Body Talk: Moving Beyond Speech When Analysing Literature Discussions

The intent of this article is to promote a methodological approach to the analysis of literature discussions by adopting a conversational analytic approach that views embodied actions (including talk) as multimodal. It is argued that this analytic approach moves beyond previous studies that focus on talk only, when analysing literature discussions, and further, that such an approach contributes to an increase in our understanding of the complex meaning-making processes that are set into play when students interact on literary texts in school.


Introduction
During their time at school children and young people alike interact with a great variety of texts which they may process in different ways in the classrooms. Regarding their encounters with literary texts there is in our schools a strong belief in literature discussions as a form of a teaching and learning process. With a point of departure from a sociocultural perspective on learning and development (Säljö, 2011), the literature discussion is seen as a significant pedagogic element through which different readers will have an opportunity to join in dialogues with each other about their readings. The fundamental idea is to let students sit together, with or without their teachers' direct participation, in small groups and discuss the same literary text they have read. Such literature discussions are supposed to create opportunities for various readings and perspectives to be met and confronted with each other, with the aim of supporting children and young people as developing readers (see Tengberg, 2011, for a more extensive discussion).
One important notion in this development is the the reader response theory perspective that emphasises the dialogic element between the reader and the text in the reading process (McCormick, 1994). Under the influence of Louise Rosenblatt's work (1938Rosenblatt's work ( /1995Rosenblatt's work ( , 1978Rosenblatt's work ( /1994 on the act of reading and her theoriesreader response theory and transactional theory -we now know the importance of creating opportunities in our classrooms where our students can respond to texts according to their own unique lived-through experience. In fact, in her work, Rosenblatt even argues that we perhaps should consider "the text as an even more general medium of communication among readers" (Rosenblatt, 1978(Rosenblatt, /1994 where different readers can interact and discuss their interpretations in order to develop as readers and as citizens in a democratic society. Research on literature discussions in school settings has generated increased interest recently, probably mainly due to the emphasis in various research contexts on the importance of collective forms of learning, such as classroom discourse and discussions, for students' learning processes (Chinn et al., 2001;Langer, 2011;Nystrand, 1997Nystrand, , 2006Soter et al., 2008).
Many of these studies on literature discussions examine, with different discussion approaches and with different aims, how the use of group discussions can help students promote high-level comprehension and responses to texts (Almasi et al., 2001;Annmarkrud & Bråten, 2012;Chinn et al., 2001;Evans, 2002;Maloch, 2002;Marshall, 1989;Murphy et al. 2009;Raphael et al., 1998;Reninger, 2007). There is also a growing body of research that focuses on different identity construction aspects that take place in literature discussions and examines how these correspond with the readers' reception of the literary text and how students construct themselves as readers in a school context. Many of these studies also deal with power issues circulating in these literature discussions. Gender has been a thoroughly studied topic (Dutro, 2001;Evans, 1996aEvans, , 1996bEvans et al., 1997) and there are also studies that examine how racial-identity formations are constructed in literature discussions (Rogers & Mosley, 2006). Asplund (2010Asplund ( , 2012 examines how adolescent boys construct class and gender in literature discussions, and in another study, Asplund & Pérez Prieto (2013) show how adolescent boys use literature discussions in the construction of identity and place.
Research on literature discussions is often done through video recordings, which is the case in the majority of the studies mentioned above. Video-based methods offer many opportunities for the researcher to evidence things in the interaction in a completely different way than audio recordings can, which has led to an increase over the past decade in the use of video recordings in the research on literature discussions. Common to these studies is that they focus on spoken language in their analyses of the literature discussions. Although video recordings also register the visual aspects of who or what the participants are directing their attention to, and how participants use their bodies and physical objects in the environment as semiotic resources, little interest (with the exception of Asplund, 2010and Asplund & Pérez Prieto, 2013 has been given to adopting a multimodal approach to the study of literature discussions. In this article, such an approach will be used, with the aim of examining how multiple semiotic resources are attended to, collaboratively recognised and integrated into the ongoing organisation of a literature discussion.

Studies of Multimodal Interaction
The study of multimodal interaction has been carried out from a number of different perspectives. What these studies have in common is that they deal with embodied interaction in the material world, including material objects in the surrounding environment, in the process of meaning-making and action formation. The term "multimodal(ity)" has been used by various groups of researchers in the study of human interaction. From the perspective of the social semiotic research paradigm (Jewitt & Kress, 2003;Kress 2010;Kress & van Leeuwen, 2001) there are several studies that focus on communication and meaning-making practices in school contexts. Some examine how different school subjects such as Science, Geography and English are shaped by the organisation of different semiotic resources Kress et al., 2005;Taylor, 2012). Others focus on if and how different digital media can promote the construction and reconstruction of knowledge, learning and literacy (Jewitt, 2003(Jewitt, , 2005(Jewitt, , 2008, how drama, visuals and music are used by students to interpret texts and to construct meanings (Lee, 2013), or the use and representation of multimodal texts in kindergarten (Granly & Maagerø, 2012) and in learners filmmaking practices (Gilje, 2009).
In another line of research, with an approach that will be adopted in this article, examinations have been made of how participants simultaneously use multimodal resources to organise their interactions. This kind of analysis (conversation analysis; henceforth CA) focuses on how multimodal resources contribute to the ongoing interaction, and studies how these multimodal resources are coordinated in interaction between participants (Goodwin, 2000(Goodwin, , 2003(Goodwin, , 2007Goodwin & Goodwin, 2012). This research has been carried out in a range of different contexts, including educational settings (see e.g., Asplund & Kilbrink, 2016;Käntää, 2014;Melander & Sahlström 2009;Tabensky, 2008). While the social semiotic approach tends to focus on the semiotic products independently of whether and how the participants orient to them, CA takes up a radical participant perspective in that it studies what participants themselves orient to and treat as relevant in the interaction. Conversation analysts do not claim to unveil the expression (or the intention) of the individual mind, rather they aim to describe "what is out there in the material world, as it is perceived and oriented to by participants" (Sahlström, 2009, p. 105).
Within the research field of reading, some researchers have devoted their attention to how other semiotic resources than talk, such as gestures and artefacts, are used by teachers in the teaching of English (Bourne & Jewitt, 2003) and in reading strategy instruction (Shanahan & Roof, 2013). Analysing the practices that takes place in the classroom from a social semiotic perspective, these studies tend to focus on bodily conduct more as phenomena that have an inherent meaning (shaped through cultural, historical and social usage) rather than public social ones. That is, they do not, as this study does, attend to how temporally unfolding deployment of bodily conduct serves as a public, interactive resource on which participants may build locally relevant action. Neither do they approach human action and interaction with the microanalytic focus on the formation of action sequences that has developed within the research field of CA, the approach adopted in this article.

Theoretical and Methodological Framework
Taking as a point of orientation that action is constituted through participants' simultaneous use of different semiotic resources in face-to-face human interaction, a CA approach to multimodal interaction will be used below when analysing a literature discussion carried out by four boys in upper secondary school. CA is a research discipline that has its origin in social constructionism, and one of its points of departure is that both meaning and understanding are "established, sustained and changed in and through the coordinated interaction of people" (Sahlström, 2009, p. 103) and the interaction between people and artefacts in specific social and cultural contexts (Drew & Heritage, 2006;Goodwin, 2000;Schegloff, 2007).
Another central principle in CA is that it takes a radical participant perspective. What is to be studied is how conversation participants carry out courses of collaborative actions together, how they show their interpretation and understanding of each other's actions and what this generates in terms of new actions (Goodwin, 2000;Hutchby & Wooffitt, 1998;Schegloff, 1996Schegloff, , 2006. In the work of identifying the aspects of a conversation that participants themselves make relevant, and to show this in a convincing way to other researchers, CA research has developed a method that Sacks et al. (1974) calls "proof procedure". This involves the use of the participants' demonstrated understanding of each other's actions as a resource in the analysis and as a means of enabling readers to judge the claims of the analyst. In this work, CA uses detailed representations of the interaction, which will be presented in this article as detailed transcriptions of spoken data, as well as drawings of visual phenomena in the literature discussion.
In CA, verbal and non-verbal language are seen as tools through which participants in a conversation construct and reconstruct their social reality; thus not only talk but also other semiotic resources such as gazes, gestures, body movements and physical objects are seen as constitutive of and integral parts of the ongoing interaction. A reconceptualisation of embodied interaction (including talk) in the material world as multimodal, and a recognition of the importance of drawing attention to how physical objects in the environment are drawn into the production of inter-subjective understanding, are priorities in using this analytic framework (Streeck, Goodwin & LeBaron, 2011).
One of the leading CA researchers who has developed such a perspective is Charles Goodwin (2000Goodwin ( , 2003Goodwin ( , 2006Goodwin ( , 2013; other works have emerged in line with his research that examine how talk and embodied actions mutually contextualise each other on a moment-bymoment basis, and how these semiotic resources provide a resource for participants to organise, interpret and understand each other's contributions to the interaction in the ongoing activity (Goodwin & Cekaite, 2013;Heath, Luff & Knoblauch, 2004;LeBaron, Glenn & Thompson, 2009;Streeck, 1993Streeck, , 1994Streeck, , 2003.
In this article the relationship between talk and bodily conduct will be viewed as processes that mutually contextualise one another, providing public resources for the participants to organise relevant actions that, in concert with each other, contribute to the ongoing activity (Asplund & Kilbrink, 2016;Goodwin, 2000Goodwin, , 2006Streeck, Goodwin & LeBaron, 2011). This approach will be used with the aim of showing that actions are constructed through the simultaneous use of multiple semiotic resources, and that it is important to pay attention not only to the vocal communication between participants in a literature discussion but also to the embodied practices of both the speaker and the listener. In this respect, the article takes its point of departure in spoken language, embodied actions and artefacts as used in concert with each other to construct meaning and understanding in the interaction, which is also very much tied to the specific local context in which it occurs (Goodwin, 2000(Goodwin, , 2007 and to the fact that it is essential to take this into consideration when analysing literature discussions. Thus, approaching multimodal interaction in a CA framework, this article aims to broaden and develop the methodological discussion of using and analysing video data in the study of literature discussions, and to contribute to a growing body of knowledge about the complex meaningmaking processes that are set into play when students interact with literary texts in school.

The Research Design
A sequence from one literature discussion will be presented and analysed in this article. Those carrying out the discussion are four boys (ages 16-17) who attend a traditionally maledominated vocational programme (the Vehicle Engineering programme) providing training for jobs such as car mechanics, truck drivers or excavators. The boys are in the same class, there are no girls in the class and the upper secondary school where the boys study is located in a rural setting in central Sweden.
During the years 2005-2008 the Author of this article lived and worked as a researcher and teacher in the classroom that was being studied, collecting extensive ethnographic data (Delamont & Atkinson, 2008) such as observation notes, all the pupils' written work, surveys, interviews and about 30 hours of audio and video recordings of some of the teaching lessons. In this article I will concentrate on one of the video-recorded peer-led literature discussions in which Niklas, Krister, Peter and Jonas (all pseudonyms) participated as part of their Swedish language instruction, and I will focus on a specific sequence in this discussion with the aim of examining how the boys use talk, embodiments and artefacts from the surrounding context when talking about their readings of a translated version of John Marsden's novel Tomorrow When the War Began (1993). The novel is about a group of Australian teens who take action towards waging a guerrilla war against a foreign force that has occupied their country. The novel, which depicts the teens struggling for survival, was part of a larger thematic project on heroes and role models that the class was involved in as part of their Swedish language course when the literature discussion took place. The boys had been instructed by their teacher to discuss the plot of the novel briefly and to talk about their reading experiences and discuss the various characters' actions and motivations and relate them to the thematic work on courage, morality and heroism.
Since the article's primary aim is to focus on methodological issues surrounding how to approach video-recorded data when analysing literature discussions, I will not discuss the boys' comprehension and responses to the novel in great detail. For the same reason, I will not discuss any educational settings that may have had an impact on the discussion.
At the beginning of the boys' conversation, the teacher moved between their group and another in the classroom. In the sequence analysed in this article, the teacher was not present in the group. The film camera recording the boys' conversation was equipped with a wide-angle lens and positioned at the front of the classroom on a TV stand.
The analysis of the boys' literature discussion is presented below with a transcript that reflects certain aspects of the interaction. Square brackets ( [ ) in vertical line positions indicate simultaneous talk. A parenthetical dot indicates a short break, a "micropause" (<1 s), while numbers in parentheses indicate a pause in seconds. Double "more than" (>>) means that the turn continues on another line, while text in double parentheses is relevant information from the Author. Empty parentheses are used when something is being said that was inaudible, while a whole word (or part) is underlined when it was uttered in a louder voice than usual.
The article focuses on how the participants in the conversation used their bodies and how they brought physical objects into the interaction; in this context I make use of illustrations to show how the boys used their bodies and artefacts as semiotic resources in the interaction (an approach which is commonly used within the CA research tradition (see e.g., Goodwin, 2000Goodwin, , 2006Melander & Sahlström, 2009). Five illustrations in which the boys' bodily conduct is especially evident are represented in the article. For reasons of lack of space, I have chosen not to use more illustrations but to describe the boys' bodily conduct in double parentheses. To show more precisely where in the conversation the illustrated act was performed, I present the transcript, with the words being said at the time, using an external frame. Arrows in the drawings are used to indicate the directions of the gestures, lines are used for gaze. For ethical reasons, I chose not to make use of still images of the recorded material.

Findings The Boys' Literature Discussion
Parking a tanker truck. The boys discuss the novel for about 20 minutes. In the sequence presented below, they are sorting out one of many events in the novel when a controversy emerges regarding how one of the characters in the novel, Ellie, parks a tanker truck under a bridge. Ellie and her friends will then set the tanker truck on fire so it will explode, and the bridge will be destroyed. It is of no importance to the plot how the tanker truck was parked, but the boys' dialogue shows that this is important to them and their readings of the novel. Krister emerges as the most ardent proponent of the conception that the tanker truck was parked perpendicular to the bridge (i.e. intersecting the driving direction), while primarily Jonas claims that the truck was parked parallel with the bridge (in the driving direction). In the excerpt, the boys then relate to these two attitudes.
Using different kinds of signs. The sequence begins when Krister, on lines 1 and 5, says that he doesn't really understand how the tanker truck was parked. In Krister's view an extremely wide bridge is required to allow a tanker with its trailer to fit under the bridge, and Krister says he doesn't really understand how this could be done. Krister's opinion prompts Jonas (lines 8, 10-11) to claim that parking the tanker truck is quite possible to envisage, if one imagines that it was parked parallel with the bridge: "Well, it depends on how you park it. If you park it parallel with the bridge like this." While making his statement Jonas takes his pen and puts it on the table, and when he says "like this" (see line 10) he moves his right hand straight forward. As soon as Jonas puts the pen on the table, it becomes implicated in the interaction and when he says the word "bridge" and moves his right hand forward along with the pen, the pen turns into a bridge and it is in relation to this pen/bridge that Jonas highlights how the tanker truck was parked.
What is thus happening in lines 8-11 is that Jonas, in reaction to Krister's questions and comments about how the tanker truck was parked, builds up a scene from the novel, fully visible on the desk in front of the other boys. His action is constructed here through the simultaneous use of different semiotic resources; his talk is filled with the combination of a pen and his body movements, and as we follow the boys' interaction in the excerpt we can see that the other boys show that they understand what Jonas is telling them.
This can be seen for instance by the fact that the pen, from this point on, is the basis and starting point for the boys' continued discussion about the parking scene in the novel, i.e., the boys continue to relate to the pen as a bridge; it becomes implicated in collaborative, symbolic action (see Streeck, 2011). The discussion that continues deals with how the tanker truck was parked, i.e. if it was parked parallel with the bridge or if it was parked in a direction intersecting the driving direction.
What is interesting in this context is that the other participants' actions are also constructed through the simultaneous use of multiple semiotic resources, and in order to understand, as an analyst (and participant), what they are communicating we must take into account all the different kinds of signs that are being used in the interaction. If we only took into account what the boys say, i.e. their talk, we would find it difficult to understand parts of their conversation, as some of the phrases they use are inconclusive.
When, as an example, Peter says "The pillar could be like this", after Krister has said that the tanker truck was parked next to a pillar in such a way that the passenger door could not be opened (lines 18, 20-21), it is difficult to understand what Peter means if we only focus on what he says. That the pillar can be "like this" can of course mean many different things, since Peter never specifies "like this" with his talk. But if we pay attention to what he is doing with his body while he makes his utterance, we find that he completes it by moving his index fingers towards the desk.
Then, in the same way as the pen was made into a bridge by Jonas's action earlier in the excerpt, Peter's index fingers are transformed into two pillars for a brief moment, until he folds his arms and puts his hands together again. Thus, taking Peter's embodied actions into consideration, his talk suddenly makes sense, and as we can see, the other participants don't treat his talk as faulty, rather they show that they understand him by relating to his action in various ways (see also Streeck, Goodwin & LeBaron, 2011).
A similar action is the one done by Jonas in line 29. In the line above (line 28), Krister repeats his perception of how the tanker truck was parked (supplementing it this time by pointing straight ahead). Jonas objects by saying, "Yes, but it may well be straight under like that too." Again, if we only take into account what Jonas says it would be hard as an analyst to form an idea of what Jonas is trying to say, since he doesn't clarify what he means by "like that." However, taking into account what Jonas does with his body during the talk we can see that he nods towards Peter when he says "like that." Parking the tanker truck "straight under like that" thus means to park it the way Peter has argued earlier in the conversation (when he uses his index fingers). Jonas thereby takes the same stance as Peter.
Co-occurring actions. The significance of paying attention to how participants use semiotic resources when interacting with each other is particularly clear when we take a closer look at what is happening in lines 30-46. There were namely several things co-occurring in the previous illustration that give us information about what was happening in the interactioninformation that would have gone unnoticed if we had focused on the spoken language only.
In lines 44-45, Jonas repeats the stance he presented at the introduction of the excerpt (i.e. that the tanker truck was parked parallel with the bridge, in the driving direction), using his talk, gestures and the pen in the interaction. When he does that, Niklas is not paying attention to what he is communicating. Instead, Niklas is occupied with writing something on a piece of paper in front of him. His gaze is pointed down, towards the paper on the desk, with his pen against it, indicating that he is writing or drawing something on it. This is an act that the other participants can observe, and as we can see, Peter is the one who is following Niklas moving his pen with his gaze while Jonas directs his focus on Krister.
The fact that Niklas is paying attention to something other than Jonas's act can be interpreted as his having lost interest in the discussion of how the tanker truck was parked. This interpretation is reinforced if we also pay attention to what Niklas is doing just before and just after this action. In lines 30 and 34-35, Niklas expresses the idea that the boys should just agree that the tanker truck was parked under the bridge and that it is less important to sort out (and agree on) how it was parked. In lines 34-35, he also turns to Jonas and says "Hey you! They parked the lorry under the bridge, so it's cool, okay?" while looking at him and grabbing his arm.
By analysing how Niklas uses talk and embodied actions it is possible for us to observe that he is addressing Jonas with his talk. Niklas gets no support, however, for his appeal; instead Krister and Jonas continue their exchange of views on how the tanker truck was parked, whereby Niklas exclaims "Oh my, oh my, oh my!" as he shakes his head and looks at Peter, probably to seek support for his disappointment in a continuation of the discussion. Then right after Jonas has repeated his stance (lines 44-45) and Niklas has turned his attention to his paper and pen, Niklas changes the topic in the conversation by saying, "And then they ran like crazy!" By taking a closer look at how Niklas uses talk and embodied actions in concert with the other participants' actions during a longer sequence in the conversation, it is possible to systematically examine the practices he uses in the literature discussion and to understand what he is doing. If we had only paid attention to Jonas's talk in lines 44-45, and not examined other kinds of semiotic resources and meaning-making practices that the boys themselves treat as relevant in the interaction, the analysis, and our understanding of the boys' interaction, would have been incomplete.

Conclusion
Through the analytic framework used to examine the boys' literature discussion, this article moves beyond the previous studies of literature discussions that focus on verbal language as the baseline for their understanding of the meaning-making practices that take place within literature discussions. If such a perspective had been the baseline of this analysis, it would have been much harder, if at all possible, to make any sense of how the boys construct meaning from their readings when they discussed the novel.
The actions that take place when the boys discuss their readings of the novel are, as have been examined in this article, simultaneous and collaborative through the boys' constructed actions, and the meaning-making practices are constructed through the mutual elaboration of several different kinds of semiotic resources, such as talk, gestures, gazes and physical objects. When discussing the parking scene in the novel, the boys are eager to explain to each other how they construct meaning from it, and the analysis shows that their perceptions differ depending on the projected images lived in their literary experiences. Using Rosenblatt's terminology (1978Rosenblatt's terminology ( /1994 here, it is the efferent mode of reading that comes to the fore in the focus on the parking scene in the boys' conversation. In contrast to the aesthetic mode in which the reader's attention is centered on what he is living through during the encounter with the text, the reader's primary concern in responding in the efferent mode is with what he will carry away from the reading. In the efferent mode of reading, the reader has a clearly focused attention on what the words point to, in order to acquire and retrieve particular information that will serve the reader's practical purposes. When discussing the parking scene, it is clear that the boys' embodied actions help them demonstrate their readings and perceptions of the novel, here and now. As Sumara (2003) points out, "reading is the act of continually noticing and interpreting links between and among different 'bodies' that comprise our physical, psychic and ecological experience of the world" (p. 92), and when the boys are exchanging experiences of the text they have read, their verbal language is supported by other semiotic resources. Through this interchange, the boys encourage each other to express their unique responses to the text.
Thus, using a CA framework on multimodality, the article shows that it is important that we as analysts, and teachers, pay attention to what the participants themselves are orienting to and treating as important in order to gain an understanding of how participants construct meaning and understanding when interacting with literary texts in a school context.
The production of the boys' embodied actions are thus not merely private psychological phenomena, nor pre-determined, but, as shown in this article, publicly visible and, therefore, to be seen and taken into account as meaning-making actions. These meaning-making actions, constructed by a speaker, are interpreted and acted upon by the speaker's co-participants (Goodwin, 1996(Goodwin, , 2000LeBaron & Streeck, 2000). If we as analysts and teachers wish to increase our understanding of what is going on when our students perform literature discussions in our classrooms, we need to pay attention to how participants' talk, bodily conduct and use of artefacts contextualise each other, and how they provide a resource for the other participants to organise and construct relevant actions when participating in a literature discussion.
Both the social semiotic approach to multimodality, with its emphasis on the sign-maker and his or her intentions, and the view of sign-making as a social process (Jewitt, 2009), along with the CA approach adopted in this article, are valid and can contribute to advancing the study of literature discussions. But when there is a specific interest and analytical focus on interaction, and when researchers want to examine how multiple semiotic resources are handled, collaboratively recognised and integrated into the organisation of a literature discussion, it is my belief that the CA perspective, with its microanalytic focus on interaction, is especially appropriate.
The findings in this article suggest that a CA approach to multimodal interaction that takes a radical participant perspective (Schegloff, 1996) and assumes that different semiotic resources interact with each other to build locally relevant action (Goodwin, 2000(Goodwin, , 2013Schegloff, 2006) can contribute to increasing our understanding of the complex meaning-making processes that are set into play when students interact on literary texts in school. The approach thus makes it possible to show how theories of multimodality are set into practice when students discuss and negotiate their interpretations of what they have read, and this study may, therefore, develop teachers' analytical "metalanguage" (New London Group, 1996, p. 77)i.e. their professional language for talking about reading instruction and meaning-making interactions in classroom practice.