The Construction of Interpersonal Meaning Realized in a Conversation between EFL Students

____________________ Abstract ___________________________________________________________________ Interpersonal meaning enables speakers to create a conversation by taking roles and establishing relationships. This study attempts to explain how interpersonal meaning is constructed in a conversation between EFL students by focusing on (1) how they take up roles, (2) how they establish relationships, and (3) how the roles and relationships are negotiated. This qualitative study involved a small group of three university students taking part in a conversation within an informal setting. This study drew on analyses on moods, speech functions, and conversational exchanges based on the theory suggested by Eggins and Slade (2004). The results indicate that mood choices allowed the students to take two main roles: initiator and supporter. The initiating role was realized in full declaratives and full interrogatives. The supporting role was achieved through minor clauses, elliptical clauses, and giving opportunities and assistance to the others to deal with incomplete clauses. The speech function choices show that the students established harmonious relationships because they frequently sustained their talks and provided support instead of confrontation. However, this does not seem to represent a good-quality conversation as they tended to avoid different ideas to explore relationships. The way the students in turn showed dominance indicates that they negotiated roles and relationships dynamically. This study can provide a new insight to improve EFL students’ ability to participate in a


INTRODUCTION
The status of English as a global language implies that many learners currently find it essential to develop the ability to use English for oral communication. As a foreign language in Indonesia, English is not spoken in daily activities. Consequently, learners only have limited opportunities to interact in English. Instead, the interaction often occurred within classroom settings. Regarding this issue, various attempts have been made to develop learners' speaking skills. The communicative approach to language teaching, for instance, has provided plenty of opportunities for learners to interact with the teacher and the other learners (Alghamdi, 2014;Astuti & Lammers, 2017;Dendup & Onthanee, 2020;Namaziandost et al., 2019). However, how EFL learners communicate using English when engaging in a conversation is still under-investigated. Therefore, there is a need to examine this issue to provide a comprehensive understanding of how a conversation between EFL learners unfolds, which will eventually improve the quality of teaching and learning speaking skills.
For some reason, conversation plays an essential role in English language learning. First, it is regarded as everyday language. It facilitates learners to communicate in the target language to accomplish various purposes such as social, business, and education. Thus, the ability to do conversation is a priority for many learners. More importantly, learners' mastery of speaking skills is often measured on the basis of how well they take part in a conversation (Sadeghi & Richards, 2015). Second, conversation creates ideal opportunities for learning. It happens because while learners participate in a conversation, they use various strategies to negotiate meanings. These strategies help learners to transmit their ideas and compensate for linguistic problems (Goh, 2016;Nakahama et al., 2012;Walsh & Li, 2013). For instance, learners who do not understand what they hear or are unable to convey meanings may seek assistance. They may also modify their message into a more comprehensible form.
As mentioned earlier, EFL learners often engage in classroom conversation. This type of conversation is undoubtedly different from ordinary conversation in which there seem to be unequal roles between the teacher and the learners. In a classroom conversation, the teacher plays a crucial role in facilitating negotiation of meanings and learner involvement through interactional features (Cancino, 2015;Canh & Renandya, 2017). These features can encourage learners to express their ideas and give more contributions to the conversation. Conversely, learners will have equal roles in a conversation that happens in an informal context. As this type of conversation aims to maintain interpersonal relationships, learners need to work collaboratively to keep the conversation going. Thus, it will be intriguing to investigate this issue.
This study will extend the issue noted earlier by examining how a conversation is created from a systemic functional theory. This theory suggests that a conversation is not merely a turn-taking process producing sounds and words but "a process of making meanings" (Eggins & Slade, 2004, p. 6). During a conversation, speakers take turns negotiating ideational, interpersonal, and textual meanings. Of these three meanings, interpersonal meaning is the focus of this study. This meaning is concerned with how speakers use language to take up their roles and establish relationships with others through systems of mood and speech function.
Mood relates to a grammatical resource for expressing interpersonal meaning. It deals with "types of clause structures" (Eggins & Slade, 2004, p. 74) such as declarative, interrogative, or imperative. Mood choices provide an overview of learners' roles and contributions in the conversation. For instance, a learner dominantly using declaratives tends to take the role of providing information. Conversely, when a learner uses many interrogatives, he aims to elicit information from others. In this case, learners need to be aware of the mood choices so that they are able to take up different roles appropriate to the given context of situation.
Speech function is concerned with a discourse resource for expressing interpersonal meaning. It refers to "a functional label for what a speaker achieves in a particular move in dialogue" (Eggins, 2000, p. 136) such as statement, demand, offer, or question. Speech function choices indicate how learners act on each other to establish relationships. The choices may function to sustain or terminate the conversation. For instance, challenging responses such as rejection or refusal prompt further responses that keep the conversation going rather than supporting replies such as acceptance or compliance. Thornbury and Slade (2006) argue that conversation is interpersonally-motivated so that the speakers need to choose speech functions that keep the conversation going. Learners need to master different speech functions and structure them to achieve specific communicative purposes.
Interpersonal meaning has been the focus of several studies in various registers such as classroom interaction, conversation, and conversational texts. Some studies on classroom interaction show that the realization of interpersonal meaning is aimed to make learning come alive. The teacher tries to be equal with the students (İlhan & Erbaş, 2016;Mahardhika et al., 2019;Moore & Schleppegrell, 2014;Yuliati, 2013). It is because classroom interaction aims to encourage active participation from learners, and the teacher plays a vital role in making this happen. Other studies argue that grammatical choices may reveal some information about interpersonal relationships between speakers in conversation (Banda, 2005;Lam & Webster, 2009;Nguyen, 2017;Sari & Sari, 2020). The relationship is determined based on the response types, either supporting or confronting. The dominant use of supporting replies may indicate a low-quality conversation because the speakers do not explore their relationship. Despite having a similar communicative purpose with real conversation, dialogues written in English textbooks show an inappropriate realization of interpersonal meaning (Achsan & Bharati, 2015;Arifuddin & Sofwan, 2015;Khalim & Warsono, 2017;Meiristiani, 2011). These dialogues are mainly used as learning materials for EFL learners. Consequently, they should portray authentic features of real conversation, including the grammatical patterns and speech acts performed; thus, learners can better acquire conversational skills.
To sum up, this study aims to investigate how interpersonal meaning is constructed in a conversation between EFL learners. This study explains this issue by focusing on (1) how learners take up roles, (2) how learners establish relationships, and (3) how roles and relationships are negotiated.

METHOD
This qualitative study included text analysis and observation (Denzin & Lincoln, 2018) to obtain an in-depth understanding of the construction of interpersonal meaning in a conversation. More specifically, this study adopted a case study approach where the construction of interpersonal meaning was investigated through a group of EFL students participating in a conversation.
Three students, coded as A1, B1, and C1, were selected purposefully, i.e. based on the recommendation given by a lecturer as active students in the classroom, to take part in this study. They were first-year university students taking a conversation course in the English Language Education Program. The students took part in a conversation conducted for 40 minutes in an informal context. Conversation has been considered a relevant context for the students to use English communicatively. Indeed, it allowed the students to freely express their ideas, feelings, or messages (Walsh & Li, 2013), especially between EFL students (Dobao, 2012). In other words, it offered a considerable opportunity for students to exchange interpersonal meaning. The conversation between the three students was also considered more interactive; thus, it provided a good data source for the analysis.
As the primary data, the conversation was collected by means of audio recording and field notes. While the recording was used to capture the verbal language, the field notes displayed non-verbal behaviours found in the conversation. The data collected through these instruments was further transcribed based on the convention adapted from Eggins and Slade (2004).
This study drew mainly on mood analysis and speech function analysis to explain the construction of interpersonal meaning. The mood analysis explained how the students took roles in the conversation, while the speech function analysis explained how the students established relationships with others. In addition, this study examined conversational exchanges to explain how the students' roles and relationships were negotiated as the conversation unfolded. These analyses were all based on the theory suggested by Eggins and Slade (2004).

Students' Roles in the Conversation
The mood analysis dealt with types of clause structures used in the conversation. The results of this analysis are summarized in table 1. Imperative Minor clause 26 84 39

Number of clauses
The number of clauses produced indicates speakers' contributions in a conversation. Based on the summary of the mood analysis, C1 was the most dominant speaker in the conversation. He produced nearly half of the total clauses in the conversation (376 out of 849 clauses). In contrast, A1 made the smallest contributions in the conversation. He produced 202 clauses of the total, approximately half of C1's total clauses. Subsequently, B1's contributions were more than A1's but less than C1's. She produced 271 clauses of the total. The proportion of clauses produced by the students only gives preliminary evidence of their overall roles and contributions in the conversation.

Number of incomplete clauses
Despite his dominant roles in the contribution, C1 produced the highest proportion of incomplete clauses in the conversation compared to the other students. B1 produced a slightly smaller number of incomplete clauses compared to A1. On the contrary, A1 produced the smallest number of incomplete clauses. The high proportion of incomplete clauses suggests that C1 and B1 tried to keep speaking even when they stumbled or hesitated. It can be seen from how they often said a series of incomplete structures before reaching the complete ones. In addition, the incomplete clauses occurred because both C1 and B1 frequently left their utterances unfinished, or they tried to create alternative structures. On the contrary, the small number of incomplete clauses produced by A1 suggests that his speech was careful and planned.

Declarative
All students relatively produced a high percentage of declarative clauses in the conversation. C1 used the highest percentage of declaratives and most of which were full declaratives. Besides, the number of declaratives produced by B1 was slightly higher than that of A1. The summary of the mood analysis shows that C1 tended to initiate negotiation most frequently through declaratives, while A1 and B1 did not seem to take the initiating role as often as C1 did. The analysis also indicates that the students produced declaratives to respond to prior initiations or provide requested information. The responses involved both full and elliptical declaratives. In this case, A1 provided supporting responses through minor clauses more frequently than the others. Most of the declaratives produced in the conversation were also used to provide explanations or additional information to the previous messages. At other times, declaratives were used to ask for confirmation (e.g., it's not your your own project?).

Polar interrogative
The summary of the mood analysis shows that all students produced an equal proportion of polar interrogatives. In this case, B1 produced a slightly higher number of polar interrogatives than A1 and C1. Full polar interrogatives were frequently used to initiate exchanges by asking for confirmation or agreement (e.g., do you have um something to do in Ramadhan, like um something special?). On the contrary, elliptical interrogatives were frequently used to respond to prior initiations as the students attempted to check or confirm what they had heard (e.g., do you?). Elliptical polar interrogatives used in the conversation often consisted of finite and subject. However, in some cases, the students omitted the finite and used rising intonation (e.g., you sure about that?). Concerning this case, asking for confirmation through elliptical clauses indicated support among the students because they depended on each other to receive confirmation.

Wh-interrogative
The use of wh-interrogatives in the conversation varied between the students. C1 produced the most significant number of wh-interrogatives, which mostly were full ones. A1 used a higher number of this clause type than B1, most of which were in full form. In contrast, B1 produced the smallest number of this clause type. Full whinterrogatives in the conversation were frequently employed to initiate exchanges by demanding information involving either fact or opinion (e.g., what do you do during the daylight?). In addition to its function to initiate exchanges, full wh-interrogatives were often used in the conversation to ask for help due to difficulties to find specific words in English (e.g., What should we say pahala in English ya?). Wh-interrogatives, particularly elliptical ones, were used to elicit further information so that the students took the role of an interrogator (e.g., then what?).

Imperative
There were quite a few imperatives (3 clauses) used in the conversation. C1 produced two imperatives, and A1 only used one imperative. This finding means that there was no significant power shared among the students. In this case, most of the imperatives were used to encode advice rather than to elicit compliance (e.g., If … if you wanna cry, just cry.).

Exclamative
There was only one exclamative used in the conversation. It means that the students quite rarely gave judgement or evaluation of events discussed in the conversation.

Minor clause
The summary of the mood analysis shows that B1 dominantly used minor clauses compared to A1 and C1. She produced 84 minor clauses while A1 and C1 used 26 and 39 clauses. The result indicates that B1 frequently provided encouragement support to the other speakers' utterances. Despite the difference, it is reasonable to argue that the students actively encouraged each other to keep the conversation going. The frequent minor clauses included lexicalized items such as "oh my God!", "okay", "really?", "right?" and positive and negative polarities such as "yeah", "yes", "no". The students used these minor clauses to give responding contributions. Moreover, the students used non-lexicalized items such as "uh huh" and "mmm" to provide feedback and backchanneling.
As suggested earlier by Banda (2005), grammatical patterns allow speakers to make contributions and take up roles in a conversation. Concerning the findings of the mood analysis, the students took two primary roles: initiator and supporter. First, the initiating role is typically realized in full declaratives or full interrogatives. The findings show that each student produced a significant number of declarative clauses. It implies that the students frequently took the initiating role by putting forward information to be negotiated. In contrast, the number of full interrogatives (polar and wh-interrogatives) was much smaller than that of declaratives. It indicates that the students occasionally initiate exchanges by demanding information from others. More specifically, the number of polar interrogatives, in general, was higher than that of wh-interrogatives. It means that some initiations tended to seek agreement or confirmation rather than request explanation from the others.
The significant number of declaratives among the students implies that they competed for initiating exchanges. This result corresponds to the fact that conversation provides speakers with equal rights to speak (Thornbury & Slade, 2006). Other types of spoken interaction (e.g., classroom interaction, interview) are likely to produce different results. For instance, interrogatives are more often used by the teacher in a classroom interaction to elicit students' responses (Mahardhika et al., 2019;Yuliati, 2013).
In addition to the initiating role, the students took supporting roles in the conversation through different mood choices. First, the support was achieved by encouraging each other through minor clauses. The high proportion of minor clauses indicates that the students actively encouraged each other to participate in the conversation. Another way to give support was through elliptical clauses. Each elliptical clause served different functions in providing support. Elliptical declaratives were generally used to provide answers, while elliptical polar interrogatives were employed to ask for confirmation or agreement from the others. The students also used some elliptical whinterrogatives to request further information as they shared the same knowledge. In addition, the students gave support by providing additional information through full declaratives. The students often could not express their ideas in one utterance; thus, they elaborated them in other utterances. Banda (2005) and Eggins and Slade (2004) argue that the supporting role can be seen as evidence of inability to take the initiating role. It makes sense because the student who dominantly took the supporting role initiate the conversation less frequently.
The supporting role was also achieved by providing opportunities and assistance to deal with incomplete clauses. The findings show that all of the students produce a different proportion of incomplete clauses. These incomplete clauses were generally caused by their difficulties expressing their ideas or feelings. When one of the students encountered a problem, another student sometimes gave an opportunity to repair the utterances or provide assistance by guessing the proposed ideas. This result corresponds to Nguyen (2017), suggesting that the existence of incomplete clauses tolerated and repaired by the other speakers indicates supportive contributions towards the success of a conversation.

Students' Relationships in the Conversation
The speech function analysis was conducted to explain how the students acted on each other during the conversation to establish relationships. This analysis deals with types of moves. The summary of the speech function analysis is presented in table 2. She produced 177 turns out of 472 turns and 237 moves out of 720 moves.

Opening moves
The summary of the speech function analysis shows that most students preferred to initiate exchanges by giving statements involving both facts and opinions (e.g., Well, I don't know why, but … last few days all I thinking about is my parents). More specifically, the students tended to give statements in the form of facts, while in some cases, they expressed opinions. The high proportion of statements was because conversation allows them to express ideas and feelings towards different topics. The summary also shows that the students rarely initiated new exchanges by asking questions about different aspects of the topic discussed. The use of questions can be seen as a means to keep the conversation going because it prompts the other students to respond. Most of the questions used in the conversation were in closed form to demand confirmation or agreement (e.g., Um … so … do you have um something to do in Ramadhan, like um something special?). On the contrary, the students used a limited number of open questions to demand facts or opinions (e.g., How do you feel being … the youngest in the -in your family?). The other opening moves used in the conversation were attending and offer. In sum, C1 took the initiating role most frequently in the conversation. This role was achieved mainly through providing information to be negotiated and asking open questions to create chances for the other students to make contributions to the conversation. In addition, A1 and B1 shared relatively equal proportions in initiating exchanges, mainly through giving information and asking for confirmation or agreement.

Continuing moves
While opening moves function to initiate the negotiation of a proposition, continuing moves aim to maintain the negotiation on the same proposition. Continuing moves are achieved through three main types: monitoring moves, prolonging moves, and appending moves. The summary of the speech function analysis shows that the students frequently maintained their negotiation by prolonging it. The students tended to elaborate their previous moves most frequently in prolonging their talk. They clarified, restated, or exemplified their previous moves to make them easier to understand through elaborating moves (e.g., I think it's not effective if I … only come home for two days. I mean actually I could …). There was also a significant number of extending moves, in which the students gave additional or contrasting information to their previous moves. At other times, the students used enhancing moves to qualify or modify their previous moves by giving temporal, spatial, causal, or conditional details (e.g., I I did not do that because … when the meeting is done it's already night). In addition to using prolonging moves, the students continued their talk as soon as they regained the turns they had lost through appending moves. Once the students regained the turns, they frequently elaborated the previous propositions by restating, clarifying, or exemplifying them. After regaining moves, the students sometimes extended their moves by giving supporting information, contrasting information, or enhancing their previous moves by giving causal details. At times, the students maintained their talk through monitoring moves to check whether the other students followed a talk or to invite another speaker to take a turn.

Supporting moves
The summary of the speech function analysis shows that the students tended to give supporting reactions rather than confronting ones. The supporting reactions were achieved in two ways: responding supporting and rejoinder supporting moves.
Responding supporting moves allowed the students to give support which moved the negotiation toward completion. The analysis results indicate that the students employed responding moves more frequently than rejoinder moves. They frequently provided supportive encouragement through registering moves (e.g., oh, yeah, okay, uh huh, etc.). In addition, the students sometimes repeated what the previous speaker said to indicate that they followed the interaction. The support among the students in the conversation was also achieved through supporting replies. The replies involved acknowledging, agreeing, answering, affirming, and accepting. The findings indicate that the students frequently showed their acceptance of the others' proposals through agreement. The agreement generally was given through minor clauses such as "yes", "yeah", or elliptical declarative such as "yes we are", "yes you are". At other times, the students showed their alignment by acknowledging the information given by the others (e.g., Yeah I see. I see). The other supporting replies used by the students in the conversation were accepting, answering, and affirming. These replies are commonly used as a response to opening moves (e.g., Yes I have). The students frequently develop the others' propositions by elaborating them. The elaboration was achieved by clarifying, restating, or exemplifying the previous propositions to make them clearer to understand. Compared to elaborating, the students extended the others' propositions less frequently in the conversation.
In addition to responding supporting moves, the students also supported the others through rejoinder supporting moves to prolong the conversation. These moves consist of tracking moves and supporting responses. In general, the students used clarifying and probing moves more frequently than the other tracking moves. The clarifying moves were used to seek additional information to understand the previous moves (e.g., why?), and the probing moves were used to volunteer further information to be confirmed (e.g., So, it's it's it's not your your own project?). At other times, the students gave support by checking what had been said (e.g., what?) and seeking confirmation of what they had heard (e.g., You're getting fatter?). In addition to the tracking moves, the students provided support for one another through supporting moves, which consisted of clarifying (resolving moves) and correcting forms of language (repairing moves).

Confronting moves
Besides providing support, the students also confronted each other through confronting moves. However, the students only produced a pretty small number of confronting moves. The small number of confronting moves also corresponded to the considerable proportion of supporting moves, which have been discussed in the previous chapter. It means that the students tended to keep the conversation going by giving support rather than confronting each other. The students sometimes used challenging moves to attack previous moves. They either sent the interaction back to the first speaker by questioning the relevance or veracity of the other students' moves (rebounding) or offered alternatives or counter-position of a situation raised by a previous speaker (countering).
The findings suggest that the relationships among the students were established through different moves. First, the students collaboratively initiated exchanges. It can be seen from a slight difference in the number of opening moves among the students. As discussed earlier, the students frequently used declaratives to initiate exchanges. It corresponded to the finding of opening moves, showing that most of the students initiated exchanges by giving information.
Second, the students also established relationships by sustaining their talks. It can be seen from the number of continuing moves, which was the most significant among the other moves. The students frequently sustained their talks because the conversation provided opportunities for them to express their ideas. In this case, the students often revealed their personal stories; thus, it required them to create sequences of moves so that the others could easily understand the messages. At other times, they checked whether the others were paying attention through monitoring moves when they produced long utterances. The fact that the students often sustained their talks confirms some previous studies (Dobao, 2012;Walsh & Li, 2013) suggesting that sustained conversation offers speakers to articulate their point of view freely towards a particular topic.
Besides sustaining their talks, the students established relationships through supporting moves. Most of the support was given as responding moves, which bring exchanges towards completion. The findings show that the students frequently produced long utterances. Therefore, these sustained talks allowed the other students to provide supporting responses such as encouraging the speaker to take another turn, showing acceptance, giving agreement and answers, and acknowledging the given information. At other times, they developed the others' propositions by restating them and providing supporting or contrasting details. In addition to supporting moves, the students established relationships through confronting moves. However, there was only an insignificant number of these moves. The fact that the students tended to provide support rather than confrontation is relevant to the study conducted by Nguyen (2017), suggesting that support among EFL students is motivated by their attempt to maintain a harmonious atmosphere during a conversation. However, this does not seem to represent a good-quality conversation because they do not explore relationships with others.

Negotiation of Roles and Relationships
The findings of the conversational exchange analysis show dynamic negotiation of roles and relationships among the students. The findings show that A1 and B1 dominated the first few exchanges. A1 frequently initiated the exchanges, while B1 cooperatively responded to A1's initiations. A1 also actively engaged B1 to respond by demanding further information regarding the topics discussed. B1 dominated the following few exchanges primarily by providing facts about herself, which means that she tried to construct personal stories about her feelings. At the same time, A1 and C1 cooperatively responded by either giving supportive encouragement or demanding further details to prolong the negotiation. Similarly, as B1's stories seemed to end, C1 started to dominate the following exchanges by presenting his personal stories. At the same time, A1 and B1 took the supporting role by providing encouragement and demanding additional information.
The dynamic shifts of each speaker's dominance indicate how the students cooperatively kept the conversation going. One student's dominance could be motivated by another speaker's inability to open up a new exchange. Because one student encountered linguistic difficulties in constructing ideas or was unwilling to provide further information, another student took the initiating role and dominated the following exchange. The fact that the students, in turn, dominated the conversation provides new insight into the study conducted by Nakahama et al. (2012), suggesting that EFL learners are likely to assist each other to avoid communication breakdowns. Besides, personal experiences seem to underlie the dominance. When a student presented his/her personal experiences, he/she tended to produce long utterances by developing ideas. It implies that he/she took the role of giving information and put the others to respond.
The findings of the conversational exchange analysis also capture further evidence relating to the way support and confrontation were given. For instance, C1 sustained his talks most frequently than the others, while B1 produced the most considerable number of supporting responses such as registering, agreeing, and acknowledging moves. It means that the support given by B1 was often addressed to C1's propositions. The support and confrontation among the students extend the findings of a previous study which suggests that positive or negative responses contribute to the realization of one's orientation towards others (Lam & Webster, 2009).

CONCLUSIONS
This study indicates that the conversation was possible because the students constructed interpersonal meaning. This meaning was constructed through grammatical patterns, discourse patterns, and conversational exchanges. First, grammatical patterns allowed the students to take two significant roles: initiating and supporting. Each role was realized in different clause types. Second, Discourse patterns enabled the students to establish relationships. The relationships were revealed through how the students initiated exchanges, how they sustained talks, and how they provided support and confrontation for each other. Third, conversational exchanges portrayed how roles and relationships were negotiated dynamically. The students, in turn, showed their dominance in the conversation, and their personal experiences underlay their dominance. The high proportion of a particular move made by a student implies that another student dominantly produced related moves. This study has provided new insight into how EFL students engage in a conversation. In the end, some suggestions can be made to improve EFL students' ability to participate in a conversation.