Stance Taking and Identity in Classroom Interactionsː A Small Scale Study

This present article is mainly concerned with the nature of stance-taking and identity in classroom interaction. The data of the research were taken from the interactions in an English Foreign Language class. Going through the framework of stance triangle, the author explored the features of stances that are frequently taken in foreign language classroom interactions and the identities enacted from the interactions. The finding of the research suggested that epistemic stance was dominantly taken in the interactions, especially by the students. This unequal distribution of stance-taking is likely to occur due to the teacher’s teaching style and students’ lack of evaluation skill. In addition, the finding indicated that both teacher and students, when they take stances, constructed diverse discourse identities, including speaker, answerer, and evaluator. These diverse identities show that the classroom interactions are fairly communicative and dynamic evaluator. These diverse identities show that the classroom interactions are fairly communicative and dynamic. of turn-taking and intersubjectivity that various practices and actions and answers, turn-taking systems, and various linguistic resources of stance-taking activities. a similar the of stance a her pilot study, the to look at the relationship of stance used in the politeness The findings that stance markers - epistemic, evidential, mirative, and

(1) That's certainly a major cause of absenteeism.
(2) I really hope [that it doesn't take that long to put the whole thing together].
(Adapted from Biber, 2006:94) Stance-taking, on the other hand, is not limited to the expression of speakers' attitudes or feelings towards something, but an activity that shows a speaker's position in an interaction. Jaffe (2009) elucidates that stancetaking is an activity of taking up a position with respect to either the form or the content of one's utterances. Through the act of stance-taking, in a spoken interaction, stance can be seen as dialogical and intersubjective activity. It is in line with Du Bois' (2007:163) argument about the nature of stance. He specifically defines a stance as "a public act by social actor, achieved dialogically through overt communicative means, simultaneously evaluating objects, positioning subjects and aligning with other subjects, with respect to any salient dimensions of the sociocultural field" . It is obvious that stance is dialogical and intersubjective in nature. The dialogicality of the stance concerns with the speakers' engagement and/or involvement with prior utterances in a conversation, whereas intersubjective results from the relation between the subjectivity of one speaker to the subjectivity of others. These slightly different concepts indicate that the nature of stance and stance-taking are quite complex.
The complexity of stance/stance-taking has encouraged the researchers to investigate the phenomenon, especially in spoken discourse. It is evidenced by the increasing number of studies throughout the years. Haddington (2004), for instance, explored stance-taking in news interview. By combining conversation analysis and 'theory of stance' frameworks in his study, he investigated interview's turn-taking and intersubjectivity stance-taking activity called positioning/alignment. From his study, he concluded that stance-taking activity (positioning/alignment) is characteristic of a news interview. He further argued that various practices and actions in questions and answers, turn-taking systems, and various linguistic resources contribute to the production of stance-taking activities. With a similar thread, Chojnicka (2015) investigated the notion of stance in a Latvian radio interview. In her pilot study, the researcher tried to look at the relationship of stance used in the interview and politeness strategy. The findings suggest that stance markers -epistemic, evidential, mirative, and hedging devices -tend to be used by both speakers as a negative politeness strategy. These studies along with some other studies (Maschler & Estlein, 2008;Matoesian, 2005;Precht, 2003) indeed provide useful information about stance/stance-taking in interactional settings, but they were not specific to classroom contexts.
With regard to the notion of stance/stance-taking specific to classroom/academic contexts, only a few number of studies has been noted. Yang (2014), for example, explored stance and engagement of academic speech at the university. She took the data from two corpora to see the different use of stance and engagement across soft and hard science disciplines. The results showed that the use of hedges, boosters, self-mentions and pronouns to address speakers and audience are not significantly different. However, her study seems to only report the frequency stance markers used in academic speeches, not necessarily discusses the meaning and position when the stances are taken. Furthermore, Kirkham (2011) conducted a study of stance-taking in the university classroom discussion. He particularly examined how students position themselves through the epistemic stance markers I don't know and I think. Through detailed interactional analysis, he found that the same stance markers may not represent the same identity/position in interaction. In a different line of inquiry, O'Boyle (2010) sought the use of stance-taking triangle framework (Du Bois, 2007) in interpersonal communication of ESL classroom interaction. She found that stance-taking concept was applied during the interaction especially when the students evaluate the object of attention. They do not only do the evaluation, but also align themselves to the position taken by the other interlocutors. These previous studies on stance/stance-taking provide valuable information on the related issue. This present study sought to explore the types of stance in classroom interaction and the participants' identities constructed when they are taking the stances. The relationship between stance and identity is explained in the following section.

Stance and Identity
The notion of stance/stance-taking and identity seem to be closely related in interaction. Johnstone (2007), for example, contends that the stance of the speaker in interaction is seen as a form of identity construction. In a similar thread, Thompson and Hopper (2001) and Kärkkäinen (2006) separately argue that individuals do not only focus on expressing their opinions, thoughts, judgments, and evaluations when they take stances, but also show their identities and relationships with other interlocutors. Furthermore, Du Bois (2002) argues that stancetaking with its linguistic features marks the speaker's alignment in conversation and can be described as 'modus operandi' to construct identity. These indicate that when interlocutors interact, they take stances and concurrently co-construct their identities. Therefore, the speakers' identities can be recognised from their stances and the act of stance-taking.
To date, the concept of 'identity' is somehow still seen as an enigma. There is no absolute and clear summary that define the concept as every expert perceives it differently (cf. Hogg & Abrams, 1988;Bethan & Stokoe, 2006;Berger & Lukmann, 2001;Bernstein, 2000;Blommaert 2005;Bucholtz & Hall, 2005;Clifford, 1988;Erikson, 1974;Jenkins, 1996;Morita, 2004;Wendt, 1992Wendt, , 1994Wenger, 1998;White, 1992). Despite the differences, most scholars argue that identity is regarded as fluid rather than fixed. Bernstein (2000), for instance, states that identity is a dynamic entity that can be negotiated, transformed, and achieved. Blommaert (2005) asserts that identity essentially is who and what an individual is, depending very much on the context, occasion, and purpose of interaction. Bethan and Stokoe (2006), further, ascertains that 'identity is actively, ongoingly, dynamically constituted in discourse' (p.4). Moreover, Bucholtz and Hall (2005), Erikson (1974), Hogg and Abrams (2006), Jenkins (1996), Morita (2004), and Wenger (1998) separately conclude that identity is not fixed and is constructed from the interaction with others. These suggest that identity is a changeable entity that can be constructed individually and socially. A speaker's identity can shift and even vary in an interaction depending on different situations he is in and different roles he assigns. Zimmerman (1998) proposes three types of social identity in relation to interactional processes, including situated, transportable, and discourse identities. Situated identity is predominantly determined by a particular situation. A subject's identity may change when he is engaged in different social activity. In classroom context, for instance, a certain subject acts as a lecturer, and other subjects as students. Once the class is over, the lecturer leaves the classroom and chats with other lecturers at the office, and in that situation the subject adopts the situated identity of colleague. Additionally, transportable identity is the identity which subjects carry with them, such as sex, and skin colour. Lastly, Discourse identity refers to the identity that a subject has within in a given speech situation and integral to each turn of the interaction, such as speaker, listener, questioner, and answerer. These identities may shift and/or change as the conversation unfolds among interlocutors. For this small-scale study, since the interaction is in the classroom context, the researcher focuses on exploring discourse identity by applying stance triangle framework (Du Bois, 2007). The notion of this framework is comprehensively discussed in the next section.

Stance Triangle Framework
As its name suggests 'triangle', stance-taking is a tri-act. It is a visual geometric model that represents interrelation between three elements of stance-taking proposed by Du Bois (2007). This model emphasises the dialogic and intersubjective nature of stance-taking in the interaction (Damiri, 2009;Du Bois, 2007). Du Bois (2007), further, outlines that it is a model for attending to the structured interrelation among the acts and entities which comprise stance and allow the analysts to draw inferences by triangulating from the explicit components of stance to the implicit (Du Bois, 2007:165). That means that this model can be used as a framework for explaining and understanding the socio-cognitive relations among stance elements and entities and the way the relations are built through explicit and implicit information in dialogical interaction, including to identify the discourse identity. The following figure illustrates stance triangle. The stance triangle, in figure 1, consists of three different entities in the stance act which are Subject1, Subject2, and Object. The terms Subject1 and Subject2 refer to the co-participants, while Object is a term indicating the focus of the interaction, such as a person or an event. The act of stance-taking simply begins when Speaker1 takes a stance by introducing a stance object in an utterance and concurrently evaluates it. The next act occurs when Subject2 evaluates the same object that Subject1 has evaluated, positions himself in relation to it, and thereby aligns with Subject1. These indicates that stance-taking act is intersubjective.
Besides, stance triangle also comprises three different elements, including positioning, evaluation, and alignment (Du Bois, 2007). According to Du Bois (2007:143), positioning is the "act of situating a social actor with respect to responsibility for stance and for invoking sociocultural value". This implies that the focus of 'positioning' is the stance takers themselves. This act is usually formed by epistemic and interpersonal stance features. To illustrate, when a speaker says "I am happy", it indicates that the speaker is positioning that he is happy. The second element is evaluation which is also known as assessment (Goodwin, 2006) and appraisal (Martin 2000). This is the most salient and recognised element in stance-taking which has been the focus of attention throughout the years (Lemke 1998;Conrad & Biber 2000;Macken-Horarik & Martin 2003;Linde 1997). In contrast to positioning, evaluation is a stance element which "orients to an object of stance and characterizes it as having some specific quality or value" (Du Bois, 2007:143). This is obvious that 'evaluation' act focuses on the object of interaction, not subjects/ stance takers. For example, when a speaker says "that's interesting", it means that s/he is evaluating the specific target in the interaction. The last element of stance-taking is alignment. Alignment is defined as "the act of calibrating the relationship between two stances, and by implication between two stancetakers" (Du Bois, 2007:144). This shows that this stance-taking aspect is different from the other two in which it is more interactional. The notion of interactional is seen as this aspect includes another party in the interaction. By taking an act of alignment, the speaker essentially engages with the stances of other speakers. When a speaker, for instance, says "I agree", the speaker aligns himself to prior speakers. However, the act of alignment in interaction is not always explicitly expressed as indicated in the example. Some speakers are prone to use gestures like thumb-up, a nod or a headshake, or stance markers like "yes" or "no", or any other forms that index the degree of alignment (Du Bois, 2007).
The overall discussion on stance triangle suggests that it is an appropriate model to understand the dialogical nature of stance-taking in interaction. Although this model has distinctive elements of stance, such as positioning, evaluation, and alignment, they are principally not separated from one another. These three stancetaking aspects can be considered as subsidiary acts of a single stance act. Therefore, in taking a stance, the stance taker positions as subject, evaluates an object, and aligns him or herself with other subjects. This model was applied as the framework to explore the participants' identities in this study.

Method
In advance of analysing teacher and students' stances and their identities, the researcher previously provided classroom-setting interactional data which were drawn from the secondary data (Abrar, 2013). He carried out research in classroom discourse analysis for his master's dissertation. In his study, he observed teacher-students' interactions and collected the classroom interaction data from two classes of different education institutions in one of Indonesian provinces with audio recording and note-taking techniques. One of the classes was a class at public senior high school with the total number of students was 35, while another class was an intermediate class of private English course with only 4 attendance. However, the students-participants' age and the recording activity in both classes were similar whereby the approximate age of participants were 16 and the activity during the recording was mainly speaking. The collected data consisted of four sets of classroom interaction audio recordings and non-verbal communication features, identified from note-taking technique with the transcript, amounting to approximately 14399 words. The transcription procedures for the recording data were on the basis of orthographic principles. Most written punctuation conventions -the use of capital letters and punctuation marks -were disregarded, but long pauses, fillers and any forms of interruption and delay in speech flow were deliberately marked. Due to the length span of the data transcription, for the purposes of this small scale study, the researcher decided to only analyse two transcription sets of classroom interaction data taken from a private English course class.
To analyse teacher-students' identities when taking stances within classroom interactions, the framework of Du Bois' (2007) stance triangle was applied to the data. The use of this model seems appropriate as the data are interactional. Given the fact that stances are frequently communicated across stretches of talk, rather than being located in isolated features (Kirkham, 2011;Local & Walker, 2008), the researcher began the analysis by identifying all epistemic and attitudinal stance markers and expressions from the recording transcription data sets (see Table 1). Identifying stances' markers and expressions is a significant procedure as they explicitly show whether the participants take stances or not in the interaction. After that, the collected data were categorised into Kiesling (2009) stance category to determine the participants' frequency in taking certain stances. Finally, the stances and transcriptions were analysed by using stance triangle framework to explore the identities constructed by participants in classroom interactions.

Research Findings and Discussion
The main purpose of this study is twofold, exploring the teacher and students' stance-taking and their constructed identities in the interaction. Therefore, this section has two main parts. The first part of this section presents the frequency of stance-taking used by the participants in their interaction and second part comprehensively provides the analysis on the identities enacted from the act of stance-taking.

Stance -Taking in Classroom Interaction
The analysis of stances' markers and features has so far shown that both epistemic and attitudinal stance markers were identified in recording transcription as presented in Figure 2. This clearly indicates that the teacher and students take both stances in their classroom interaction.

Teacher Students
As demonstrated in Figure 2, the total of 59 stances' markers and features were found in the interaction and more than 50% of them belong to epistemic stance category. This suggests that the participants in general take epistemic stance more frequently compared to attitudinal stance in their interaction. In other words, the stance takers mostly convey their opinions, knowledge, and belief frequently rather than evaluate the proposition. However, by looking at the specific types of stance, the frequency of stance-taking between teacher and students is different. The students take more epistemic stances; on the other hand, attitudinal stances are mostly taken by the teacher. The findings might be considered as consequences of the classroom atmosphere and individual's cognitive capacity. According to Iinuma (2015) and Rismark and Solvberg (2011), a school/classroom primarily functions as a place of sharing new thoughts, knowledge, and beliefs as well as building common grounds. It is a place whereby a teacher facilitates the learners with knowledge and the students exchange their personal views and thoughts on it. Therefore, it is not surprising when the participants especially students take more epistemic stance in classroom interaction. Additionally, the fact that students take less attitudinal stance inherently relates to their cognitive capacity particularly their level of thinking. Bloom's revised taxonomy (Anderson, Krathwohl, Airasian, Cruikshank, Mayer, Pintrich, Raths, & Wittrock, 2001) lists evaluation as the second highest level of thinking. It is arguably challenging because evaluation requires relevant content knowledge, understanding others, and analysing their thoughts. Figure 2 also shows that three epistemic expression categories were identified in the transcription, such as verbal, adverbial, and others, among other things, modality and gestures expressions. Of the three categories, epistemic verbal expressions, such as I (don't) know and I think were mostly used in addressing epistemic stances with the percentage nearly 75%. This finding corroborates Biber, Johansson, Leech, Conrad, and Finegan (1999), Caliendo and Compagnone (2014) claims that lexical verbs (I know, I think, I guess) are the most epistemic expression in spoken interaction. On the other hand, in relation to attitudinal stance, shown in Figure 2, attitudinal adjectives were noted as the most frequently found items in the interaction to express participants' evaluation and attitude towards the proposition. This study result seems reasonable since attitudinal language is not avoided in spoken discourse and therefore evaluative expression with basic evaluative adjectives frequently appear (Mauranen, 2003). In addition, this finding confirms Swales and Burke (2003) study. They analysed a spoken corpus and revealed that adjectival evaluative is prevalent to use in interaction.

Stance -Taking and Identities
As scholars argue (cf. Du Bois, 2002;Johnstone, 2007;Kärkkäinen, 2006;Thompson & Hopper, 2001) that stance/stance-taking and identity are related functions, it is possible to identify the interlocutors' identities constructed from their act of stance-taking. Going through Du Bois (2007) framework of triangle stance, some teacher and students discourse identities were identified. The frequency data of discourse identity identified from their act of stance taking is shown in Figure 3. Since they generally show different discourse identities, the discussion on the report is separated. In addition, the presentation of identities is completed with selected relevant extracts taken from different part of interaction. Teacher and students discourse identity through stance-taking

Teacher Students Total
As illustrated in Figure 3, both teacher and students constructed a multitude of discourse identities when they took stances, including speaker, answerer, and evaluator. This shows that the interaction is fairly communicative and dynamic as the teacher is not the only participant who takes control of the class, but both teacher and students alternately take stances in classroom interaction. However, the type of dominant discourse identity constructed by teacher and the students is obviously different. The teacher predominantly positioned herself as an evaluator, while the students constructed dominant identity as the answerer.

Teacher Identities
The identity that a teacher constructs through the act of stance-taking are quite diverse. The teacher has a multitude of identities throughout the interaction, and in certain situation she shifts her identity when taking a stance. This is in line with the nature of identity itself that it is fluid and not fixed (Bethan & Stokoe, 2006;Bernstein, 2000;Blommaert 2005;Bucholtz & Hall, 2005;Erikson, 1974;Hogg & Abrams, 2006;Jenkins, 1996;Morita, 2004;Wenger, 1998). An example of the teacher's identity is presented in Fragment 1. In this extract, the teacher takes an epistemic stance with a lexical verb expression in a response to the student's statement (5). The expression 'I think' is an epistemic phrase which indicates degrees of certainty or commitment (Biber, Johansson, Leech, Conrad, & Finegan, 1999;Kärkkäinen, 2003). By uttering 'I think yes although it's boring', the teacher shows her identity as a resistant speaker as she used it to convey her perspective that opposes the student's view on the preposition. Her identity when she takes this epistemic stance changes from her initial identity as a listener. In turn (3), she identifies herself as a listener to her student's stance and even shows her alignment with 'ya' expression. In the following turn (4), the student changes his view on the proposition and it triggers the teacher to oppose the view. The fact that the teacher shifts her identity from a listener to a resistant speaker when taking an epistemic stance supports the fluid and non-fixed nature of identity (Bethan & Stokoe, 2006;Bernstein, 2000;Blommaert, 2005) Fragment 2 1 Teacher : Ok. Did they speak English? 2 Student 2 : Yeah, they speak English very well. 3 Teacher : Ha? 4 Student 2 : Yeah. Because my gang (1.0) none of them can speak English very well. So, they ask me. "Okay, you speak to them", they said. I don't know how to speak. Then, "okay…okay I'll speak", I said. 5 Teacher : That's a good thing for you. 6 Student 2 : Ya, then I speak blab la bla 7 Student 1 : Are they handsome? 8 Student 2 : What? 9 Teacher : Why didn't you ask me to join? 10 Student 2 : I forgot because it's holiday. 11 Teacher : Oh 12 Student 3 : Why don't you ask me? I am their twins. 13 Student 2 : No you are uglier. 14 Teacher : Excellent! You stay in Jambi but it's still good.
This extract is part of large data when the teacher interacts with student 2 talking about her holiday. The teacher mainly plays a role as a questioner who initiates the student to talk by asking questions. Questioning is characteristic in teacher-students' classroom discourse. It is considered as an important form of instructional interaction that stimulates students' thinking, learning, and interaction (Wilen, Ishler, Hutchison, & Kindsvatter, 2000;Wragg & Brown, 2001). In this extract, the teacher takes attitudinal stances (turn 5 and 14) that signal evaluation or judgment (Kiesling, 2009) as the responses to the student's story. By saying 'That's a good thing for you' and 'Excellent! You stay in Jambi but it's still good', the teacher positions herself as an evaluator toward the story and the student's activities during the holiday. Her evaluation is expressed in the form of positive judgement or praise by using positive attitudinal adjectives expression 'good' and 'excellent' (Martin, 2000). Another interesting point from the data is the other students' involvement in the classroom interaction. The teacher is not the only interlocutor who initiates student 2 to speak in conversation, other students do (turn 7 and 12). The involvement between teacher and students in the data indicates the classroom is not a teachercentred class (Duckworth, 2009;Garrett, 2008)  This extract presents another attitudinal stance taken by the teacher in classroom interaction. When she takes the stance, the teacher constructs her identity as an evaluator. The phrase 'That's Indonesian' is an evaluative expression addressed to respond the students' previous statements about the broken rule of law in the country. Unlike the teacher's attitudinal stance in Fragment 2, the teacher's attitudinal stance in this extract also aligns with students' responses. The notion of alignment act in the stance-taking may not be easily noticed as there is not any explicit expression indicating such act, such as agree, yes, or no. However, it can be seen from the initial turns in interaction. As indicated in turn (2) and (3), student 4 and 3 show alignment in their answers responding to the teacher's query. They both argue that rules are made to be broken. The teacher's stance which comes after does not oppose their responses, but provides aligned judgment by saying 'That's Indonesian'. This finding confirms Du Bois (2007) triangle framework that when a speaker evaluates, he positions and aligns himself with other subjects. Moreover, this is in line with O'Boyle (2010) findings in her study.

Students Identities
Like the teacher, the students also reveal some identities in the interaction through the act of stance-taking. One common discourse identity that the students constructed is as answerer. This means that the students take the stances to respond the initiation given by the teacher. An example of this discourse identity are presented in Fragment 4. Fragment 4 This extract is part of large data when the class is discussing the pictures about rules and freedom in the textbook. The teacher asks the students about certain type of bank cards and the students respond them with epistemic stance expressions and make alignment. Student 3 (turn 2) immediately takes an epistemic stance as a response to the teacher's query. The expression 'I don't know', proceeded with deictic 'I', is an epistemic phrase which indicates the speaker's lack of knowledge (Biber, Johansson, Leech, Conrad, & Finegan, 1999;Kirkham, 2011). The following utterance of the turn 'I have never heard it' reinforces the indication of student's limited knowledge on the proposition. The teacher afterwards reinitiates the students with another question (turn 3) and all students take stance in responding the question through gesture 'shrugging' (4). Their gesture is included as epistemic stance as it signals uncertainty with the answer. This finding supports the study results of Roseano, González, Borràs-Comes, and Prieto, (2016) that epistemic stance can be expressed through gestures and corroborates Du Bois (2007) claim that certain gestures can index the act of stance taking. The epistemic expressions 'I just know' and 'I don't know' (turn 5 and 7) are the confirming answers of teacher's questions and alignment expressions with other interlocutors. The meanings of the expressions are completely opposite. The hedge 'just' (turn 5) indicates that the speaker is fairly acknowledgeable, while the negative expression 'don't' (turn 7) means that he seems know nothing. Another student's identity is shown in Fragment 5. In this extract, student 1 constructs his identity as a speaker, not answerer. When the student takes the stance, he is not initiated by a question, but he self-selects to speak in the interaction. The teacher (turn 2) does initiate the student to talk, but not practically point one of the students to respond her initiation. The stance that the student take is attitudinal stance. It shows his feeling toward what he did during his holiday with an attitudinal adjective 'boring' and a booster 'really' as the markers.
The analysis on stance-taking and identities obviously shows that the teacher and students constructed a multitude of identities through the act of stance-taking in the classroom interaction, such as speaker, answerer, and evaluator. They also engaged with one another in learning as the teacher was not the only speaker who initiated the conversation and took control of the conversation (see Fragment 2). These indications suggest that the classroom interaction is more dynamic and communicative.

Conclusion
This small-scale study aims to explore the use of stances and their identities constructed by teacher and students in Indonesian classroom interaction. The investigation on the stances show that the participants in general take more epistemic stance than attitudinal stance in their conversation. Interestingly, the distribution of stance-taking between teacher and students in each stance were found distinctively different and unequal. The teacher, for instance, is prone to take more attitudinal stance (evaluation) than students in the interaction. This unequal distribution of stance-taking among the participants is likely to occur due to the teacher's teaching style and students' lack of evaluation skills. To address this issue, the teacher may apply some teaching techniques that enhance students' evaluation skills, such as self and peer assessment (Bound, 2013;Falchikov & Bound, 1989;Falchikov, 1994), and collaborative learning (Johnson, Johnson, & Smith, 1998;McKeachie, 1999).
Additionally, the results of the study have shown that the identity is a dynamic entity as the participants constructed a multitude of discourse identities when they took stances, including speaker, answerer, and evaluator. These diverse identities also imply that the classroom interaction is fairly communicative and dynamic. The teacher is not the only participant who dominantly takes control of the class as both teacher and students alternately take stances in their classroom interaction. It is found that although the interactions are communicative, the teacher still frequently asks questions to the students. The teacher, in substance, can encourage the students to ask more questions and be inquisitive, so the classroom becomes more active and student-centred (Eekelen, Boshuizen, & Vermunt, 2005).
Even though this study has discussed the use of stances and the identities enacted from stance-taking, this is not to say that the study has no limitations. First, since the data were gained from one class only, they may not represent foreign language classroom interaction in Indonesian institution and any generalisation on stances and identities cannot be made. Further research could include the interaction data from various classes. Additionally, this study examined the stances and identities constructed by teacher and students in general without specifying gender and age differences, so further studies may explore stance-taking acts and their identities from gender and age perspectives.