Readers’ Interactions in an Online Reading Group

This study is an investigation of readers’ posts to an online forum devoted to the discussion of short stories and poems. The ELT Online Reading Group (ORG) ̶ an electronic version of a face-to-face reading circle ̶ aimed at promoting the development of reading skills and the reading and discussion of literature among teachers, trainee teachers and other professionals in the field of English language teaching (ELT). This paper focuses on the multi-layered relationships between readers in the group by looking at some of the distinctive features of participants’ posts. Results suggest that there is a dialogical orientation in the communicative strategies participants employed in the forum and that those are similar to the ones observed in face-to-face reading groups. The internal ‘stratification of discourse’ and the ‘diversity of social speech types’ that Bakhtin (1981: 262) argues are present in the Novel have also been observed in the forum posts. The findings of this study may be relevant to an assessment of the viability and usefulness of establishing online reading groups as tools to promote the development of members’ reading skills, the creation of communities of readers, and a re-evaluation of the role of literature in language


Introduction
This paper reports on a study conducted on an online reading group for teachers, trainee teachers and other professionals in the field of English language teaching (ELT). The Online Reading Group (ORG) was hosted on the British Council website TeachingEnglish and active between 2007 and 2014. The data for this study was collected between 2010 and 2013. The study consisted of an investigation of readers' responses to literary texts and their comments on such texts posted to a dedicated online discussion forum. This paper discusses the extent to which participants' posts to the forum related to the group interaction and which compositional features in such written comments indicate these relationships.
The development of interactive technologies at the end of the 20th century and their considerable expansion in the 21st century has made it possible for people living in diverse places and time zones to get together and exchange information and ideas via social media and online interactive channels. These new possibilities have been recognized by people interested in reading and discussing the books they read with other readers without the temporal and spatial limitations inherent to face-to-face reading groups. An online search conducted on "reading groups" on Google in May 2016 generated 'about 216,000,000 results' while a new search in early February 2017 produced 'about 408,000,000 results'. The vast majority of these reading groups are hosted by mainstream media, including newspapers and publishers, as well as various other organizations. They are often linked to the home pages of such companies and organizations even though their commercial associations and affiliations may not be immediately obvious thus suggesting complex relationships between publishers, social networks, and readers . In spite of their growing numbers and Internet presence, such reading groups are mainly designed for readers who have access to physical copies of books in their lists of recommended reading, either by acquiring them or borrowing them from local libraries. Moreover, as a rule, they do not explicitly target non-native speakers of English who largely do not have access to printed versions of texts, as it was the case with the members of the online reading group investigated in this study.
This investigation was based on the notion that the meanings attributed to the texts read, produced, and discussed in the research setting were partially the result of the complex relationships between those texts and the myriad of other reading and life experiences readers brought with them. In this paper, I argue that the internal dialogue that Bakhtin (1981) sees as the main characteristic of the Novel as a literary genre is also present in the dialogue among literature readers in the ORG forum. At surface level, this dialogue is visible in the interactive strategies participants employ, such as greeting, naming, posing questions, giving answers and openly agreeing or disagreeing with something that had been previously said by another poster. At a deeper level, dialogic interaction is found in the posters' drive towards what others have said, such as borrowing ideas from other posters and engaging in an unfolding exchange about other participants' ideas for a string of comments.

Face-to-Face and Online Reading Groups
The first organized reading societies appeared in England in the eighteenth century with members meeting face-to-face to discuss their reading (Hartley, 2002: 18). Since then, groups of readers, variously called reading groups, reading circles and book clubs, have spread around the world and have eventually reached the cyberspace. Reading groups could be simply defined as groups of people who regularly, or irregularly, meet to discuss the books they read. One of the very first studies on reading groups in the UK was conducted at the beginning of the 21st century by Hartley (2002) whose survey disclosed a wide variety among reading groups in terms of composition, objectives, mode of interaction, reading selection, frequency of meeting, patterns of participation and whole group dynamics. Despite such diversity, Hartley emphasizes that the social interactive aspect is shared by almost all groups surveyed since 'a reading group isn't just about reading; it is about reading in a context, a context which is fostered by the group, and which in turn affects the whole experience of reading' (p.22). Commenting on Hartley's early work, Sedo (2011: 9) recognizes that the publication 'provides useful background for an understanding of the cultural, social and educational roles of book clubs', but argues that it ignored human complexities and the complexities of the relationships between books and readers. Sedo (2003) conducted her own survey on reading groups collecting data mainly among Canadian and American readers. Her research, besides focusing on the gendered aspect of membership, considered the influence of the Internet on book clubs, as well as the differences and similarities between face-to-face and online reading groups. She concluded that for both kinds of groups 'the popularity of book clubs lies in the idea that people want to regain a sense of community and that sharing ideas about books is one way to do that' (Sedo, 2003: 85). Although Swann and Allington (2009: 247) intended to investigate both face-to-face and online reading groups, their research was eventually limited to a face-to-face reading group in Scotland. Their study aimed to 'provide an example of how "ordinary readers" (…) interpret and evaluate literary texts'. They drew on sociolinguistics to provide an analysis of spoken interactions in the reading group under investigation and concluded that 'where participants appear to make reference to their subjective responses to texts, this often has the function of presenting evaluations of those texts in mitigated form.' They observed that in the discussion of texts, participants directly invoke their own and others' knowledge and past experiences and identified 'a significant amount of co-construction in which interpretations' were 'collaboratively developed ' (p.262).
Despite the similarities between face-to-face and online reading groups, the environment in which exchanges occurred in the ORG was considerably different from most research settings mentioned above since the overwhelming majority of group members had little or no offline contact with each other. Granting that there are considerable differences among various face-to-face reading groups, some of the aspects that were regularly observed by Hartley (2002) in her seminal study can still serve as a point of reference for the analysis of the similarities and differences between face-to-face reading groups and the ORG (Table 1). Worldwide community; open, fluctuating membership; members belong to one professional group; diverse cultural backgrounds, no specific information on participants' gender.

Interaction
Usually pre-determined, regular, scheduled meetings; predominantly face-to-face contact and interaction; mostly oral communication; typically immediate response to other members' comments; the discussion is accessible only to group participants and, potentially, those in contact with them.
Flexible, unrestricted, and unscheduled access to the discussion; predominantly, and often exclusively, online contact and interaction; textbased communication; mostly delayed response to other members' comments; the discussion is accessible to any individual visiting the Group as long as the website remains live.

Reading
Texts available as hardcopies; texts usually selected by group members; wide range of criteria for text selection; typically all members read the same text for a given period.
Texts available as electronic copies; texts mostly selected by the Group coordinator; choice of text strictly subject to copyright restrictions according to the UK laws and regulations; members may be reading different texts available on the website in a given period.

Readerresponse
Mostly oral commentary; personal and ethically oriented responses to texts; comparisons to other books read together.
No data regarding the use of quotes or other linguistic and compositional features observed in the texts read in participants' comments and discussions.
Predominantly written commentary; personal and ethically oriented responses to texts; frequent intertextual references; presence of other linguistic and compositional features observed in the text; frequent use of quotes; creative writing output triggered by texts discussed in the Group.
Some more recent studies on online readers' include Lima and Lamy (2013) discussion of the social and professional development interactions taking place in the same online reading group that is the subject of this paper. Curdwood (2013) also focus on reading on social media but his investigation consists of a single case study of a 13-year-old boy and his 'literacy practices across modes, texts, and contexts associated with The Hunger Games trilogy'; while Vlieghe et al. (2015: 25) observed 'literary practices within social media environments' on the website and a Facebook group established by the Flemish government aimed at promoting the reading of literature. Yet, despite recent investigations, there is still a considerable gap in published academic research on online reading groups. One of the main contributions of this study to the field is that the unique nature of ORG provided a research setting where it was possible to examine online interactions among readers who chiefly depend on the Internet as their means of contact and as the main source of reading material.

The Educational Roles of Reading Groups
The potential for learning that reading groups offer was early recognized by educators who started using face-to-face reading groups in diverse educational contexts both in schools and in further education for the development of literacy, reading strategies, critical reading, literary reading, and for a plethora of other educational purposes (Daniels, 2002;Hartley, 2002;McMahon & Raphael, 1997). For example, Cumming-Potvin (2007) conducted a case study with a grade-7 boy in Australia to investigate the role of scaffolding in literature circles to facilitate the development of critical reading skills, whereas Duncan (2009) explored the roles of reading circles to advance adult reading in formal adult education in London. Peplow (2011) used conversation analysis and the concept of communities of practice to analyse the face toface group interactions in a book club in south-east England in 2009. Levy (2011) looked into the connections between literature reading circles, developmental reading and basic writing in classrooms in a community college in New York, while Sanacore (2013: 117) reflected on how literature circles could be used 'as a vehicle for nurturing personal and critical responses to text'. More recently, Hyder (2016) investigated reading groups in public libraries and how they have expanded to include a wide range of groups, such as visually impaired readers, in order to promote social inclusion.
Reading groups have also been used in English language teaching to promote extensive reading in foreign and second language acquisition and recent research in the field tends to support the claim that participating in reading groups can be beneficial to language learners in a number of aspects. For example, working with Vietnamese learners Shelton-Strong (2012: 222) argues that, as long as appropriate scaffolding is provided, literature circles can 'foster learner autonomy, while providing opportunities for focused extensive reading and collaborative, purposeful discussion' and serve as 'platforms to promote collaborative and multidimensional learning'. Beglar et al. (2012) claim that Japanese university EFL students showed great gains in terms of reading rate and reading comprehension by taking part in pleasure reading groups. The concept of reading groups has also been explored by some ELT publishers (Macmillan, 2016;Oxford University Press, 2017) who support their creation as a way of exploring their Graded Readers series in classroom-based discussion groups. Much rarer is literature on reading groups among English language teachers. Working in Japan, Fenton-Smith and Stillwell, (2011) applied the concept of reading groups to explore how English language teachers engaged with ideas in texts; however, instead of literary texts, their reading group members discussed professional literature.

Reading Groups as Communities
Communities can be categorized and defined in different ways according to different criteria, such as the profession, interests and language their members share. In this sense the ORG could be seen as a community since its members interact in a shared virtual space and have the common characteristics of being all connected to English language teaching and interested in literature. In his seminal work, Wenger (1998: 4-5) proposed the concept of communities of practice which is based on the assumption that learning is the result of active social participation in the 'practices of the social communities' that in turn leads to the construction of 'identities in relation to these communities' (emphasis in the original). Both practice and identity are key components of Wenger's social theory of learning, along with meaning and community, all of which are 'deeply interconnected and mutually defining' (p.5). Wenger's theory is that learning is an integral part of our lives and it occurs in the various communities to which we belong (p.8) 'through our engagement in actions and interactions' that are embedded 'in culture and history' (p.13). Learning is not seen as a separate activity but something that occurs at all times. It can be intensified at time when our familiarity with situations is shaken as problems and challenges are posed to us and force us to engage in new practices to be able to respond to them. Moreover, 'even failing to learn what is expected in a given situation usually involves learning something else instead' (Wenger, 1998: 8). Lave and Wenger (1991) call situated learning the learning happening in a particular context and social situation. Considering learning a response given to a particular problem in a certain context and situation is close to Bakhtin's (1990) notions of answerability or responsiveness. For Bakhtin, life is an utterance, a response by someone in a certain time-space situation (chronotope) given to someone else in another time-space situation. There is no moment or place in life that is not a situation. Life is dialogical because we are naturally oriented towards a response. We cannot fail but answer to the other and to the world, even if we refuse to answer. That is why some form of learning always takes place when individuals interact with each other (Holquist, 2002: 153).
Communities have also been defined in linguistic terms. The fact that members of the ORG are all speakers of English, albeit with different levels of language proficiency, and possibly making some use of the ELT professional jargon, may lead to the question whether the ORG can be considered a speech community. Wardhaugh (2011: 118-134) traces the historical definitions of speech community and the difficulties of coming to a consensus on what it actually constitutes. Swales (1990: 24) argues that the concept of speech cannot be adopted as 'an exclusive modifier for communities that are often heavily engaged in writing' and, instead, advocates a distinction between socio-linguistic grouping and socio-rhetorical grouping based on their communicative needs and goals, group formation and forms of recruitment into the community. Swales' taxonomy and principles have also been contested, alternative terminologies have been proposed, and further differentiation between groups and communities suggested (Zhu, 2005: 37). Linguists' difficulties in producing encompassing definitions for speech and discourse genres and to account for individuals' apparently inconsistent and multiple discourses inside a given community illustrate their 'failure to acknowledge the actual possibility of specific genres coexisting' (Bakhtin, 1981: 5) within a given body of discourse, which Bakhtin had already identified in his critique of the Poetics of the nineteenth century.
Developments on communication technology have brought the concept of communities to the cyberspace and there is no shortage of publications in the field of online learning communities. From the beginning of the century, researchers have examined a wide range of issues related to technology-based learning from the instructional, technical and collaborative point of view (Andreson & Elloumi, 2004;Lambropoulos & Zaphiris, 2007;Luppicini, 2007;Palloff & Pratt, 2007;Palloff & Pratt, 2010;Pozzi & Persico, 2010;Prestridge, 2010;Stephenson, 2001;Warschauer, 2002). There are also edited publications specifically connected to the setting of this study, such as England (2012) who focuses on online education from a TESOL perspective and Lancashire (2009) who brings together a collection of case studies in language and literature online teaching.
A seminal work on the use of technology in language education was published by Lamy and Hampel (2007) where they compare and contrast two studies on online communities: one 'inspired by Hallidayan text linguistics' and, thus, positioned in the 'more cognitive learning framework' (p. 108); and another that investigates teachers' collaborative use of technology from a constructivist point of view and is, therefore, positioned 'within the more sociocultural of the learning frameworks' (p. 110) but they concluded that research in the field of online interaction continued to miss 'a theory of collaboration' (p. 112).
In a report published by the European Commission, Aceto et al. (2010) analysed twelve different online learning communities, including one devoted to English language learning (pp. 59-64) and one created to facilitate the exchange of books among individuals mostly living in the US and Europe (pp. 69-72). According to the report, in spite of all the differences among them, in all analysed communities 'members perceive that significant learning happens unintentionally (with very few exceptions) by means of interaction, knowledge and experience sharing, and material creation and/or sharing' (p.6). Aceto et al (2010: 105) conclude that the usual definition of an online learning community as 'a group of individuals who have a common learning goal' fails to adequately describe the complex interactions (contextual, motivational, personal, behavioural, values) which 'combine to create a satisfactory and effective learning experience'. Instead, they propose a definition of learning community as a group of 'individuals who share common interests, aims, passions, objectives or circumstances' and who 'expand their level of understanding, awareness, knowledge, experience and horizons' by engaging with other members in the group through electronic media.
In this study, I adopted Aceto et al.'s (2010) broad definition of online community based on the sharing of a common interest in literature instead of a goal oriented definition since survey data showed that participants' objectives when joining the ORG tend to vary. I also propose employing Bakhtin's concepts of dialogue and heteroglossia to fill in the theoretical gap Lamy and Hampel (2007) identified and, hence, explain how online community members collaborate and construct knowledge through language.

Research Methodology Participants
Although the ORG membership consisted of over 1,400 members based in over 75 countries at the end point of the data collection (October 2013), research participants considered in this paper are only the members who posted in one or more of the 14 online discussion threads dedicated to the discussion on a single short story or poem. Analysis of the forum posts shows that posting was done mostly by a core of 23 long-term members and three new members that took up posting with enthusiasm, besides occasional contributions by various other participants.
Due to data protection restrictions imposed by the online host organization, it was not possible to establish the mother tongue of each group member or accurate information on members' professional status. However, a survey sample of 126 ORG members, conducted between February and May 2012, shows that the majority of respondents were non-native speakers of English (89.69%) who were mostly ELT practitioners (73.8%) or teacher trainers (17.5%). The survey participants' ELT history is much more heterogeneous; however, approximately one in four respondents reported having been working in the field for more than 20 years.

Data Collection and Analysis
From the 38 ORG active discussion threads (total of 1,611 posts), 14 threads on the discussion a single short story or poem were selected for analysis of the readers' forum interactions, with a total of 382 posts (Table 2). Threads were selected based on the following criteria: a) they contained a minimum of twenty replies each; b) they roughly corresponded to different periods in the ORG existence, thus adding a longitudinal dimension to the study; and c) they discussed texts written in different styles, by writers of different nationalities and from different literary periods, albeit still limited by online accessibility to texts and the UK legal copyrights constraints. August 2012 20

TOTAL 382
The first stage in the analysis was to look at the posts in relation to the literary texts on which they commented in order to identify discursive features present in, or showing some connection with, the reading source. However, in this paper, I focus on the second analytical stage when posts in the same thread were analysed in order to identify discursive features and interactive patterns between posters. As in face-to-face reading groups, the act of reading presupposes engaging in 'talking about' the text read since posters know that their comments are open to others' replies where their arguments can be reinforced and/or contested and where meaning may have to be negotiated. In this paper, I discuss three main patterns of relationships among participants that were observed in the ORG forum: use of direct quotes, interactions and adoptions.
Although the analytical categories used are unique to this study, Hall's (2015: 157-230) comprehensive review of research done in the field of reading literature, with both native speakers of English and second language learners in different educational contexts, reveals that some of the categories adopted by other researchers bear some similarities with the ones employed here, albeit differing in the terminology adopted. The main differences between the vast majority of the studies Hall (2015) discusses and this investigation are: a) those are mostly controlled experiments in extensive reading, b) they tend to consider solely the relationships between texts and individual readers, without considering reader-reader interactions, and c) they tend to focus on linguistic features of readers' responses to texts.

Theoretical Framework
My choice of analytical approach was based on conceptual understandings strongly influenced by Bakhtin's (1981) views of the dynamic multiple relations between the self and others encapsulated in the concepts of dialogue and heteroglossia. However, there is not yet a single Bakhtinian approach with a set of specific analytical tools, such as the ones developed by conversation analysts or Foucauldian researchers. Therefore, researchers analysing discourse from a Bakhtinian perspective tend to develop their own codes based on the 'complex' and 'valuable conceptual resources ' Bakhtin offers (Coulter, 1999: 12). According to Maybin (2001: 70), Bakhtin's master concepts have practical implications for those analysing discourse: a) 'both spoken utterances and written texts need to be understood in terms of how they are responding to, and anticipating, other utterances and texts'; b) 'a speaker may explicitly or implicitly report or appropriate other voices from written texts, authoritative figures, or a comment earlier in the conversation'; c) 'the words and phrases which speakers use bring with them their own social history and association, and introduce a wealth of nuances and connotations into the current speech context'; and d) 'individual utterances and texts will reflect the heteroglossia of language itself, and the conflicts that permeate it, between centrifugal and centripetal forces and between authoritative and inwardly persuasive discourses.'

Findings and Discussion
Direct quotes were usually used by writers to enhance the reliability and authority of their statements. Analysis of the forum posts reveals that although posters quoting from each other were far less frequent occurrences than posters quoting directly from the literary texts (247 in 382 posts), there were still a number of occasions in which writers' borrowed each other's words. There are 46 instances in the fourteen threads analysed when posters inserted a direct quote from other posters into their comments. The examples below show how participants quoted each other in order to agree with a particular statement, disagree with it, pose questions, or develop a topic further: I fully agree with your suggestion that 'the characters bear their fate stoically', but in my reading I…. (P11 quoting P10) Why [P11] when you said "finally where is the ghost in this story? Not where he supposed to be"… (P11 quoting P33) The second part conveys destruction, extinction, a damaged environment, where "the purity of the land is lost to civilization" (P12 quoting P21) The fact that quotes from other posters are less numerous than quotes from literary texts and less frequently used to build up an argument may be seen as an indication of how much less authoritative posters considered the others' contribution to the forum. In comparison, it may be also seen as an indication of how much participants regarded the literary text itself as the final authority in matters of interpretation.
Interactions are communicative instances where posters clearly addressed each other by naming and/or signalling a response to another participant's post by framing it with pertinent language. The most common interactive strategies observed were naming, questions and answers and agreeing and disagreeing. Interactions are overt ways of actively responding to what is posted by others, either as an immediate reply or as a delayed response. Moreover, the very acts of posting and replying can be seen as actively sought forms of interaction and a response to someone's utterance. The most frequent interactive strategy employed was the opening address when participants greeted each other, either speaking to the group as a whole or directing their responses to particular individuals. General ways of addressing found are 'Hi All' and 'Hi everyone' with variations using the words hello instead, but more elaborate versions, such as 'Dear Colleagues' and 'Hi Readers', also appear. The usual greeting used when directly addressing another ORG member was 'Dear' followed by the name of the person. In general, the tone of addressing was polite and frequently posters addressed the previous commentator by name. Names were also frequently used in conjunction with a general greeting, as in 'Dear Chris and All'. The general address, as in All, may be understood as a form of addressing all the other posters involved in the discussion in a particular thread, but it may also be seen as a way posters had to acknowledge that their comments would be read by a larger online audience and as a way of reaching out for other potential ORG members who had not engaged in posting.
Another interactive strategy posters adopted was the use of questions that invited a response, as the in the following examples: , it is just to answer your question addressed to the group (…). I am also leaving the same question to our dear colleagues: Should the narrator tell everything to his/her reader? (P04) Hello All, I'm [P30], from the south of B____. I'm a university teacher and I've been following your discussions with interest. As for our friend's question about the narrator's role, my opinion is no, I don't think he should tell us everything! (P30) Although indirect and rhetorical questions cannot be considered a 'direct' forms of interaction, they may be seen as attempts to prompt some form of response. The examples below illustrate how posters made use of rhetorical questions and indirect forms to invite replies: We found the last line interesting: "Heaven is for them that bring the rain." Shall we compare this "rain" to God's blessings?... Adoptions are instances where a poster borrowed and developed ideas initially posted by others in the same discussion thread. Adoptions can be seen as posters' attempts to build on an idea or theme raised by others. These may be instances where the comments of one poster were prompted by another's previous comments on a theme or topic triggered by the reading of the literary text. There are 167 identified instances of Adoptions in the 382 posts devoted to the discussion of a single short story or poem. The example below shows a clear reference to the Gunesekera's Goat: As for the narrator not having a name, I agree with you, [P12], that it is not relevant, and I also agree with you, [P11], that the narrator is probably the author himself (both were born in Colombo; both are now living in London). (P23) Adoptions are also instances where one poster presented ideas or raised issues that had been previously proposed by someone else in the same thread without making a direct or indirect reference to the previous post. The poster then simply re-stated a similar point ignoring that it has just been mentioned. In the fifth post into the discussion thread on Wilde's The Nightingale and the Rose, a participant made a reference to the religious overtones of the tale. In the post that immediately followed it, another member made no reference to the point brought up by the previous commentator and presented the same idea in a way that suggests that she has either decided to ignore the other member's comment or simply had not read it and had come to the same conclusion inspired directly by the text. In the example below, the two posts are very close to each other and there seems to be no attempt to develop the topic first presented: This text has much of a whole spiritual dimension; it has much of similar to The Lord's sacrifice for mankind salvation. So has the Nightingale's sacrifice towards the young Student's Love salvation. (First post by P12) This story of love comes from the remote time, even in the bible. We can see that God loves the world and gave his only Son who was killed by the same people. (Follow-up post by P04) Another similar instance occurred in the thread on Tagore's The Parrot's Tale when a participant commented of the historical context of the short story creation and the second poster after her made a similar comment ignoring the previous remark: The moment I read this tale was first published in Bengali in 1918 it made me think of Bengali situation at that time. It was still a British colony and striving for independence (together with India and other places on that region) (First post by P23) Though the story was written long back in the context of the colonial education system of India, its appeal is universal. (Follow-up post by P62) However, instances where posters developed a topic previously brought up by another poster are much more frequent. Previous ideas were then presented under a different angle adding a different nuance to the discussion, directly or indirectly acknowledging other participants' contributions. The exchange below between two posters in the thread on Wilde's story may serve as an example: The Nightingale interprets things through her own emotions, for as Wilde himself had remarked, "No great artist ever sees things as they really are." (P19) [P19] has given us the key: literature sometimes is just like a mirror and Oscar Wilde saw nothing but absence of love in it, or maybe he did not find one. Literature is just a feeling we have in front of a mirror (P18) [P18]: Now that you speak of mirrors, it almost seems as if Wilde has held up an enchanting mirror through his story to reflect real life and society. (P19) It is important to mention that the relationships between readers in the forum were not limited to a single analytical category in each post as generally presented above. Instead, they are all interwoven and many times overlap. Posts often show a combination of features that denote both connections with the literary text and with other posts as exemplified in this long comment on Gunesekera's short story Goat: Hi folks (interaction), this was the first time I'd encountered Romesh Gunesekara and I loved this story. Thanks once again for leading me to another fine new (for me) author. I agree with many of the comments made here (interaction), especially the one made about the cultural mishmash present in the story. I'm always drawn to this as I've experienced first-hand something similar (adoption). But what struck me most on my initial reading of the story were the two characters of Byron and his friend (I've only now realised that he's not actually named in the story which highlights what I want to say) (interpretation). Well, I'm off to think some more about the role of the camera in the story. Thanks for those who pointed this out (interaction). I hadn't really noticed the importance of the cultural mixture (adoption). (P24) Analysis of the relationships between comments on the forum posted by different readers showed that participants used direct quoting and a range of interactive strategies to establish a conversation with their fellow readers in the ORG. These are the most noticeable distinguishing features of the interactions in the forum and are understood here as an attempt to extract from other group members a written reply that would contribute to the ongoing dialogue. At a deeper level, dialogic interaction is found in a poster's drive towards what others have said (adoption) evidenced by the borrowing and developing other poster's ideas, even when not openly acknowledged by later posters. Such borrowing of ideas may be the result of not having read the other's post, but it may also be that posters had reached such a high degree of agreement with what was said before, or had already internalized the message to a point, that they already saw these ideas as theirs. Instances of borrowing are, however, much less frequent than topic development. More common are the situations in which posters addressed an issue or topic related to the text from a particular viewpoint and others engaged in an unfolding exchange about it for a string of comments. It can be argued that this serves as evidence of the desire to establish a dialogical relationship with other group members and create an online community of readers.
Moreover, the internal 'stratification of discourse' and the 'diversity of social speech types' that Bakhtin (1981: 262) argues are present in the Novel are also observed in the discourse building strategies posters employ in the ORG forum. Participants adopted, at different occasions and often within the same comment, a variety of styles and authorial voices which can be also equated to some of the social and interactional roles performed by students in face-to-face ELT classroom based reading circles. The combination of direct quotes, interactive strategies and the inclusion of others' ideas in a participant's posts points towards the heteroglossic nature of their writing.

Conclusions
This paper examined to which extent participants' posts to the online reading group forum were shaped by the group interaction and which compositional features in their written comments indicated that. The findings of this study confirm previous research on interactions in face-to-face reading groups and online communities and show that there is a dialogical orientation in the communicative strategies that participants employ in the forum. This may help those interested in exploring computer mediated communication in teaching and learning to understand how teachers and TESOL students construct their comments when communicating in online professional forums and social media. Practitioners interested in conducting research projects with their own students in distance learning and/or blended courses may find the categories used to analyse posts in relation to each other useful to better understand their students' online interactions and communicative strategies. Online forum moderators working in educational contexts may find that the use of literary extracts and quotes can be a useful tool to trigger greater interaction in the forum and facilitate cognitive and emotional engagement with the topic proposed for discussion.
Above all, the findings of this study may be relevant for an assessment of the viability and usefulness of establishing online reading groups as tools to promote the creation and development of communities of practice. They may also contribute to a re-evaluation of the role of literature in language teacher education and the employment of alternative methodological approaches to the integration of technology mediated communication, literature and language in educational contexts.