Exploring Pragmatic Competence in Formal Communicative Contexts: The case of thanking by native and non-native speakers of Greek

espanolEl objetivo de este estudio preliminar es el de explorar el acto de habla del agradecimiento en contextos comunicativos formales. Mas especificamente, se investigan las posibles diferencias en la expresion del agradecimiento por parte de hablantes nativos y no nativos de griego. Los participantes fueron treinta y un (N = 31) estudiantes de griego como segunda lengua (L2) que pertenecian a cinco niveles de competencia diferentes (de A2 a C2) matriculados en un curso intensivo de griego de verano en Atenas, Grecia. Ademas, participo un grupo de hablantes nativos de griego (N = 30). Los datos orales se obtuvieron a traves de una serie de juegos de rol abiertos que representaban tres situaciones comunicativas formales con diferentes parametros sociales. Tambien se llevo a cabo un protocolo verbal retrospectivo con el grupo de alumnos de L2. Los analisis se centraron en el tipo y la cantidad de estrategias utilizadas por los dos grupos. Los resultados mostraron una diferencia en el tipo y el numero de estrategias que emplearon estos dos grupos. Tambien parecen indicar que el entorno social y la direccion de imposicion fueron factores cruciales para la expresion del agradecimiento de ambos grupos. La expresion de los alumnos tambien se vio afectada por su familiaridad con cada situacion. EnglishThe aim of the current study is to investigate the role of second language (L2) proficiency in the pragmatic competence of learners of L2 Greek. More specifically, it aims at exploring whether L2 proficiency can have an impact on the strategies that L2 learners use in order to express the speech act of thanking in L2 Greek. Participants were thirty-one (N=31) learners of Greek at different proficiency levels (from A2 to C2) who were enrolled in a summer intensive course of Greek at a university language school in Athens, Greece. Additionally, a group of native speakers of Greek has been included as a baseline (N=30). Oral data were elicited through a series of open role plays which represented various communicative situations with different social parameters. Furthermore, a retrospective verbal protocol has been used in order to gain more insights into the way L2 learners perform the speech act of thanking. Results showed that there is a difference in the number of strategies of L2 learners across the different proficiency levels, which is more evident between beginners and advanced learners. However, findings seem to suggest that the role of proficiency is not clear-cut and it also depends, firstly, on the social parameters of each communicative situation and also on learners’ familiarization with each situation. The study concludes by discussing some implications for L2 pedagogy. catalaL’objectiu d’aquest estudi preliminar es el d’explorar l’acte de parla de l’agraiment en contextos comunicatius formals. Mes especificament, s’investiguen les possibles diferencies en l’expressio de l’agraiment per part de parlants nadius i no nadius de grec. Els participants van ser trenta i un (N = 31) estudiants de grec com a segona llengua (L2) que pertanyien a cinc nivells de competencia diferents (de A2 a C2) matriculats en un curs de grec intensiu d’estiu a Atenes, Grecia. A mes a mes, hi va participar un grup de parlants nadius de grec (N = 30). Les dades orals es van obtenir a traves d’una serie de jocs de rol oberts que representaven tres situacions comunicatives formals amb diferents parametres socials. Tambe es va portar a terme un protocol verbal retrospectiu amb el grup d’alumnes de L2. Les analisis es van centrar en el tipus i la quantitat d’estrategies utilitzades pels dos grups. Els resultats van mostrar una diferencia en el tipus i el nombre d’estrategies que van emprar aquests dos grups. D’altra banda semblen indicar que l’entorn social i la direccio d’imposicio van ser factors crucials per a l’expressio de l’agraiment d’ambdos grups. L’expressio dels alumnes tambe es va veure afectada per la seva familiaritat amb cada situacio.


Introduction
C ommunicative competence (Hymes, 1972) is considered a fundamental purpose of language acquisition and teaching. When interacting, speakers need to employ a variety of linguistic means in a manner that is appropriate for each communicative situation (Usó-Juan & Martínez-Flor, 2008). In other words, they need to develop their pragmatic competence-an ability which is integral part of the communicative competence (Canale, 1983;Canale & Swain, 1980). Pragmatics and Interlanguage Pragmatics (ILP)-the field of Applied Linguistics that explores how learners of second language (L2) develop their L2 pragmatic competence-have typically focused on the analysis of speech acts (Searle, 1969). Speech acts constitute "actions performed via utterances" (Yule, 1996, p. 47); the ability to successfully comprehend and produce them is a major component of the pragmatic competence. However, empirical evidence has shown that acquiring L2 pragmatics can be a challenging task for L2 learners and that the development of pragmatic competence does not always follow the development of the grammatical one (Bardovi-Harlig, 1999). Moreover, it has been argued that native speakers (NS) and non-native-speakers (NNS)/learners of a language tend to present differences in the way they produce speech acts (Bella, 2014(Bella, , 2016)-and also that their speech acts differ in the amount of strategies used Ozdemir & Rezvani, 2010). Different speech acts have been analyzed extensively during the last decades. Nevertheless, thanking has received less attention so far. Thanking is a communicative strategy frequently employed in everyday interaction in many cultures. Either because of its usually brief and simple structure or because of its responsive nature, it is usually regarded as a secondary and peripheral language component and is one of the most neglected parts of language research and teaching (Hinkel, 1994a, p. 2). However, it is an important component of language use and pragmatic competence and the way it is expressed is closely linked to interpersonal rapport and politeness.
The purpose of the current preliminary study is to try to fill this gap in the literature by further examining the possible differences in the expression of the speech act of thanking by NS and NNS of Greek-an under-explored target language. More specifically, it aims at investigating whether there are any differences in the type of the speech act and the amount of strategies that these two groups use in three different formal communicative situations. This study builds on previous research carried out by Gkouma (in press) and Andria et al. (in press) and it is part of the LETEGR2 project, a larger project focusing on the acquisition and the teaching of Greek as an L2.
The following section (Section 2) offers a review of the literature, with reference to pragmatic competence in L2 acquisition (Section 2.1). The speech act under analysis (thanking) is discussed in Section 2.2, while Section 2.3 presents the research question that guides the current study. Next, Section 3 explains the methodology and Section 4 examines the results of the analyses. The results are discussed in Section 5, while Section 6 contains some concluding remarks.
A nuclear point of interest in the to-date research in the field of ILP is the comparison of the pragmatic competence of NS and NNS. Empirical studies have shown that there are differences between the pragmatic systems of these two groups (Barron, 2003;Díaz-Pérez, 2005;Félix-Brasdefer, 2007;Flores Salgado, 2011;Göy et al., 2012;Schauer, 2009;Trosborg, 1995;Woodfield, 2012). Research also indicates that the L2 proficiency level seems to be an important factor for the pragmatic success or failure of the L2 learners' communicative attempts, although pragmatic competence does not always follow the development of the grammatical one.

The speech act under analysis: Thanking
In this study, we investigated the pragmatic competence through the production of the speech act of thanking in formal communicative situations. Thanking is a fundamental speech act with high social importance, as it constitutes a widely used language component in every-day communication, both formal and informal. Prototypically, it expresses gratitude and constitutes a reactive speech act. This means that it is always preceded by an action or utterance that calls for gratitude or acknowledgement (Coulmas, 1981;Haverkate, 1993). Thanking is frequently used in daily encounters and is highly formulated by and subject to cultural norms (Coulmas, 1981;Hinkel, 1994a). In many cultures, Greek included, it is explicitly taught to children from an early age and is regarded as an important element of communicatively adequate and polite behavior. Even though it bears little information, it reinforces social relations (Bodman & Eisenstein, 1988;Eisenstein & Bodman, 1993;Jung, 1994;Lakoff, 1973) and it reflects each individual's perception of cultural norms. It is also important to mention that usually thanking is reinforced with supportive moves, in order to underline the sincerity of the gratitude, the appreciation or the emotional engagement (Aijmer, 1996;Leech, 1983), and that social parameters affect its realization (Díaz Pérez, 2005).
Despite its importance to the success of communication, thanking is an under-explored speech act, in comparison to others (for example, requests or refusals), especially from the SLA point of view. There has been an increasing interest lately for this speech act (Bardovi-Harlig et al., 2008;De Pablos-Ortega, 2010, 2015Demir & Takkak, 2016;Díaz Pérez, 2005;Kia & Salehi, 2013;Lanteigne & Crompton, 2011;Ozdemir & Rezvani, 2010;Schauer & Adolphs, 2006, among others), but more research is still necessary. Important cross-cultural differences related to this superficially simple and peripheral aspect of L2 are sometimes ignored or overlooked (Hinkel, 1994a, p. 2). The acquisition of thanking by L2 learners is usually regarded as easy to learn, since it is one of the first linguistic expressions that beginners are exposed to. However, expressing this speech act in a communicatively appropriate way seems to be a more challenging task for L2 learners (Schauer & Adolphs, 2006). This is, first, due to its expressive and reactive nature that imposes instant and proper utterance production and second, to its dependence on social parameters. Previous studies have shown that even high-proficient L2 learners encounter difficulties in the production of this speech act (Bodman & Eisenstein, 1988;Eisenstein & Bodman, 1993;Hinkel, 1994b;Ozdemir & Rezvani, 2010).
Regarding the target language under analysis, Greek, research on thanking is very limited. Gkouma (in press) and Andria et al. (in press) have recently presented, within the framework of the LETEGR2 project, a first empirical approach on the acquisition of thanking by L2 learners of Greek. Gkouma (in press), in her qualitative study, has examined the way NS and NNS of Greek perform the speech act of thanking in a variety of communicative situations. She also presented a data-driven taxonomy for thanking in Greek. Findings have shown that NS of Greek either used multiple strategies to reinforce their expression of thanking in the majority of the contexts under examination or, in informal situations, usually opted for an indirect strategy to perform the speech act of thanking. As far as L2 learners of Greek are concerned, they showed progress from beginner to advanced levels. However, even the advanced learners lagged behind native speakers' performances. In the same line, Andria et al. (in press) adopted a quantitative approach in order to investigate thanking performance of NS and NNS on a series of different formal and informal contexts.
Building on the two above-mentioned studies (Andria et al., in press; Gkouma, in press), the current research aims to delve CLIL Journal of Innovation and Research in Plurilingual and Pluricultural Education, 3(2), 2020: 57-70 more into the above mentioned LETEGR2 database and further explore the expression of thanking after a request by NS and NNS of Greek in formal communicative situations. The focus on the analysis of formal communicative contexts was motivated by the fact that they are usually regarded as a communicative challenge both for NS and for NNS.

Research question
In light of the literature review presented above, the research questions that guide this study are the following:

RQ1:
Do native speakers and L2 learners of Greek differ in the type of strategies that they employ for the speech act of thanking in formal communicative situations?

RQ2:
Are there any differences in the number of strategies that native speakers and L2 learners of Greek use for the expression of the speech act of thanking in formal communicative situations?
Based on previous empirical findings (Bella, 2012a(Bella, , 2014(Bella, , 2015Félix-Brasdefer, 2007;Gkouma, in press;Ozdemir & Resvani, 2010), it could be hypothesized that there will be differences between these two groups. In addition, differences are also expected among the various proficiency levels.
In the next section, we will present the method that was followed in order to answer these research question.

Participants and context
T hirty-one (N=31) L2 learners of Greek participated in the study (4 male, 27 female). Their age varied from 17 to 48 years (M=24.27 years old, SD=7.2) and they were of different first-language (L1) backgrounds (See Appendix A for L2 learners' countries of origin). They were attending a five-week intensive summer course at a university language school in Athens, Greece. Prior to the beginning of the course, the participants had taken a placement test. The test was administered and evaluated by the teaching staff of the language school. Based on the results of this placement test, the participants were assigned to a specific proficiency level group. It should be noted that the classification per proficiency level follows the criteria described by the Common European framework of reference for languages (CEFR) (Council of Europe, 2001). In addition to the L2 learners' group, a group of native speakers of Greek (N=30) was recruited in order to obtain a native baseline for the speech act under analysis.

Instruments
For the data collection, three instruments have been used: a series of role plays, a retrospective verbal report and a questionnaire. The role plays have been designed specifically for the purposes of a broader research investigating the pragmatic competence of L2 learners of Greek. Drawing on this database, for the purpose of the current study a series of three role plays that concern formal communicative contexts was selected (see more information below). Role plays as an elicitation technique for studies in Pragmatics, and especially for studies focusing on speech acts, have been found to present several advantages (Alcón-Soler & Martínez-Flor, 2008; Cohen & Olshtain, 1993) the most important ones being that enable the researcher to control a series of contextual parameters and that they have interactive nature (Félix-Brasdefer, 2010, p. 47;Yuan, 2001, p. 284).
Despite the simulating character of this elicitation technique, role plays have the dynamic of interaction that lacks in Discourse Completion Tests (DCT)-the most popular instrument in ILP. Their interactive nature is particularly important in the investigation of expressive and reactive speech acts, such as thanking, which have to be instant and spontaneous. It should be noted that in this study open role plays were used; participants were given a general scenario of a formal communicative situation (in which they had to request something) without predetermining the course and the outcome of the conversation, in order to examine if and how the speech act of thanking is performed in every situation.
Since data collection was part of a larger project on Pragmatics, participants were asked to play a role with another interlocutor (a researcher) in a total of nine (9) communicative situations. Six (6) of them were target situations, that is, they were designed to elicit the speech act of thanking and three (3) were distractors. As far as the six (6) target situations are concerned, they included formal and informal contexts. However, as it was mentioned above, in the current study we will focus our analysis only on three of those situations, namely those that concern formal contexts.
All three formal communicative situations under examination had the social parameters of distance and power in common, but differed in terms of the direction of the imposition and the social setting of interaction. More specifically, in the three role plays under examination the L2 learners interacted as "boss", "employee" and "university student" (see Appendix B). In the case of "boss", the participant dominates over the researcher while in the case of "employee" the researcher dominates over the participant. Finally, in the case of "student" the researcher again dominates over the participant, but now the social setting is different; it is an academic context while the former two were taking place in a working context. In other words, superficially the situations of boss, employee and student have the same combination of social parameters. However, in the case of "boss" the social power is reversed, in contrast to employee and student. Furthermore, the roles of employee and student differ in the social setting (working versus academic context). The reliability of the instrument was checked and it was found that it had high internal consistency, with a Cronbach α coefficient of .801.
The L2 learners were also asked to provide retrospective verbal report. This instrument was used in order to shed more light on and complement the oral data from the role plays, as it allowed us to obtain information on the learners' pragmatic and sociocultural awareness (Félix-Brasdefer, 2010, p. 53). Finally, a questionnaire was used to elicit the participants' biodata and background information.

Data collection
For the L2 learners, data collection took place during the first week of the intensive course in a quiet room of the language school. Before the data collection, the participants were asked a few ice-breaking questions. After that, a researcher explained the instructions for the role plays. The situations were presented to each participant with a different order. The scenario of each situation was presented in a colored card. Apart from the brief written instructions, a visual input (picture prompt) was also included in each card. The picture prompt was used in order to facilitate the understanding of each situation, especially for participants of initial levels whose reading skills in Greek were still limited (Alcón-Soler & Martínez-Flor, 2008, p. 215). Once the series of role plays was over, the retrospective verbal report was carried out. The oral data were voice-recorded. It is important to mention that before the data collection all the participants had been informed about the study and they had given their In order to answer the first research question, the percentage of each of the above-mentioned categories was counted for native and non-native speakers in each communicative situation under analysis.
For the second research question, the total number of strategies that was used by each participant in each situation was counted. The total number of strategies included the head acts, as well as all the internal and external modifications.
I knew that you wouldn't disappoint me (external modification)! Thanks (head act) a lot (internal modification).
The total amount of strategies in this example is three (3).
The information that was elicited by means of the retrospective verbal report was also transcribed and was used for the qualitative analysis of the data.

Statistical analyses
For the statistical analyses, descriptive statistics were computed, first, for the type of each act and, second, for the number of strategies that were used in each situation by the participants. Next, with regard to the number of strategies, a preliminary ANOVA was run in order to better understand the differences presented by NS and NNS in the production of thanking. For the statistical analyses, the Statistical Package of Social Sciences (SPSS 25) was used.
permission by signing a consent form. The questionnaire was administered at the end of the process. In addition, data were obtained by 30 native speakers of Greek as a baseline.

Coding
After the data collection, the oral data were transcribed. Next, the participants' strategies were categorized using the categorization system of thanking in Greek suggested by Gkouma (in press). This categorization is, to our knowledge, the first one that attempts to present a classification of thanking strategies in Greek. Its creation was data-driven, but it also incorporates elements of previous categorizations in the field (Aijmer, 1996;Blum-Kulka, et al., 1989;Haverkate, 1984). In this categorization, thanking performance is identified in six (6) different types: 1. Simple direct strategy of thanking (use of performative statement for gratitude in Greek ευχαριστώ "thank you") 2. Direct thanking strategy with external modification (use of supportive moves) 3. Direct thanking strategy with internal modification (use of internal sentence components) 4. Direct thanking strategy with combination of external modification and internal modification 5. Indirect thanking strategy (absence of performative statement which is substituted by a supportive move)

Results
O ur first research question asked whether there are any differences in the type of speech act that NS and L2 learners use for the expression of thanking. Table 4 below offers the descriptive statistics for each role play. The L2 learners' results are presented per proficiency level. Table 4 indicates that NS mainly use a combination of external and internal modifiers in all contexts and sparsely opt for indirect or only internally or externally modified thanking strategies. None of them uses a simple head act or avoids thanking. In the L2 learners' groups there is a greater variety of strategies. Beginners usually resort to simple direct strategy or end the role play without thanking. The majority of the thanking strategies of intermediate level students were only internally modified, while some of them were both internally and externally modified. Finally, high-proficient students' behavior was more similar to NS'; however, it appeared that there were still remarkable divergences.  Our second research question sought to examine whether there are any differences in the number of strategies that NS of Greek and L2 learners use for the speech act of thanking.  Then, Figure 1 presents the descriptives graphically. According to the descriptive statistics, NS present the highest number of strategies in all the three formal communicative situations. Regarding L2 learners, levels A2 and B1 behave in a quite similar way and show a relatively low amount of strategies (two or less in all cases). Moreover, there seems to be gradual increase in the amount of strategies that L2 learners use, with the larger amount being employed by the highest level in our sample (C2).
In order to better understand how significant the differences in the number of strategies employed by the NS and the different proficiency groups of L2 learners were, a oneway between-groups analysis of variance (ANOVA) was conducted for each role play.  The actual differences that were found in the mean scores of the above mentioned role plays were large, as it can be seen by the effect sizes.
Afterwards, post-hoc comparisons using post-hoc tests were carried out in order to explore whether the differences in the mean scores of native speakers and each proficiency group were significant. Table 7 below summarizes the results of the post-hoc comparisons calculated using the Tukey HSD 1 .

NS-NNS Role play 1 (boss)
Role Play 2 (employee) Statistical differences in the amount of strategies appear to be significant between NS and beginner levels (A2 and B1) in all the role plays. More advanced levels (B2, C1) still present differences in certain communicative situations. The highest level in our sample (C2) does not present any differences with NS in the number of strategies.

Role play 3 (student)
In the next Section, the Discussion of these results, as well as the qualitative analysis of the oral data and of the retrospective verbal reports will be presented.

Discussion
T he aim of the current study was to investigate the speech act of thanking by NS and NNS of Greek. More specifically, we focused our analysis on the type and the amount of strategies that these two groups used in formal communicative settings after the fulfilment of a request. Below, results are discussed, first, for NS of Greek (Section 5.1) and then for L2 learners (Section 5.2).

Native speakers of Greek
NS of Greek mainly preferred to use a combination of externally and internally modified direct strategy for the expression of thanking in the three formal communicative situations. The percentage of this use was higher for the case of "employee" (83%), as compared to that of "student" (76,6%) and "boss" (66,6%). Absence of any thanking strategy or cases of a simple direct strategy had not been observed in the NS data. This finding seems to imply that thanking in formal settings is very important for NS of Greek and that they need to reinforce it with further modifications apart from a simple "thanks".
In line with these findings, NS of Greek used a largercompared to L2 learners-amount of strategies (mean score more than four (4)) for the expression of thanking. This result seems to also be related to the above-mentioned idea on the importance of reinforcing the expression of thanking with more elaborated utterances, especially in formal settings as the ones examined here. This might be in order for NS to underline the sincerity of gratitude; in other words, in order to avoid giving the sense of a mechanic and phatic answer. This corroborates the idea of previous studies on how the reinforcement of thanking enhances speakers' positive aspect of public self-image (the so-called face by Brown and Levinson (1987)), while at the same time renders their speech more honest and convincing (Aijmer, 1996;Jung, 1994;Leech, 1983).
Even though the number of the NS in our sample is limited and precludes making generalizations, it allows us to detect some preliminary tendencies about the way NS of Greek perform the speech act of thanking. Example 2 below provides a typical case from the NS data. It comes from the role play 3 (student), where thanking occurs after the teacher agrees to extend the deadline of an assignment. The strategies appearing in this example are four (4):
Oh, perfect! Thanks very much! I will do my best.
Example 2 seems to indicate that the participant, in order to prove his sincere gratitude to the teacher, reinforces his thanking strategy with internal and external components-leading, thus, to the use of a series of strategies. This observation seems to be in line with Aijmer's (1996, p. 51) suggestion that elaborated thanking and length of the utterance itself is a factor contributing to the politeness of speech act. Therefore, elaborate thanking strategies in formal contexts seem to serve speakers' intention, on one hand, to correspond to the challenge of formality and, on the other hand, to verbally reciprocate the interlocutor's beneficial action.
Delving more into each role play, it could be argued that NS performed thanking differently in every situation. More specifically in the case of "boss", qualitative analysis seems to point that many participants opted for other types than the prevalent (combination of strategies). That is, 26,6% employed only internal or external modification and 6,6% employed only an indirect strategy. In concordance with this, the quantitative analysis indicated that in this role play NS employed less strategies than in the other two. Taking into account these two remarks, it could be argued that the sense of dominance over the interlocutor seems to minimize the need for elaboration and proof of the sincerity of gratitude. Thus, in the case of the formal setting with speakers imposing on the interlocutor, an important part of the NS in our sample seems to consider simpler and formulaic structures more adequate.
In the case of "student", some NS opted for an indirect strategy for thanking (13,3%). This could be explained by the sense of solidarity and closeness that characterizes students of the Greek community towards their professors (Koutsantoni, 2004, p.122). Finally, in the case of "employee" both the qualitative and the quantitative analysis indicate that NS used the most elaborate and intensified thanking strategies. Taking into account the social parameters that guide the three formal settings under analysis, it could be said that modification strategies of the speech act are also used as a declaration of respect towards hierarchy.

L2 Learners of Greek
L2 learners' expression of the speech act of thanking in terms of type and amount of strategies varied depending on their proficiency level (for examples of L2 learners' speech acts, see Appendix C). Beginner levels (A2 and B1) seemed to encounter difficulties in the completion of these role plays, and in case they did manage to complete it, most of the times tended to make use of a simple direct strategy. This tendency was also evident in the small amount of strategies that they employed (always one or two). This finding could be interpreted by the fact that participants in these levels either have not yet acquired the necessary pragmalinguistic knowledge or they lack the required grammatical competence. In fact, participants mentioned in their retrospective verbal reports that they considered formal communicative situations highly demanding for their proficiency level, because they did not have the necessary vocabulary or/and grammar to complete the task. The more advanced levels handled every communicative situation differently. Level B2 tended to use more elaborated utterances than the beginner levels. However, they also encountered difficulties which, in their verbal protocols, attributed to the formal nature of the communicative situations. The number of strategies also differed from those of the NS. Those differences appeared to be statistically significant in the case of the "employee" and the "student", but not in the case of the "boss".
Both B2 and C1 level learners employed a combination of internal and external modifications in high percentage in the case of "boss" (66,7% and 83,3% respectively), but not in the other two contexts. In the "employee" and "student" role plays, they mainly opted for internal modification, which is highly formulaic and constitutes a default solution. However, this option seems to imply distance and therefore, it led them to pragmatic failure. The amount of strategies they used in the role play of "boss" is in line with this. The number of strategies that both these levels employed in this particular role play increases impressively. Hence, it seems that for these levels the direction of imposition is an important factor for the performance of thanking. The sense of power and dominance on the interlocutor, characteristics of the role play of "boss", seemed to have made them more verbose and fluent when thanking.
The performance of C2 level learners was very close to that of NS, except for the case of "employee". Regarding the type of strategies, they resorted in high percentage (57.1%) to the internal modification, a very infrequent option for NS (3.3%). As far as the amount of strategies is concerned, the results from the three contexts are similar; however, the number of strategies in the case of the "employee" was found to be smaller than the equivalent in the other two role plays. The combination of qualitative and quantitative analysis indicates that, in this case, that is, in a working context where the hearer imposes on the speaker, C2 learners tried to elaborate their performance, but finally deviated from the NS pattern in our sample. It could be assumed that C2 level participants might not have been aware of the Greek community tendency to reinforce thanking by a combination of both external and internal modification. This performance, in combination with a lower number of strategies employed-as compared to the one of the NS-, seems to create the impression of distance, mechanical response and finally lack of sincerity in the expression of gratitude.
In the case of "boss", the C2 learners' verbosity could be possibly explained by the social power they had in this role play-as it was also found for the previous two levels. In the case of "student", this tendency to expansive talkclose to the native-like pattern-could be attributed to the learners' familiarity with this communicative situation. The participants of this group mentioned in their verbal protocols that they were familiar with formal and academic contextssince all of them were academic students of Greek Philology in their countries-and they had already encountered similar situations in their real academic life, most probably also in Greek.
To sum up, the type and the number of strategies that was used by L2 learners in the formal contexts under analysis seemed to be related to their proficiency level in Greek-a finding which is consistent with the literature (Bella, 2015;Félix-Brasdefer, 2007;Kuriscak, 2010). However, discrepancies from the more native-like pragmatic pattern even at more advanced levels seem to suggest that pragmatic development is not always concomitant with grammatical competence (Bardovi-Harlig, 1999). L2 learners' performance of the speech act of thanking appeared to be related to how familiar they were with each communicative situation and also to the specific characteristics that this situation presented. "L2 learners' performance of the speech act of thanking appeared to be related to how familiar they were with each communicative situation and also to the specific characteristics that this situation presented. "

Conclusions
T he purpose of the current study was to explore the expression of the speech act of thanking in formal contexts by NS and NNS of Greek. Our first research question asked whether there were any differences in the type of strategies that these two groups employed for thanking. Results showed that the NS' expression tended to differ from that of L2 learners. The differences in the type of strategy were more obvious for the beginner levels, while the most advanced levels (C1, C2) tended to be somewhat closer to the native-like pattern.
Our second research question sought to examine whether there were any differences in the amount of strategies that NS and NNS employed for thanking. Similarly to the above question, the NS made use of a larger amount of strategies in order to reinforce thanking in formal contexts. Again, the beginner levels lagged behind the NS, and differences were still identified even at more advanced proficiency levels. Statistically significant differences, however, were mainly identified between NS and A2, B1 and B2 levels. More statistically significant differences between NS and NNS have been observed in role play 2 ("employee"). Differences that were observed among the three formal communicative situations seem to suggest that L2 learners' performance in such contexts seem to be influenced, firstly, by the social power and the direction of imposition; secondly, by participants' familiarity with such context. The case of thanking by native and non-native speakers of Greek The findings of the current study also highlight the importance of a pedagogical intervention for a better acquisition of L2 pragmatic competence (Bardovi-Harlig, 2001;Félix-Brasdefer, 2008;Schmidt, 1993). The majority of L2 learners in our sample reported that they had not received any explicit instruction on pragmatics in their countries and, as it was observed from the analysis of the data, they encountered difficulties with the speech act of thanking in formal contexts. Had the learners been instructed, they might had been more prepared and more familiarized with formal situations. The explicit sociopragmatic instruction and exposure to rich input could familiarize students with pragmalinguistic norms. These interventions are of high importance both in L2 textbooks and in teaching practices, especially in foreign language contexts where access to L2 input is limited (Alcón-Soler, 2008; Bardovi-Harlig & Vellenga, 2012;Celaya et al., 2019). They could help learners build their communicative competence and enable them to have a wide repertoire and consciously select the way they want to express themselves.
Although we tried to systematically investigate our topic, our study has certain limitations, one being that it was a small-scale study with a limited number of participants-a fact that makes the generalizations of the findings hard. Further research with a larger sample is necessary in order to examine more profoundly the possible differences between NS and NNS. In addition, here thanking was explored through oral plays, but it would be fruitful to also explore this speech act with other instruments and methods (for instance, with DCTs or by using ethnographic methods). It would be also interesting for future studies to include longitudinal designs in order to investigate the development of pragmatic competence in time. This could enable us to gain more insights into the way L2 proficiency affects the L2 learners' pragmatic development throughout the different stages of L2 acquisition. To recap, this study attempted to contribute to the field of ILP by bringing to light new, insightful aspects from an under-explored target language and a relatively under-explored speech act.