmatic competence in L2 and for informing the development of more effective pedagogical approaches to teaching politeness in EFL settings.

LITERATURE REVIEW

Politeness Theory

Several prominent scholars in pragmatics, including Lakoff (1977) and Leech (1983), have recognized the Politeness Principle as a fundamental aspect of this field. Lakoff (1977) proposed that individuals select appropriate politeness rules based on contextual factors, such as social status differences, familiarity, and cultural norms. She defined politeness as encompassing two key elements: clarity, influenced by Grice’s (1975) maxims of conversation, and politeness itself, which she further categorized into three rules: formality, hesitancy, and equality. Leech (1983) made significant contributions to our understanding of politeness.

Brown and Levinson (1987) leveled criticisms against Leech’s politeness model arguing that it was too theoretical and abstract. Brown and Levinson (1987) added another dimension to their model in addition to status differences among interlocutors and their relative power: Perceived imposition rate inherent in the utterance. Most cross-cultural speech act studies have taken this model as their conceptual framework. Their politeness framework is deeply rooted in the notion of face. The right of an individual to be free from confusions, befuddlements, and obstructions is regarded as negative face, while the positive self-image that one holds for himself or herself and one’s longing and yearning to be confirmed by others can be described as positive face. Individuals might apply various techniques and strategies to implement face threatening acts.

Politeness Markers

Politeness, a fundamental aspect of human interaction, is intricately linked to successful communication. As Brown and Levinson (1987) argue, politeness arises from the inherent human need to maintain face, both for oneself and for others. Face, in their framework, refers to the public self-image that individuals strive to project and maintain. This concept has profoundly influenced politeness research, providing a valuable framework for understanding how individuals navigate social interactions and employ linguistic strategies to mitigate potential face threats.

Within the realm of second language acquisition (L2), numerous studies have demonstrated that learners often encounter difficulties in producing politeness markers appropriately (e.g., Aini et al., 2023; Dufon, 2008; Kartini et al., 2024; Kasper, 1997; Kasper & Rose, 2002). Kasper and Rose (2002) posit that even lexically and grammatically proficient learners may struggle to effectively employ politeness markers due to the significant cognitive demands involved. Pragmatic competence, they argue, requires careful consideration of social context, interlocutor relationships, and the potential impact of utterances on others’ face needs.

A cornerstone of politeness research lies in the categorization of politeness markers. While various taxonomies exist (e.g., Brown & Levinson, 1987; Crystal & Davy, 1975; Edmondson, 1977; Holmes, 2000), the typology proposed by House and Kasper (1981) remains highly influential. Their framework identifies eleven subcategories, including hedges, apologies, and compliments, providing a comprehensive overview of the linguistic strategies employed to signal politeness and respect:

1. Politeness markers (PM): Some expressions like “please” which demonstrate the speakers’ respect to the addressee and display collaborative manner.
2. Play-downs (PD): They include some syntactic patterns like past tense (e.g., “I wondered if”), continuous accompanied by past tense, a modal verb accompanied by interrogative, and a modal accompanied by negative interrogative. Such patterns soften the perlocutionary impact that an expression is likely to leave on the addressee.
3. Consultative devices (CD): Statements such as “Would you mind” and “Could you” which demand the addressee to cooperate and involve him/her.
4. Hedges (HD): Tools like “somewhat”, “rather”, and “kind of” by which the interlocutors make an expression more ambiguous or shun providing a precise propositional content and allow the addressee to maintain their own intention.
5. Understaters (US): Adverbial mitigators such as “a moment”, “a bit”, and “a second” that are utilized by the interlocutors to lower the imposition of a remark or expression by underrepresenting the propositional content of an expression.
6. Downtoners (DT): Adverbials like “maybe”, “perhaps”, and “possibly” that can be used by the interlocutors to decrease the impact of an expression.
7. Committers (CS): Statements like “I guess”, and “in my opinion” that are employed to reduce the interlocutor’s commitment and obligation to the propositional content of an expression.
8. Forewarning (FW): Strategies such as complementing or using expressions like some sort of metacomments on a face threatening act made by the interlocutor.
9. Hesitators (HS): Non-lexical phonetic materials such as “er”, “uhh”, and “ah” that can be utilized by the interlocutors in the conversation to fill pauses.
10. Scope-staters (SS): Statements that aim at indicating the interlocutor’s personal idea about the subject matter of a debate or discussion.
11. Agent avoiders (AA): The application of passive patterns in sentences such as “people don’t do X” in which the doer is not present (Watts, 2003).

Measures of Pragmatic Competence

A mounting interest in pragmatic research and speech acts has coincided with adopting different measures (e.g., Bardovi-Harlig & Shin, 2014; Ellis et al., 2024; Golato, 2003; Mohammad Hosseinpur et al., 2021; Yuan, 2001). Such analyses have delved into various measures of pragmatics to come up with the best possible methodology. Methodologies like WDCT, ODCT, and Role-plays are among the most commonly-used measures that can be adopted by the researchers within the field of pragmatics.

Naturally-Occurring Data

To glean natural data, the genuine talk-in interactions of the learners in the actual settings are observed and field notes are recorded online or with audio-visual apparatus to be transcribed for further thorough examination (Taguchi, 2018). The inherent advantage of such approach rests in the genuineness of the obtained data and the fact that it reflects the actual characteristics of the authentic collaborations (Schauer, 2009). Nevertheless, the existence of the observer or audio-visual tool can undermine and lessen the genuineness and accuracy of the information and cause what Labov termed the "observer’s paradox" (House, 2018). Absence of researcher’s control over the extraneous variables under examination (Bardovi-Harlig, 2018) and the influence of inappropriate elements on the final outcomes such as power, social interaction, distance, position, sex, and age discrepancies are the major constraints of this method (Taguchi & Roever, 2017).

Written Discourse Completion Task

In Written Discourse Completion Task (WDCT), students are provided with the explanation of a scenario and are asked to present a suitable response for the depicted scenario. This tool enjoys high popularity among interlanguage pragmatics (ILP) researchers since it can be readily and easily administered, makes an enormous extent of data elicitation possible within a short amount of time, and enables the investigators to exert control over the study’s variable as well (Bardovi-Harlig, 2018; Golato, 2003). Nevertheless, such approach suffers from its own constraints and drawbacks. The non-authentic and non-interactive nature of the collected data via WDCTs are the most crucial flaws of this data elicitation technique. Previous studies (e.g., Bardovi-Harlig, 2018) point to the fact that there is a failure on the part of respondents to actually mirror and manifest the genuineness of spoken interactions in their written responses and the elicited written data differs from the genuine interactions.

Oral Discourse Completion Task

Oral Discourse Completion Task (ODCT) can be regarded as a particular pragmatic apparatus through which learners can be asked to collaboratively listen to the situations from a recorder and then pass on their verbal and spoken answers to another tape recorder (Brown, 2001). ODCTs fail to be truly reflective of all the inherent characteristics of the natural interactions since they are merely a verbal and spoken answer to a spoken question. As a matter of fact, no genuine interaction is established between the one who poses the question and the ones who provides the response. Thus, ODCTs suffer from all the constraints of the WDCTs except its dispatch mode and its capability to inspect and securitize speech characteristics that are the primary reasons for the essence and presence of this data-collection instrument (Taguchi, 2018). Furthermore, recording tools will have to be utilized and collecting data through ODCTs is a time-consuming process.

Role-play

Role-plays are considered an appropriate option because the intrinsic drawbacks of the natural methodology, like the vastness of scope in analyzing a specific facet and absence of genuineness, flexibility, and versatility of the gathered data via WDCTs and ODCTs could be either mitigated or kept in check by the researchers (Bardovi-Harlig, 2018; Taguchi, 2018). Moreover, role-plays present language learners with a good chance to be cognizant of each scenario’s details including the intent and social status of the interlocutors (Félix-Brasdefer, 2010).

Though RPs can be viewed as advantageous over ODCTs and WDCTs because of their proximity to natural methodology and having more researcher control over the variables, such pragmatic measures typically take a fair amount of time to complete (Eslami-Rasekh & Mirzaei, 2014). However, the capability to actually play roles and visualize oneself in specific circumstances might be well beyond the scope and limits of the average language learner (Kasper, 2000; Schauer, 2009). To the best knowledge of the researchers, no study has been carried out specifically on the representation of politeness markers across different measures of pragmatic competence. Therefore, the researchers formulated the following research questions:

• Does WDCT elicit the same data, in terms of politeness markers, as Natural Methodology does?
• Does ODCT elicit the same data, in terms of politeness markers, as Natural Methodology does?
• Does Role-play elicit the same data, in terms of politeness markers, as Natural Methodology does?

METHOD

Participants

To carry out the research, 56 language learners’ dialogues and institutional exchanges with their instructors in a language academy in the Iranian EFL setting were gathered as the natural data. After examining the transcribed data, 27 learners were identified as meeting the inclusion criteria for the research, and they accompanied the investigators in other phases of the study. The stipulated requirement for the participants to be included in the study was to make various requests. The participants’ overall language proficiency was intermediate. Participants were classified as intermediate (B1 or B2 levels on the CEFR scale) based on their performance on the test. Specifically, a score of 60-79 on our 100-point scale was considered intermediate, aligning with the CEFR B1 and B2 criteria. Their age ranged from 21 to 32.

Instruments

Four measures of pragmatic competence (Natural methodology, Role-play, ODCT, and WDCT) were employed to conduct this research. To gather the natural data, the investigators recorded the learners’ talk-in interactions. Next, they examined the data to identify and pinpoint the most significant characteristics of the elicited data so as to pick the most suitable one. This was done to guarantee that the chosen scenarios were truly reflective of the natural conversations’ features and to make the analogy among the aforementioned instruments more tenable and persuasive.

The commonness and popularity of the requests prompted the researchers to pick requests with these appropriate characteristics in the natural data. Similarly, the scenarios for the other three measures (Role-plays, ODCTs, and WDCTs) were chosen according to the same circumstantial characteristics. As a result, each learner made four requests for each measure and each participant generated 16 requests altogether in this study. Social status refers to an interactant’s standing in relation to other individuals within the classroom context, and imposition signals the burden imposed by a request.

Data Collection Procedure

All the learners received fifteen weeks of instruction to finalize the research. First, the talk-in interactions of the learners were recorded during 12 weeks. The first three weeks were not included in the data collection procedure because of the observer paradox. Furthermore, 27 out of 56 learners generated various requests in the natural data. Accordingly, these 27 learners ultimately fulfilled the stipulated requirement to continue with the rest of the study.

When the investigators were done with collecting the natural data, they involved the learners in role-plays in the 13^th week. All the learners were invited to the class and the written scenarios were handed out to one of the student in the pair to be properly studied before making a request. The other learner was merely there to respond to the requesters’ request off the cuff. As for the first two scenarios, the two role-players were learners due to the low-status of the scenarios. As for the next two scenarios, one of the players had to be the teacher. The instructor had to play requestee’s part due to the high-status inherent in the presented scenarios. Each scenario was carefully read by the requester who then made their request individually. The researchers recorded and transcribed the interactions of the participants for later in-depth and thorough analysis. When the researchers were done and over with recording the scenarios, they invited the learners to the class individually to produce requests according to the recorded scenarios of the ODCTs. The learners produced requests modelled on the presented scenarios. The researchers then recorded and transcribed the students’ provided answers for further analysis. In the 15^thweek, the researchers administered the WDCTs whose scenarios were different from those included in the Role-plays and ODCTs, but enjoyed similar contextual characteristics of status and imposition. The learners were then required to go through the scenarios and provide their responses in the blank space right below them. Finally, all the data collected through four measures of pragmatic competence were analyzed and thoroughly examined based on House and Kasper’s (1981) classification of politeness markers to ascertain how representative of politeness markers each measure was and to discern how differently and similarly politeness markers were represented by each measure.

RESULTS

Exploring Research Question 1

Concerning this question, Table 1 displays the statistical analyses for the politeness markers elicited through WDCT and Natural Methodology. The statistics are descriptive indices, based on which no conclusion can be reached regarding any significant difference between cell frequencies, and cells can be compared modelled on the descriptive statistics. If Std. Residuals are higher than 1.96, it can be concluded that the observed frequency was significantly beyond what was expected; and if Std. Residuals are higher than -1.96, it can be claimed that the observed frequency was significantly below expectation.

            Modelled on the outcomes illustrated in Table 1, it can be said that WDCT elicited play downs (1 %, Std. Residual = -2.1 > -1.96) substantially less than Natural Methodology (8.1 %, Std. Residual = 2.4 > 1.96), and Natural Methodology elicited politeness markers (8.7 %, Std. Residual = -2.2 > -1.96) significantly less than WDCT method (21.1 %, Std. Residual = 1.9 < 1.96). WDCT elicited consultative devices (37.6 %, Std. Residual = 2.5 > 1.96) significantly more than Natural Methodology (15.4 %, Std. Residual = -2.9 < -1.96) and also it elicited scope-starters (32 %, Std. Residual = 2.4 > 1.96) remarkably greater than Natural Methodology (12.8 %, Std. Residual = -2.7 < -1.96). Natural Methodology, on the other hand, elicited hesitators (50.3 %, Std. Residual = 7.1 > 1.96) significantly more than WDCT method (1.5 %, Std. Residual = -6.2 < -1.96).

Table 1 also reveals that no remarkable variance existed between frequencies of understaters elicited through WDCT (3.1 %, Std. Residual = -.5 < -1.96) and Natural Methodology (4.7 %, Std. Residual = .6 < 1.96), downtoners elicited under WDCT (2.1 %, Std. Residual = 1.2 < 1.96) and Natural Methodology (0 %, Std. Residual = -1.3 < -1.96), committers elicited through WDCT (.5 %, Std. Residual = .6 < 1.96) and Natural Methodology (0 %, Std. Residual = -.7 < -1.96), and agent avoiders elicited under WDCT (1 %, Std. Residual = .8 < 1.96) and Natural Methodology (0 %, Std. Residual = -.9 < -1.96). And finally, WDCT and Natural Methodology did not elicit any hedges.

Table 1. Frequencies, Percentages and Std. Residuals of Politeness Markers in WDCT vs. Natural Methodology

Table 2 exhibits the outcomes of Chi-square test. The analyses (χ2 (8) = 140.58, p = .000) indicated that substantial variances existed between the frequencies of politeness markers elicited under WDCT and Natural Methodology. The effect size was .640 that demonstrates a great extent (Cramer’s V^[1] = .640).

Table 2. Chi-Square Tests of Politeness Markers for WDCT vs. Natural Methodology

Table 3. Fisher’s Exact Test of Politeness Markers for WDCT vs. Natural Methodology

The following figure exhibits the representation pattern of politeness markers across WDCT and Natural Methodology.

Figure 1. Representation of Politeness markers in WDCT vs. Natural Methodology

Exploring Research Question 2

Concerning this question, Table 4 exhibits the statistical analyses for the politeness markers elicited under ODCT and Natural Methodology. It can be stated that no remarkable difference existed between frequencies of politeness markers elicited under ODCT (14.1 %, Std. Residual = .9 < 1.96) and Natural Methodology (8.7 %, Std. Residual = -1.1 < -1.96), understaters elicited through ODCT (3.2 %, Std. Residual = -.5 < -1.96) and Natural Methodology (4.7 %, Std. Residual = .6 < 1.96), scope-starters elicited under ODCT (24.5 %, Std. Residual = 1.6 < 1.96) and Natural Methodology (12.8 %, Std. Residual = -1.9 < -1.96), and  agent avoiders elicited through ODCT (.5 %, Std. Residual = .5 < 1.96) and Natural Methodology (0 %, Std. Residual = -.6 < -1.96). Table 4 also shows that ODCT elicited play downs (.9 %, Std. Residual = -2.2 > -1.96) significantly less than Natural Methodology (8.1 %, Std. Residual = 2.7 > 1.96). On the other hand, ODCT elicited consultative devices (33.2 %, Std. Residual = 2.1 > 1.96) considerably greater than Natural Methodology (15.4 %, Std. Residual = -2.5 < -1.96). Finally, Natural Methodology elicited hesitators (50.3 %, Std. Residual = 3.3 > 1.96) considerably higher than ODCT (23.6 %, Std. Residual = -2.7 < -1.96). It is evident from the table that ODCT and Natural Methodology did not elicit any downtoners, committers, and hedges.

Table 4.

Frequencies, Percentages and Std. Residuals of Politeness Markers in ODCT vs. Natural Methodology

Table 5 displays that the effect size was .371, representing a moderate effect size (Cramer’s V = .371).

Table 5. Chi-Square Tests of Politeness Markers for ODCT vs. Natural Methodology

Figure 2 illustrates the summary of the above-mentioned information. It compares and contrasts politeness markers representation across ODCT and Natural Methodology.

  Figure 2. Representation of politeness markers in ODCT vs. Natural Methodology

Exploring Research Question 3

This question intended to explore whether Role-play could elicit the same data as Natural Methodology with regard to politeness marker representation.  Table 6 displays the descriptive statistics for the politeness markers elicited through Role-Play and Natural Methodology.

Modelled on the outcomes exhibited in Table 6, it could be assumed that there was not any meaningful difference between frequencies of politeness markers elicited under Role-Play (12.9 %, Std. Residual = .7 < 1.96) and Natural Methodology (8.7 %, Std. Residual = -1 < -1.96), play downs elicited through Role-Play (5.3 %, Std. Residual = -.6 < -1.96) and Natural Methodology (8.1 %, Std. Residual = .8 < 1.96), consultative devices elicited under Role-Play (26.6 %, Std. Residual = 1.4 < 1.96) and Natural Methodology (15.4 %, Std. Residual = -1.8 < -1.96), hedges elicited under Role-Play (.4 %, Std. Residual = .5 < 1.96) and Natural Methodology (0 %, Std. Residual = -.6 < -1.96), understarters elicited under Role-Play (5.3 %, Std. Residual = .2 < 1.96) and Natural Methodology (4.7 %, Std. Residual = -.2 < -1.96), and scope-starters elicited through Role-Play (22.8 %, Std. Residual = 1.3 < 1.96) and Natural Methodology (12.8 %, Std. Residual = 1.8 < -1.96).

Table 6.

Frequencies, Percentages and Std. Residuals of Politeness Markers in Role-Play vs. Natural Methodology

 Natural Methodology elicited hesitators (50.3 %, Std. Residual = 3.1 > 1.96) significantly more than Role-Play method (26.6 %, Std. Residual = -2.3 < -1.96). And finally, Role-Play and Natural Methodology did not elicit any downtoners and committers. Table 7 demonstrates a weak to moderate effect size which was .264.

Table 7. Chi-Square Tests of Politeness Markers for Role-Play vs. Natural Methodology

Figure 3 demonstrates how similarly and differently politeness markers are represented in Natural Methodology and Role-play.

Figure 3. Representation of politeness markers in Role-Play vs. Natural Methodology

DISCUSSION

Regarding the objective of this investigation, the findings pointed to remarkable disparities between the data-gathering techniques in terms of politeness markers representation. Concurrent with the results of several analyses (e.g., Golato, 2003; Mohammad Hosseinpur et al., 2021; Turnbull, 2001; Yuan, 2001), the findings indicate that despite some commonalities, such techniques fail to gather the same data as NM with regard to politeness markers. In other words, politeness markers in WDCTs, ODCTs, and Role-plays are differently represented compared to their representation in Natural Methodology.

It seems that, due to the spontaneous, ad hoc and haphazard nature of the interfaces, and the speed of the unfolding affairs in natural situations, the pupils have relied upon their automatic knowledge to produce requests. The prevalence of hesitators, for example, suggests that the participants were mainly preoccupied with the sociopragmatic considerations in naturally-occurring situations when producing requests. On the other hand, the frequent application of consultative devices, politeness markers and scope-stators can be explained by the participants’ dependence on their declarative knowledge. Since the learners were given ample opportunity to go through and think over their responses for the scenarios, they could readily employ more consultative devices and scope-staters in their requests in comparison with the participants engaged in natural ongoing interactions, in which they had to call upon their extemporaneous capability to generate English requests.

Another possible justification for the overuse and frequency of some markers such as consultative device and scope-stator politeness markers in the previous-stated measures might be ascribed to the test/task-like nature of such measures compared with Natural methodology (Sasaki, 1998). There is an element of formality in WDCT, ODCT, and Role-play which could have contributed to the overall popularity of consultative devices and scope-stators in comparison with the Natural methodology which had fewer of both.

Yet another likely reason for significant and marked differences in politeness markers representation through WDCTs, ODCTs, and Role-plays compared with Natural methodology can be ascribed to the data collection situation. When making requests in the WDCTs, ODCTs, and Role-plays, the subjects need to visualize the proposed positions and put themselves in those situations, and not all the interactants possess this capability (Schauer, 2009). Under real-life situations, the interactants are truly involved in the interactions and there exists no need to imagine a situation. Fewer utilization of hesitators in the WDCTs, ODCTs, and Role-plays, for instance, can be logically explained by the artificiality that is inherent while making requests through such measures of pragmatic knowledge.

Transfer from the participants’ L1, Persian, might justify the frequency of some markers such as consultative devices and scope-staters in all measures of pragmatic knowledge, especially WDCT, ODCT, and Role-play. In the Iranian culture, requesters usually attempt to be courteous to display reverence as follows:

Excuse me! would you mind repeating the statement again? I couldn’t hear it. It is too noisy in here.

As shown in the above excerpt, Iranians typically present reasons and justifications for their requests, especially in high-status circumstances, to display both their deferential attitude and ensure that they are generating the right requests. Seemingly the formal element existing in the context has compelled the interactant to explain his/her request by prolonging it (Woodfield, 2012).

Finally, findings were indicative of the fact that learners were less prone to use some politeness markers such as understators, play-downs, and hedges in their interactions. This underuse or so-called unintentional underestimation of such politeness markers on the part of participants in the four measures of pragmatic competence might possibly be attributed to their unfamiliarity with such politeness markers, lack of proper L2 instruction, or an offshoot of negative transfer from their L1.

CONCLUSION AND IMPLICATIONS

Generally, it can be concluded that WDCTs, ODCTs, and Role-Plays could be more appropriate and advantageous to draw on and utilize learners’ declarative knowledge, while NM could be more suitable to capitalize or draw on procedural knowledge. If the objective is to probe the sociopragmatic aspect of the language, NM, viewed as the most intricate data collection procedure (House, 2018), best lives up to the expectations and meets the demand since the interactants do their best to comply with social regulations and norms rather than to consider the pragmalinguistic dimension during natural interactions. If a different kind of speech act had been employed, different outcomes might have ensued. Furthermore, the data type elicited through different measures of pragmatic knowledge might be either directly or indirectly impacted by the data collection procedure design. Nevertheless, the proximity of one pragmatic data gathering technique to Natural method does not necessarily indicate its superiority or advantage over other techniques. Research questions and researchers’ objectives should be taken into account to make the most appropriate and informed choice about the adopted methodology (Yuan, 2001).

Contextual information plays a pivotal part in eliciting pragmatic knowledge (Schauer, 2009). The data type elicited through each pragmatic measure can be directly influenced by the amount of contextual information included in each measure. Comparing and contrasting two various versions of such data-gathering techniques with differing levels of contextual information to discover how they differ in terms of politeness markers representation can be a promising area for further research. Utilizing mixed-methods designs (House, 2018) and matched modality tasks (Bardovi-Harlig, 2018) to investigate politeness markers can be another interesting venue for future investigation.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID