Summarize and compare the main results described in the following two research articles (in three pages)

Article1_SchwarzEtAl2012-co-alienation.pdf

Home >Education homework help >Summarize and compare the main results described in the following two research articles (in three pages)

‘Co-alienation’ mediated by common representations in synchronous e-discussions

Baruch B. Schwarz⁎, Yifat Ben-David Kolikant, Maria Mishenkina The School of Education, The Hebrew University, Mount Scopus, Jerusalem, 91905, Israel

a r t i c l e i n f o a b s t r a c t

Article history: Received 10 February 2012 Received in revised form 23 August 2012 Accepted 18 September 2012 Available online xxxx

Synchronous e-discussions have become common social practices in and out of educational institutions. In comparisonwith face-to-face dialogs, synchronous discussions seem less propitious for learning. Yet, this social practice is extremely popular. Socio-cultural psychologists have suggested that intersubjectivity is central for maintenance of communication and for productive interaction in face-to-face social practices. In this paperwe study howcommunication ismaintained in synchronous discussions and whether intersubjectivity is reached in those discussions. Four university students used a CMC tool to discuss an educational issue on learning, teaching and moderation. One week after the discussion, each student was interviewed on his/her views on learning, teaching andmoderation. Then, the technique of cued retrospective reporting was used to uncover how each student interpreted eachmove of the synchronous discussion. The cross analysis of the interviews and the cued retrospective reporting showed that actions were not co-ordinated. Agreements and disagreements were not shared, and the order of actions was quite whimsical. We conclude that intersubjectivity was not established. However, communication was maintained through a process of co-alienation— the juxtaposition of incompatible alignments of representations through a common external representation. Although co-alienation is problematic, we show that discussants could learn from the e-discussion.

Keywords: Intersubjectivity e-Discussions Moderation Peer argumentation

Socio-cultural psychologists have studied many social practices that are relevant to learning and development. In these practices, the other is significant: he or she cares before guiding, he or she shares something that fuels productive interaction (Rogoff, 1990). The notion of intersubjectivity was elaborated to account for maintenance of communication in practices that lead to development and learning. The social practices of synchronous (electronic) discussions are newcomers that seem strange to educators or psychologists: People seat alone by their computers and interact with others they don't see and often don't know. The interactions are often lopsided, interrupted and rudimentary. Yet, sometimes, people seem to learn in or from these strange interactions. How can such learning occur? Is it possible to discern any kind of intersubjectivity in synchronous discussions? We approach these questions in this paper. A necessary step in this enterprise is to come back to the origins of intersubjectivity and to the different senses it has received to account for maintenance of communication.

Introduction: Four senses of intersubjectivity to account for maintenance of communication

The historical origins of intersubjectivity are philosophical: From Husserl, to Heidegger, Levinas or Gadamer, philosophers have asked the transcendental conditions of the possibility of dialog, of the existence of the other. These philosophical steps led to consider intersubjectivity as an idea that explains how empathy develops between people or how people reach new meanings together. Socio-cultural psychologists adopted this togetherness view of intersubjectivity to become a central idea in their theory of human development. Intersubjectivity was primarily meant to express a general idea of overlapping of subjectivities or prolepses

Learning, Culture and Social Interaction xxx (2012) xxx–xxx

⁎ Corresponding author. Tel.: +972 2 5881064; fax: +972 2 5880037. E-mail address: [email protected] (B.B. Schwarz).

LCSI-00020; No of Pages 16

Contents lists available at SciVerse ScienceDirect

Learning, Culture and Social Interaction

j ourna l homepage: www.e lsev ie r .com/ locate / lcs i

Please cite this article as: Schwarz, B.B., et al., ‘Co-alienation’mediated by common representations in synchronous e-discussions, Learning, Culture and Social Interaction (2012), http://dx.doi.org/10.1016/j.lcsi.2012.09.002

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

mailto:[email protected]

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

http://www.sciencedirect.com/science/journal/22106561

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

(Rommetveit, 1979, 1985)— communicative moves speakers take for granted things that not have been (yet) discussed. Even before language is mastered, a baby is born anticipating a sympathetic, interactive social environment, in which he or she will encounter fellow-human loving, care-taking and responsive subjects, available as companions in the baby's encounterwith theworld, interested in the baby's experience and enjoying the baby's responses or sharing his/her fears (Trevarthen, 1993). Another sense of intersubjectivity refers to the construction of a sharedmeaning by people to interpret social and cultural life, or a specific situation. This sense is very frequent in adults' guidance of children's explorations. It has also been adopted to describe some situations of collaboration among peers, for example, when jointly solving problems. A different sense given to intersubjectivity— shared divergences of meanings, is necessary to understand how practices such as lying, disputes and jokes “work”. Lying is genuinely inter-subjective as it operates between two subjective definitions of reality. It entails partially shared divergences of meaning: the liar shares with his/her interlocutors the meanings of a situation— divergent for him/her, he/she wants to “share”with them. Disputes are fueled by the common recognition of divergent views on a situation. So far, the senses of intersubjectivitywe reviewed have in common the sharing of subjective states by two or more individuals, and stress that shared cognition and consensus is essential in the shaping of our ideas and relations.

Still, all of us intuitively know that this communal view does not render the richness of human communication. And indeed, some distanced themselves from the sharing perspective in intersubjectivity presented so far to suggest ideas such as taken-as-shared beliefs (Cobb, Stephan, McClain, & Gravemeijer, 2001) to describe communication among children participating in successive social activities designed to promote learning. For Cobb et al. (2001), the term taken as shared “…leaves room for the diversity in individual students' ways of participating in these activities” (p. 124). But if these beliefs are, as posited by definition, subjective and not necessarily mutual and not shared by the community, how various social activities in which divergent views emerge may lead to new understandings, to new meanings? If Cobb is right, then he leaves room for diversity but also for mystery, as it is difficult to understand how communication can be maintained when nothing is necessarily shared. An additional sense of the notion of intersubjectivity was necessary. To make clear this sense, Matusov (1996) observed a succession of discussions on play craft among elementary school children mediated by a teacher. Matusov showed that in the disagreements some students raised no idea or no opposition to any idea. Rather, although a fierce disagreement arose among children, the disagreement for some turned around ownership of the play crafted while the other students thought it was motivated by a better play craft. Matusov concluded that a new sense of intersubjectivity should be elaborated, a process of coordination of participants' contributions in joint activity. For Matusov, not only agreement, but disagreement is at the basis of social activities if it is shared (and in this sense, it differs from Cobb's stance). Matusov gives examples for which understandings emerge from the coordination of contributions without any common belief. In those cases, an observer can extract from the social activity a thread of compatibility of actions, that is, an interpretation encompassing the goals of each participant, and which is coherent. In Matusov's terms, “The participatory notion of intersubjectivity as a coordination of individual contributions to the joint activity allows researchers to incorporate both participants' understandings and misunderstandings of each other, and their similarities and differences as the participants are simultaneously in agreement and in disagreement” (p. 29). Matusov considered other situations in which meaningful communication can be maintained — situations in which teachers or students care for others and help them understand, develop, or own what they already master. Communication in this case is maintained but calls for another sense of intersubjectivity, that of agency to features concerns by teachers (or students) that are shared by the ones about which they care (Matusov, 2001).

So far, among the senses given to intersubjectivity, the sharing senses do not seem to fit the experience we all have of synchronous discussions. Rather, the incorporation of compatible contributions/actions, compatibility being defined by the coordination of actions undertaken by the participants reflects more synchronous discussions. A different research direction done in linguistics may constitute a first step in studying how communication is maintained in synchronous discussion: Linguistics studies have recognized the role of grounding in attaining mutual understanding of utterances through coordination (Clark, 1994; Clark & Schaefer, 1989). The maintenance of common ground is done at different levels. It arises from contact (whether the interlocutor is willing and able to continue the interaction); perception (whether the interlocutor is willing and able to perceive the message); understanding (whether the interlocutor is willing and able to understand the message); attitudinal reaction (whether the interlocutor is willing and able to react and adequately respond to the message, specifically whether s/he accepts or rejects it). For Clark, grounding cannot be undertaken through isolated utterances but through patterns of interactions: for example, contact is maintained through a three-turn pattern between discussants: one posts a message, the other acknowledges his/her attendance, and the first continues. Baker and his colleagues (Baker, Hansen, Joiner, & Traum, 1998; Baker, Hansen, Joiner, & Traum, 1999) have related this tradition with the socio-cultural historical tradition and identified in grounding a basic mechanism that enables collaborative learning and establishes intersubjectivity as a process of coordination of participants' contribution in joint activity: Learning from grounding – collaborative learning – can be viewed as appropriation of semiotic tools, mediated by those very tools. Appropriation of cognitive tools can be viewed as a gradual transition from pragmatic level grounding – learning to understand each other, to collaborate – towards learning to understand the semiotic tools in a specific domain (that may also be languages themselves) (Baker et al., 1999).

The studiesmentioned above on the different types of intersubjectivity involved face-to-face communication. The studies done on the role of grounding in establishing intersubjectivity involved face-to-face or a-synchronous communication. The diversity of communication practices grows extremely rapidly, as various forms of e-discussions gain popularity. For those practices, especially for some cases of e-discussion, wewill argue that groundingmight be questionable if we view intersubjectivity as the coordination of contributions in joint activity. We will also show that the sense of intersubjectivity as sharing of subjective states among several individuals, cannot be retained for certain synchronous discussions. The question is then how communication ismaintained?And is it

2 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

possible for discussants to learn from such discussions? Those are harsh questions that we will approach in this paper in the case of synchronous e-discussions with a particular tool for facilitating collective argumentation in synchronous discussions (Schwarz & Glassner, 2007). This particular tool is representative of a quite large set of tools for facilitating e-discussions that displays a representation of the on-going discussion, including its whole history. We will claim that this gradually produced representation (called here an argumentativemap) which is shared by the discussants while participating in their discussion, enablesmaintenance of meaningful communication among them, although possibly no idea is shared and no coordination of actions is attained during the discussion. In an example, we will show that the argumentative map produced by the discussants will serve as an artifact for projecting personal interpretations of the synchronous discussion so far,whichwould have probably led to ruptures if communication would have been face-to-face. We will point at the cohabitation of incompatible prolepses, or what we call co-alienation along with new kind of communication.Wewill show that the discussants can learn from such discussions.Wewill discuss whether in this case, a new sense should be conferred to intersubjectivity. But, before we delve into the analysis of the synchronous discussion, we first review what is known in communication studies about synchronous e-discussions. As we will see, this review appears to be highly relevant to the issue of maintenance of communication.

Features of synchronous discussions: New turn-taking rules, social (lack of) presence, complexity, persistence and attractiveness

Apriori, it seems that e-discussions are detrimental tomaintenance of communication, hence, by definition, to intersubjectivity: In spoken discussions, joint attention is very often maintained. Turn-taking is central in this endeavor as turns are determined by facial expression, pause of the other, intonation, or simply interruption. Also, turn-taking in spoken conversation follows a normative ideal of precisely alternating turns. The word ‘precisely’ refers to the timing of the transition from one speaker turn to the next, which is ideally supposed to occur with no (or minimal) gap, and no overlap between speakers (Sacks, Schegloff, & Jefferson, 1974). ‘Alternating’ refers to the expectation that participants will take turns speaking in an orderly fashion; thus in dyadic exchanges, one person speaks, the other responds, the first speaks again, the second takes the following turn, and so forth. In small group conversation, overlaps can occur but they generally occur throughgestures, facial expressions, etc. In sum, numerous cultural tools are at the disposal of participants of spoken discussions in general and collective argumentation in particular, to maintain communication. In contrast, descriptive studies of computer-mediated communication (CMC) suggest that this maintenance may often turn out to be problematic. These studies exhibit numerous violations of both the “no gap, no overlap” principle and the principle of orderly turn alternation, unless floor control in turn-taking is ensured (automatically, by a tutor or by the discussants themselves). Concerning gaps, there is often a considerable time lag between when a message is sent and when it is responded to, especially in asynchronous forms of CMC. Synchronous CMC involves more rapid exchanges of turns, but delays may be caused by disrupted turn adjacency. Overlap in CMC is also problematic. In dyadic communication, users – unable to tell whether their interlocutor is in the process of responding or not –may become impatient and send a secondmessage before a response to the first has been received, resulting in incomplete or interleaved exchange sequences (Condon & Cech, 1996; Marvin, 1995). In group communication, unrelatedmessages fromother participants often intervene between an initiatingmessage and its response (Murray, 1989). According to Herring (2001), these problems are responsible for incoherence, and for topic decay — the fact that discussants rapidly discuss less and less the topic at stake during e-discussions. For us, these violations suggest that presence in an e-discussion maynot necessarilymean attendance, at least in real time, to turns. Of course, in spokendiscussions too, peoplemay not always notice that one of the interlocutors intervened, but this phenomenon is inherent to synchronous CMC: discussants do not feel obliged to answer a question directed to them, especially when they are busy answering to other participants.

Despite what could have been seen as numerous shortcomings of synchronous CMC systems as conversational environments, students often see them as more socially desirable than analogous face-to-face interaction—in Walther's (1996) term, interaction is ‘hyperpersonal’ rather than ‘interpersonal’ interaction: weakening of coherence between messages induces humorous style. The fact that in contrast with spoken discussions, no norms are prescribed on discussants enables them to participate in parallel discussions and to play. Reduced feedback and loosened adjacency enable a qualitatively different kind of interaction from spoken conversation, and this contributes to CMC's ‘hyperpersonal’ appeal. To grasp the social character of CMC messages researchers have defined the term social presence as “the salience of the other in a mediated communication and the consequent salience of their interpersonal interactions” (Short, Williams, & Christie, 1976). Several researchers studied the contents of CMC messages to show the presence of affective as well as interactive–cognitive expressions (Anderson, Rourke, Archer, & Garrison, 2001). A three component tool was elaborated for measuring social presence: (a) emotional responses that include uses of emoticons, humor, self-disclosure, etc.; (b) interactive responses that include continuation of a thread, referring to or quoting from others' messages, asking questions, complimenting or expressing agreement, etc.; (c) cohesive responses that include phatics (expressions of sociability), vocatives, etc. With such a tool, Rourke and colleagues (Rourke, Anderson, Archer, & Garrison, 1999) could show that students enjoy synchronous discussions more than a-synchronous e-discussions (which are deeper and more filtered, but with less emotional and more cohesive responses). Thus the “attractions” of CMC can be seen as the flip side of the “incoherence” coin—loose inter-turn connectedness and overlapping exchanges have both advantages and disadvantages, depending on the purposes for which users engage in computer-mediated interaction.

This attractiveness is not the only advantage of CMC systems over face-to-face communication: CMC communication is based on the availability of a persistent textual record of the interaction. Even the least persistent synchronous interface is more persistent than spoken language, which disappears immediately once it is uttered. Persistent conversation appears to aid the user's cognitive processing. The predilection towards meta-humor andmeta-play in CMCmay be attributed in part to the fact that CMC persists as text on a screen and is subject to conscious reflection in ways that spoken language is not, thereby facilitating a

3B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

heightened meta-linguistic awareness (Cazden, 1995). With persistent textual records, users can participate in simultaneous multiple interactions without getting hopelessly lost or confused.

The short review we undertook suggests that the issue of maintenance of communication in synchronous discussions is a challenge: On the one hand, there are many shortcomings in synchronous discussion tools that may impair sharing beliefs or coordinating actions, hencemaking it difficult tomaintain high-quality communication. On the other hand, this medium is attractive, and its persistence enables simultaneous multiple interaction. The persistent on-going record of interaction between discussants should play a central role in the maintenance of communication in synchronous discussions.

Description of the research: Studying maintenance of communication in synchronous discussions in a university course

The context of the present study is educational. Discussions took place in a University course on The Role of the Teacher in Classroom Learning. One of the themes of the course was to focus on the role of the teacher in structuring interactions in group work. In addition to theoretical lectures, the teacher provided a CMC graphical tool for facilitating group e-argumentation, and for reflecting on ways teachers could intervene during discussions (Schwarz & De Groot, 2007). This tool named Digalo enables the management of discussions and the representation of their argumentative processes and components among participants. Using Digalo consists of co-creating maps built of written notes inside different shapes that represent the permissible argumentative moves for the discussion (or what is called the ontology chosen for the discussion), and different arrows representing different connections between the shapes. The ontology chosen in the course included Claim, Argument, Explanation, Comment, and Question. There are three kinds of arrows: support, opposition, and link. The support and opposition arrows help challenging, refuting, elaborating, or agreeing. In each of the contributions, participants add one shape or more and arrows/links to shapes built by others to articulate own claims, arguments, etc., and then write their contribution inside of the shape(s) they chose. Each participant can locate his/her contribution wherever on the screen and change the location of previous shapes. Each participant receives a distinctive color and an icon that help identifying his/her contributions on the board. Fig. 1 shows the on-going construction of the Digalo map on which we will focus in this paper. Digalo enables three types of synchronicity: (a) synchronous e-discussions with floor control, (b) synchronous discussions without floor control, and (c) a-synchronous discussion. Schwarz and Glassner (2007) compared the characteristics of synchronous discussions with vs. without floor control, with and without the

Fig. 1. The Digalo map of the discussion between Rim, Judith, Fatima and Ahmad in its entirety.

4 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

ontology described above. Unsurprisingly, they found that the characteristics were deeply influenced by the conditions under which synchronous discussions were set.

Participants

11 graduate students participated in a university course the first author delivered. They differed in ethnicity, and gender. This article focuses on one of the discussions. The names of the discussants are fictitious but reflect the ethnicity (Jew, Arab) and the gender of the participant. The ethnicity is not blurred because it turned to be highly relevant to the discussions. Judith is a 26 year old Jewish graduate student in Educational Psychology. She also attends a pre-service teacher program in psychology and in biology. She uses to teach special education young delinquent girls. Fatima is a 25 year oldMuslim student in Education. Shewas accredited as a teacher at elementary school level and worked as a kindergarten teacher in the Arab sector with toddlers with problems in language development. Ahmad is a 47 year old teacher. He is an alumnus from the Hebrew University in Arab literature, and is a graduate in political sciences. He serves as a teacher for 20 years, 8 of them he taught Arab literature in high-school and junior high-school levels in the Arab sector. He also taught Civic Education, History and Arabic language. In the past he taught in one of the bi-national schools in Israel. At the time of the present study, Ahmad worked as a teacher in History and in Civic education in East Jerusalem. Rim is a 31 year old Muslim graduate student in Hebrew language and Education. She serves as a teacher of Hebrew language in an Arab school in grades 4 to 6. She also served as a project coordinator in special education (in the Jewish as well as in the Arab sector).

Procedure

In this study students participated in synchronous discussionswithout floor control. In a first session, students were introduced to Digalo and conducted two “warm-up” discussions during which they learned about technicalities and became accustomed with the ontology we chose. In the first warm-up activity students were arranged in three groups of 3–4 and were asked to resolve a moral dilemma. The presentation of the ontology was accompanied by a suggestion to follow a series of ground rules of conversation developed to instill dialogic and dialectical talk (“You should present a view point”, “you should provide reasons to support your viewpoint”, “You should challenge any idea”, “you should refer to the arguments of your peers with respect”, “you should answer to challenges or questions”, etc.). On the followingweek, the studentswere arranged again in three groups. This time theywere asked to discuss an educational dilemma.

At the end of the first session, the students were given two articles to be read before the next lesson. Both articles related to the issue of guidance in discussions. The article by Howe (2009) showed the importance of giving instructions in advance such as reaching consensus in scientific discussions and that without such instructions students may go astray. The article by Asterhan and Schwarz (2009) showed that to be productive, unguided argumentation should realize a balance between harmonious social relations and intransigence towards reasoning criteria and that this balance is extremely difficult to realize. Both articles suggest that the total absence of guidance during discussions is rarely productive but that moderation, a kind of guidance which is caring but not intrusive, may be one way to help students in their discussions. The discussion groups were chosen in advance. One moderator was designated for each group, and was asked to insure that discussants follow the ground rules we presented during the discussion. Themoderators were chosen to reflect the sectors to which the students belonged: one Arabman, one Jewish woman and one Jewish man.

One week after, the lesson on which we focus in this paper took place in the computers room. All students sat by their own computer at distance from each other and were explicitly asked not to communicate verbally.1 The issue to be discussed was “Is it indispensable for research on classroom learning to account for the teacher's activity?” The discussants were asked to capitalize on the resources they already read and to follow the ground rules presented the week before. Two experimenters helped in technicalities. The discussions lasted around 30 min.

Two weeks after the discussions, we set a meeting with each of the discussants. This meeting consisted of two parts. The first part of the meeting was a preliminary semi-structured interview in which the third author asked the students on the role of discussions in learning and on the role of the teacher in discussions: “After the Digalo discussion, what do you think of it and what did you learn about discussion with Digalo tool?”, “What did you learn about the role of discussions in learning and about the role of the teacher\moderator in the discussions?”, were the main questions asked during the interview. In the second part of the meeting, each student underwent a cued retrospective reporting (Van Gog, Kester, Nievelstein, Giesbers, & Paas, 2009) in which he was presented the progressive reconstitution of his argumentative map by using the replay option of the Digalo software: whenever the interviewer clicked, the map grew by one argumentative move. This technique has been particularly fruitful to reconstitute what participants think during e-discussions (e.g., Schwarz & Asterhan, 2011).2 Interviewed discussants were asked to explain their moves, thereby give their interpretation of each argumentative move, such as peers' or moderator's interventions, own reactions to these interventions, and to figure out the goals and expectations, and their or others' concrete decisions. Each of the meetings lasted around two hours. They were all recorded and fully transcribed. We will use interchangeably the term cued

1 The absence of verbal communication naturally occurs not only when students are distant, but even when they are at the same place. The instruction not to communicate verbally insured the deployment of this practice.

2 Cued retrospective reporting involves harsh epistemological questions: To what extent does the interaction with the experimenter reveal “what they thought in the first interaction” and/or their adaptation to the experimenter–interlocutor? Does this really reveal “what they thought” during the interaction? The semi- structured interview that was around the issues discussed in the interaction minimized possible discrepancies and validated to a large extent our interpretations.

5B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

retrospective reporting and auto-confrontation (Mollo & Falzon, 2004) in the paper to stress that the methodological aspect of the meeting has also educational implications.

Analysis of the discussion: three incompatible but cohabitating stories by three discussants

Howwas the discussion perceived by the participants in eachof itsmoves? To answer this difficult question,we had at disposal the Digalomap they gradually produced, and the auto-confrontation of Judith, Fatima, and Ahmadwith the replay of themap (the fourth discussant, Rim, did not agree to participate in the auto-confrontation). The preliminary interview of the discussants provided an indispensable background to interpret the e-discussion. To keep this paper short, we do not report on the beliefs of the discussants on the role of teachers and learners in classroomdiscussions, as theywere identified in the interviews.Wewill insert some insights from the interviews to complete the interpretations of the discussion by the discussants in their cued retrospectives.

As we will see, interpretations were quite divergent: they did not convey shared understandings, or even coordinated actions. Since, as we will show, the nature of communication was so precarious, the division of the discussion in episodes was a difficult enterprise: somemoments could represent boundaries for some discussants, while they could be apprehended as in the middle of a move by others. Our approach was to propose a division into episodes according to clear changes in patterns of interaction. The discussion includes 49 contributions. The first episode — Turns 2–20 is characterized by the non-intervention of the moderator, and the quite equal participation of the three discussants. The second episode— Turns 21–37, is characterized by the participation of all discussants, including the moderator, who sends messages to all discussants. The pattern of interaction is quite centralized. In the third episode— Turns 38–49, the moderator continues to sendmessages to all discussants but only one discussant is responsive. We begin our analysis by a first episode in which the moderator does not intervene.

1st episode: starting the discussion without moderator — accumulating similar contributions

Fig. 2 shows the beginning of the discussion (Fig. 1 shows it in its entirety). It is characterized by short contributions in which no challenge is raised. Rather, the map seems to grow with contributions that express the same point: at Turns 8 and 13, Fatima writes In my opinion research should refer to the position of the teacher, no one can overlook him! and I agree with you (what is written in the e-discussion is in underlined italics to differentiate it from what is said in the interview); at Turn 11, Rim writes In my opinion, any research on learning and instruction should refer to all the directions, students, teachers, parents, school. Judith writes at Turn 12 We all agree. Is it possible? and ostensibly links this recognition of agreement to all previous interventions with arrows of support. Ahmad did not make any contribution in the first episode because he faced technical problems. However, as he later reported, he watched the map, and thought how to act in this situation.

We will see that all participants interpreted this situation differently and that their interpretations relied on their beliefs regarding the role of discussions in learning, and the role of the moderator and of the discussants in e-discussions. Let us begin with how Ahmad interpreted this discussion in the auto-confrontation.

Ahmad: discussants do not really refer to each other; they agree instead of thinking in depth

When Ahmad observed this discussion in his auto-confrontation, he reacted: the participants began by throwing out ideas… and then everybody agrees, agrees, agrees, and in fact, I don't know about what. In his view, what happens here is disagreement which is called an ‘agreement’. When Ahmad read Judith's contribution at Turn 12 (We all agree. Is it possible?), he felt that this means that we are afraid of ourselves. Is it possible that we agreed? On what did we agree?We agreed on one sentence […], the fact that the teacher should be at the center of research and should look at the students. But we didn't formulate this properly. He recalled the importance of collaborative learning to contrast it from the ‘railroad parallel tracks’ (sic) that characterized the beginning of the talk. Ahmad thought that discussants should slow down their pace and begin to read others' contributions to uncover their real meaning.

Ahmad raises an additional idea, the importance of formulating own opinion or of creating own position. He criticized the Arab society to which he belongs, that does not enable young people to think autonomously: the Arab society is clearly patriarchal. The father decides on everything, and everybody has… sometimes, this is not the father, this is the grandfather […] and everybody must tell him ‘yes’ and nobody argues with him, nobody argues with him about things that are important in their lives. The society was built in a way that only when you get old, you can think, you are allowed to think. Before you are old, you must listen, you must accumulate wisdom. It seems that Ahmad sees the problems of his society through the lenses of Rim and Fatima's contributions. In the next episodes, we will see how Ahmad decided to act to improve the quality of the discussion. The way he chooses reflects what he thinks about what a good discussion is. But first of all, let us see what Judith has to say on the same episode.

Judith: it is impossible to learn from this discussion

As she read the contributions of the other discussants in her auto-confrontation, Judith reminded that she had no interlocutor for a real discussion. For example, her reaction to Rim's contribution at Turn 7 (I agree with your opinion Ahmad that the teacher should be a partner) is: It seemed to me that it's not enough articulated […] it means to say ‘yes’ to the main issue, but this is not enough […]. I felt that I don't have any interlocutor with whom to argue. She compares her pace in her contributions to theirs to say: see how much I write [laughing]…I felt that they don't move. Judith was interested in changing this situation: I tried again and again, as much as I could, to move things. She commented on her contribution at Turn 12 (We all agree. Is it possible?), that she linked with arrows

6 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

of support to all previous contributions as ‘a bit cynical’. For her, the contributions of the others are like building bricks, and each discussant should bring personal contributions of high quality, original, and warranted: When you express only your opinion, without going deep into it, without explanations, without bringing citations, warrants to what you say […], it's superficial, it's to remain in ‘I think that’. It is then understandable that she is disappointed by the beginning of the discussion.

In spite of her disappointment, Judith continued reacting to her interlocutors and expressed her opinion with the hope that it would develop as a more interesting discussion in which she will have opportunities to explain her position. Differently from Ahmad, she put the responsibility for the learning in the individual rather than in the group. However, she thought that the discussion quickly became purposeless. To sum up, Judith did not see the discussion in the same utterly pessimistic way as Ahmad; she was worried, but although she perceived it as shallow, hoped to personally grab later on interesting ideas from her discussion. We will see that Fatima grasped the e-discussion very differently, in fact very positively.

1. Question Is it indispensable for research in learning and instruction to address the teacher's activity to be meaningful in educational practice? 2. Judith Claim neutral link to 1 Yes and no. No – because there is also independent learning. Yes – because even the independent learning is organized by the teacher 3. Judith Claim [contribution without content posted and immediately deleted]

4. Judith Comment [contribution posted and immediately deleted] Even if the research does not study the teacher directly, it still should use him or refer to him, isn't it? 5. Judith Claim support to 2 Even if the research does not study the teacher directly, it still should use him or refer to him, isn't it? 6. Ahmad Argument [contribution without content posted and immediately deleted] 9a. Judith Argument opposition to 6, 7 and 8 that are contributions without content by then Would someone just write something!!! 7. Rim Comment support to 5 I agree with your opinion Ahmad that the teacher should be a partner 8. Fatima Claim neutral link to 1 In my opinion research should refer to the position of the teacher, no one can overlook him! 10. Judith Claim neutral link to 7 I didn't understand 11. Rim Argument support to 7 In my opinion, any research on learning and instruction should refer to all the directions, students, teachers, parents, school 12. Judith Comment support to 7 and 8 We all agree. Is it possible? [The content in this contribution appears simultaneously with the content of contribution 13] 13. Fatima Comment support to 5 and 12 I agree with you

Fig. 2. Starting the discussion without moderator — accumulating similar contributions.

7B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

Fatima: the discussion is successful because people express themselves and the moderator is not intrusive

As Fatima read Judith and Rim's contributions, she thought that the discussion was excellent, not only good at that stage: each of the discussants expresses what she has. For her, this situation is surprising and does not necessitate any moderator. Fatima asserts that the discussion is successful because it does not stop after the first contribution: They [the discussants] forgot that there is no moderator and continue talking onward. When reading Judith's Turn 12, she understands that Judith wants to serve as themoderator. She does not identify any cynicism in Judith's intervention (asmeant by Judith). Rather, she interprets this turn as a sincere request to figure out the intentions of the discussants: She [Judith] wanted to take the role of the moderator who tries to understand whether we all are on the same wavelength, whether we agree on the same things or not. In her opinion, this role is not necessary: the discussants are interested in this topic … We want to go on although we have no moderator.

Fatima's satisfaction is consistent with the opinions she uttered in her interview: for her, discussions contribute to learning by the simple expression of diverse opinions. She expresses her surprise that the discussion goes on without any guidance, since, according to her experience, students do not speak and the role of the teacher is to cause them to speak. Understandably, since students speak, the moderator is unnecessary.

Interim summary

The interpretations of this first episode were divergent: For Fatima (and probably Rim), the expression of disparate opinions is highly valuable. It conveys her belief about learning as an incremental process of completion of information by someone who intentionally wants this information to be integrated, and her beliefs about the role of discussions to provide opportunities to individuals to change their mind. This is not Judith's view that collaboration (including with the moderator) leads to interesting and founded arguments. Her cynical remarkWeall agree. Is it possible? aswell as her disappointment reported in her auto-confrontation is not shared by Fatima who is enthusiastic. And Fatima's satisfaction about Ahmad's restraint is of course not shared by Ahmad who waited to overcome technical difficulties to intervene in order to improve the quality of the discussion throughmeaningful references to all opinions and through the co-construction of newknowledge. The divergences betweendiscussantswere not shared as discussants had different interpretations about interventions: forced silence of the moderator was interpreted as willingness to let discussants free of any guidance, Judith's cynicism and discontentment were not perceived by some, etc.

Second episode: Ahmad tries to change the flow and the quality of the discussion

The discussion which began with ungrounded agreements between the discussants, developed into a harsh dispute (see Fig. 3): Fatima and Rim argued that the student is at the center (Turns 14, 16, 19), against Judith who argued that opposed this view and added that it is impossible to take into consideration all the variables that are relevant to the educational field (Turn 18). Initial agreement quickly turned to a harsh polemic. At this point, Ahmad succeeded to enter the discussion. At Turn 21, hewrites: Let's make some order here, we should know on what there is an agreement. At Turn 27, he intrudes into the gist of the discussion by asking the question:What do you think about the saying according to which the student is at the center and the teacher only disturbs in his learning? However, the discussants do not change the way they discuss the issue. The map fills up with short interventions popping out at a dizzying pace. Ahmad, who tries to stop this trend by writing (in Turn 30) Please stop throwing out things without seriously referring to what is written immediately receives Rim's reaction We don't simply throw out (Turn 31). In Turn 33, he links Turn 31 to her contribution in Turn 11where she previously wrote In my opinion, any research on learning and teaching should take into consideration all the directions— students, teachers, parents, school, to ask (in Turn 33) U1 what about opening the idea in contribution 11. This subtle move seems to point at Rim's incoherence, but she is not sensitive to it.

Very often, disagreements are signs of a high level of intersubjectivity as what fuels them is a shared divergence of meanings (Matusov, 1996). At a first glance, our general description of the episode seems to show such a divergence ofmeanings. However, we will show that understandings are local, subjective, and the participants do not react in relation to the meanings attributed by their interlocutors. They coordinate their actions, but this coordination is done according to interpretations that are totally divergent.

Ahmad: efforts to organize the discussion, and to foster collaborative learning

In his auto-confrontation with this episode, Ahmad complains that the lack of reference to others' ideas originates from the fast pace of the discussion: Too much, too fast, things they threw up. I had the feeling that it's raining. Ahmad tried to lead discussants to scrutinize the ideas of each other and to refer to previous contributions expand them, or, in his own terms, to open them. In his opinion, the presence of different discussants facilitates the consideration of multiple perspectives and naturally leads to the necessity to explain. Consequently, Ahmad tries to slow down the pace of contributions. His comment of his contribution at Turn 30 (Please stop throwing out things without seriously referring to what is written) is: So, I explained that you should be more focused, more ordered […], that you should stop, that you should think. Because there was such a deluge of contributions. This comment conveys a quite high emotional state against the behavior of the discussants.

We already saw that Ahmad entered in the middle of the discussion (e.g., in Turn 27: What do you think about the saying according to which the student is at the center and the teacher only disturbs in his learning?). Viewing Turn 27 he explained: I came to tell thosewho agreed: listen… there are other viewpoints…From the beginning, I didn't knowonwhat they agree and onwhat they disagree. I try to position a conflicting viewpoint, an antithesis…to arouse the issue.

8 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

Reaction to Ahmad in Ahmad's eyes: the discussants do not agree with his description of the situation

In his auto-confrontation with the reaction to his injections, Ahmad assessed that the discussants adopted a negative position: When in Turn 31, Rimwrites:We don't simply throw out, Ahmad comments: This means that Rim thinks that what she says is important and is not superfluous. This means that she doesn't agree with what I said [at Turn 30]. Ahmad considers Rim's contribution as a disagreement on the definition of the situation in which the discussants are evolving. To handle this disagreement, he attempts to

21. Ahmad Argument neutral link to 1 Let’s make some order here, we should know on what we agree and on what we don’t 22. Judith Claim opposition to 17 and support to 18 And not every research in learning and instruction should include the teacher 23. Fatima Comment contribution without content posted and immediately deleted. 24. Fatima Question neutral link to 1 Who else wants to add to the central question? 25. Judith Comment neutral link to 8 Of cause they are important, but there is a difference between the fact that they have impact and the claim asked here – should every research take into consideration the teacher. Not every research should take into consideration the parents!!!!!! 26. Rim Argument support to 21 There is agreement Ahmad that the teacher should be included 27. Ahmad Question no link What do you think about the saying according to which the student is at the center and the teacher only disturbs in his learning? 28. Judith Comment no link I don't understand what is going on here There is too much mess!!!!!! 29. Rim Comment support to 27 The student is at the center but the teacher does not disturb but directs the students and guides them 30. Ahmad Comment neutral links to 11 and 24 Please stop throwing out things without seriously referring to what is written 31. Rim Comment neutral links to 27 and 30 We don’t simply throw out 32. Ahmad Question neutral link to 29 Judith what does it mean the student at the center? 33. Ahmad Comment neutral link to 31 Judith what about opening the idea in contribution 11

Fig. 3. Second episode: Ahmad tries to change the flow and the quality of the discussion.

9B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

model his viewpoint in Turn 33 bywritingU1what about opening the idea in contribution 11with a link to Rim's opposition at Turn 30. Ahmad comments:

I linked this [Turn 33] to 31 to say that if people don't refer to that, this means that they write for no reason […] because she said that this is not without any purpose. And Ahmad continues: her contribution [Rim's contribution] at Turn 11 in which she says ‘In my opinion, any research on learning and teaching should take into consideration all directions, students, teachers, parents, school’ is good, there is a lot of content here, but one should develop it, I mean that we should elaborate upon it, explain and warrant it. It was written and no one referred to it. It's a pity that it passes by without mention […]. This was my intention, that if one does not develop these ideas, he writes without any purpose. (Rim did not react to this intervention).

Ahmad's comments show how he used the Digalo map to convey suggestions to the discussants to improve the quality of the discussion by referring to previous contributions. When Ahmad affirms I linked this to 31 to say that it was written and no one referred to it so it has no sense, he makes clear that the growing map is a central artifact for constructing a shared meaning. The way Ahmad reacts to Rim's disagreement totally fits his positive approach to criticism and his demarcation from blind agreement clearly stated in his interview. In the rest of the discussion too, Ahmad acted in diverse ways in order to lead discussants to refer to what their peers write. He did that by requesting them to refer to specific contributions, by presenting an idea that opposed what the discussants wrote to one discussant or to all of them. To sum up, Ahmad's comments on his own actions and on the actions of others convey a coherent picture of an educator who is dedicated to help in learning as an interactive endeavor in which diversity of ideas and disagreements are opportunities to co-construct knowledge.

Judith: the moderator was not noticeable

Judith attempts to justify the fact that she and her peers did not react to Ahmad's contributions by pointing at technical mistakes that Ahmad did, despite his good intentions. For example, in her comment on Turn 30 (Please stop throwing out things without seriously referring to what is written), she admits that this intervention can help the group in organizing and promoting the discussion: The moderator belongs to the group. And he represents the goal which is common to all members of the group […] to take a decision, or to reach a conclusion. And this is the role he has — to help in this goal. Judith comments that in his interventions Ahmad complies with his roles of moderator. However, Judith estimates that he failed because he did not benefit from the visual aspect of Digalo: He says important things, but does not locate himself properly … he does not think about that. Do people listen to him, don't listen to him, see him or don't see him? And indeed, it seems as if Judith did not notice Ahmad's contributions at all and she assumes that the same happened to the rest of the discussants. In her comment on his contribution in Turn 30, she explains: I didn't see him at all; in my opinion, it's very important, and if nobody notices you, you can't organize a discussion. She suggests that Ahmad could have created a personal reference at the right place in the middle of the map and he could link it to everybody's last contributions with arrows; not to put himself like that, on the side, in the corner. For Judith, the growing map is a central artifact for constructing shared meaning. Judith does not put all responsibility for this failure on Ahmad but mentions that the map is in disorder, and contributions hide one another, something which makes it difficult to organize a discussion.

Besides this critique (also aimed at justifying her behavior during the discussion), Judith praises Ahmad for his efforts to improve the quality of the discussion. While she notices that his contribution in Turn 30 (Please don't throw out things without seriously referring to what is written) is a bit harmful, she asserts that the idea is good: the discussants really do not devote enough thought on their contributions. Although she claimed that she didn't notice Ahmad, it seems that Judith reacted once to Ahmad's effort to promote the quality of the discussion, and when he wrote that one should take advantage of what the authors of the papers wrote to see what pros and the cons are to write in Turn 36: We saw in Howe's research that interventions by the computer are equivalent to interventions by the teacher. But I claim that still, a teacher was necessary to organize this computer event. So it's impossible not to refer to the teacher. We have here one of the rare moments in which a reaction is made on the basis of a shared understanding. Also, as Judith attends Rim's disagreement and Ahmad's reaction, her interpretation is: The moderator tries a new way, as he didn't succeed in explaining himself […] He shows her what he means, what is my problem with what you said. This interpretation fits Ahmad's explanation of his act. But Judithwas surprised to see Ahmad trying another way to lead discussants to act more effectively, rather than simply doing it instead of Rim. The difference between Ahmad's and Judith's reactions corresponds to their perception of the role ofmoderator: both of them claimed that themoderator is a participant. However, Judith lacks the vision of a moderator as a manager that Ahmad has.

In her auto-confrontation (and probably not in her discussion in vivo) Judith appreciated the way Ahmad tried to promote the quality of the discussion. However, she sees in his failure a lack of clarity: While Ahmad considered his interventions as perfectly clear and acceptable, this feeling of clarity was not shared by Judith who suggested a more egalitarian style of moderation (although she disdained her peers for what she considered as their shallow contributions). We will see in the following that for Fatima, Ahmad's intervention was very clear but unacceptable.

Fatima: the intrusion of the moderator turns the discussion to uninteresting

In her auto-confrontation, Fatima is quite resolute about her feelings towardswhat she sees as the intrusion of themoderator. She considers Turns 30 and 33 as examples of bad guidance. Turn 30 (Please stop throwing out things without seriously referring to what is written) is totally unacceptable: This is not OK…he was not with us at the beginning, and he can't tell us such things. She justifies Rim's

10 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

reaction in Turn 31 (We don't simply throw out). Such an intervention from the part of the moderator does not encourage the continuation of the discussion. She interprets Ahmad's link to Rim's contribution in Turn 31 in his contribution (Turn 33:U1what about opening the idea in contribution 11) as away to ask questions on claims thatwere far away and hewants us to return to Turn 11, what's the link? And Fatima goes on explaining: He has to know that many things happened since then. So, his reaction here [in Turn 30] is not interesting at all and is connected to nothing [Fatima does a gesture of ‘going away’with her hand]. Fatima feels that the rest of the activity of the moderator is even worse as he does not show any interest in her and Rim's reaction. Her feeling is based on the pace of his contributions in Turns 30, 32, and 33: He didn't wait for our reactions. Rim reacted but it doesn't seem that he waits that somebody else reacts, or asks a question anymore. This kind of behavior leads Fatima to stop contributing, yet she kept reading the discussion, which she found to be “not very interesting”: I only read. In the discussion, I read all the contributions and this was not very interesting. Fatima's feeling ismediated by the fast development of themap (done by the other participants) which conveys for her a lack of consideration from the part of the moderator. Fatima's interpretation is consistent with her feeling of satisfaction in the first episode, when Ahmad did not intervene. It also fits what she said in her interview about the detrimental role of the teacher in discussions between students. In accordancewith this view, Fatima opts not to react to Turn 30 that she considers as harmful. Rather, she reacts to Turn 27 (What do you think about the saying according to which the student is at the center and the teacher only disturbs in his learning?). She explains this choice by asserting: In that way, he doesn't say that all what we said before doesn't matter or is nothing…in some way, he took what we already said and the main issue and assembled them together. He gave us a second question, a question more or less connected to the main question and to what we said before. Fatima's answer in Turn 35 (The student always at the center, and the parents and the teachers around) is quite laconic of course and unsatisfactory at a university level.

We see then that the divergence of interpretations of this episode was large. We also showed that these divergences were not based on shared understandings as each participant interpreted his/her interventions and (lack of) reactions to them according to divergent beliefs about learning, and the role of the moderator in facilitating learning processes in discussions. We also alluded to the fact that themaintenance of divergences that are not based on shared understanding also relates to the tool used: For example, the fact that contributions are persistent (in contrast with face-to-face communication) led Ahmad to expect commitment to what has been said in the past, even if remote, while Rim and Fatima saw this expectation as unacceptable and even insulting. For some the high pace of interventions of the moderator clearly suggests a lack of consideration towards discussants, but for others, a high commitment and close consideration. The complexity of the display stresses that presence is not a trivial matter and that interlocutors may not be noticed.Wewill return to these points in our general discussion. But let us look at the third episode, the end of the discussion. This episode will complete the picture about the weird kind of maintenance of communication that synchronous discussions may enable.

The end of the discussion: a common summary or the moderator's aggressive coercion?

The dynamics of the end of the discussion changed in comparison to what happened so far: The moderator referred to Judith's contributions, and she reacted to them. In these interactions, Ahmad gave to Judith ideas that complete or challenge the ideas she brought forward. Judith identified this new style as that of a challenging moderator. Judith responded by adding more ideas or more details to ideas already expressed. She felt that there is a learning layer in the discussion, and also learns from the inventive strategies he adopts to go deeper in the discussion and to improve its quality. In contrast, the two other discussants vanished from the scene and did not contribute even when Ahmad turned to the whole group to refer to the discussion in a reflective way (Turn 38): Intermediate summary we have here two viewpoints Turn 35 and Turn 36; is it possible to bridge between them or do they contradict each other? Ahmad's attempt to instigate a common conclusion encountered Judith's strong opposition, who was happy that she could finally develop her own ideas and who thought that the opinions brought forward by Rim and Fatima are not academic enough. Judith saw in their desertion from the discussion an additional proof of their weakness and of their lack of motivation. For example, when Ahmad asked her to think about relations between her and Fatima's ideas at Turn 42, she wrote: They do not contradict, they are not connected. Ahmad's reaction at Turn 44 (We don't say ‘contradict’ or ‘connected’ we say ‘two positions’ and whether it's possible to integrate them in a composite sentence?) is opposed by Judith who asks Why is it needed?. Ahmad, who did not receive any reaction from Fatima and Rim, wrote up the summary by himself in Turn 47: My summary: we can't give up the role of the teacher so that we should continue to take the teacher into consideration in any activity linked to learning and teaching in school. And in the same breath, it is impossible to neglect the student when we study the teacher so that both the teacher and the learner are not disconnected from their learning environment in school and in family. Just after he posted this contribution, he wrote another contribution (Turn 48): Those who do not agree should offer an alternative proposal. Judith reacted I agree in Turn 49. Fatima and Rim did not react (See Fig. 4). Somehow, this episode should have been out of topic in our quest for maintenance of communication: Ahmad who decides to sum up the discussion, interacts with Judith whereas Fatima and Rim remain silent. However, as we will see, this e-silence is all but a lack of attendance; it is loaded, full of anger, inscribed there on a map that shows their presence. Let us see how this episode was interpreted by Ahmad, Judith and Fatima in their auto-confrontation.

Ahmad: a last attempt to create some shared knowledge

In his auto-confrontation, Ahmad explains his decision to sum up the discussion by himself: I stress that this is a summary that I suggest for the group, but meanwhile it is only mine. Because I am one among others…It's possible to modify it. Indeed, Ahmad sees the moderator as a partner who contributes, who expresses his opinion: [the moderator] should refer to what others say, and should also contribute…I mean content not only the form, to give his ideas in the discussion. If he has a position, he should present it […] he should

11B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

add and be a partner in the course of the discussion. This comment fully fits what Ahmad said in his interview on his professional commitment to influence: He should express his opinion as a model of reasoning (even if the opinion itself is controversial); Students have the obligation to propose alternatives and to compare themwith what is proposed. It seems here that Ahmad, who

38. Ahmad Comment no link Interim summary: we have here two viewpoints, Turn 35 and Turn 36; is it possible to bridge between them or do they contradict each other? 40. Judith Comment support to 39; after a few moments adds opposition to 37 Even if the role of the teacher is to prepare the setting, he still exists and it is impossible without him 42. Judith Claim opposition to 38 They don’t contradict, they are not connected. 41. Ahmad Claim support to 40 From what we are saying it appears that the teacher continues to take the central and important role in modern education system despite all the criticism said and written. If it's agreed, let's move on to the o ther things regarding the teacher. 43. Ahmad Argument support to 40 Agreed. 44. Ahmad Comment support to 38 and neutral link to 42 We don’t say ‘contradict’ or ‘connected’, we say ‘two positions’ and whether it’s possible to integrate them in a composite sentence? 45. Judith Question opposition to 44 Why is it needed? 46. Ahmad Argument neutral link to 45 It is needed because we were instructed to reach a joint statement 47. Ahmad Argument neutral link to 46 My summary: we can’t give up the role of the teacher so that we should continue to take the teacher into consideration in any activity linked to learning and teaching in school. And in the same breath, it is impossible to neglect the student when we study the teacher so that both the teacher and the learner are not disconnected from their learning environment in school and in family 48. Ahmad Question neutral link to 47 Is this summary acceptable? Those who do not agree should offer an alternative proposal 49. Judith Argument support to 47 I agree

Fig. 4. The end of the discussion: a common summary or the moderator's aggressive coercion?

12 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

was asked to serve as a moderator, behaves as a teacher. His message at Turn 48 in which he asks discussants whether they agree with him and if not, suggests to propose alternatives totally fits his pedagogical view as a teacher. On this message, Ahmad comments: I said that if somebody opposes, he should propose something else. He didn't think that this would actually happen but felt that discussants should know that the possibility is open: At that point, I knew that I won't get alternative positions because this was the end of the lesson […] But I wrote it because this was my duty.

At the end of his auto-confrontation, Ahmad reflected on his efforts to organize and to dissuade. Ahmad admits that he took on himself a role which is too complex: It is too complicated, I must see things at the beginning, at the end, I must look at what they refer, at what they didn't, I must express my opinion. I should say that this was sometimes an impossible mission.

Judith: a surprising episode in which the moderator promotes a high-quality discussion

Judith's reaction on Ahmad's request to summarize the discussion is first quite negative. Although Judith wants to end discussions with something agreed upon, she mentions that the two other discussants were stuck. Quite naturally, in Turn 42, she opposes Ahmad's request to include their contributions in the summary by writing they don't contradict, they are not connected since their previous contributions were superficial. Ahmad's reaction at Turn 44 (We don't say ‘contradict’ or ‘connected’, we say ‘two positions’ and whether it's possible to integrate them in a composite sentence) refers explicitly to Judith but she does not accept yet this explanation and asksWhy is it needed?. But Judith is open to Ahmad's position. She progressively understands Ahmad and she comments then: He explains…he simply wants to tie them, and he says ‘we don't have to decide among them but to connect them in one sentence’ and leads to the direction of a summary. And quite naturally, her final comment is Indeed, I agreed with him. Although she does not react to Ahmad in Digalo, his intervention deeply impinges on her appraisal of the role of the moderator as somebody who tries to put together all arguments, even small, those I thought to be not necessarily right and important […] in order to summarize.

Later on, in the discussion, Judith sees in Ahmad's summary at Turn 47 a successful combination of ideas of all discussants: Indeed, this is what I tried to say… so, this is why I agreed with him… His formulation is sensitive… He doesn't say one must study the parents too in all research [what the other discussants thought] but he says that it is necessary to study the teacher and the rest is important too (what Judith thought). This formulation is surprising for Judith after all the discussants expressed their opinion and tried to decide ‘who is right’ because, as she claimed in her interview, discussions are places for competition between ideas. With such a view, it is indeed surprising for the moderator to summarize by combining. The moderator is for her a secretary who helps the group attaining its goal. But as she reports in her auto-confrontation on what she felt during the discussion, she changes her own view (and as we will claim in the discussion section — she learns) and appreciates Ahmad's move: His formulation was nice. He didn't just say ‘everybody agrees’… I didn't think in that way (laughs). Judith comments on turn 48 by appreciating very much Ahmad's the importance of the discussants in his summary, although they didn't write the summary: This is beautiful […] He summarizes the discussion and writes ‘my summary’ because this is what he understood. He checks whether his understanding is correct. This is very important. Quite naturally she agrees with this summary at Turn 49.

After Judith commented on the moves of the discussion map with the replay function of the Digalo tool, she praised the moderator as somebody responsible who leads the group to go deep into the content: He knew to observe us from the top…to bring a question that brings a new angle. For Judith, this was a new insight that leads the moderator to combine ideas (she confessed that she did not play this role when she was a moderator, the week before). Nevertheless, she saw that Fatima and Rim did not react to Ahmad's summary. She interpreted this phenomenon as something to take care of, that originated from what she considered as their weak competences: How can we force weak students to participate…in order to gain as much as the others? One should be aware of the personal characteristics of students in order to adapt to them the goal of the activity and its formulation. I have been one of those with high motivation and obedient. If I am told to participate to a discussion, I do it…

Fatima: the moderator destroyed the discussion

We saw that Ahmad thought that Rim and Fatima were silent during the last episode of the discussion because they went out of the class, and didn't pay much attention to this fact. Judith explained that this behavior characterized weak students. In contrast, Fatima explained her silence as a demonstration. She claimed that by summarizing, Ahmad did not fulfill his role of moderator and became a regular discussant. She considered his moves in Turns 38, 42, 43, 44 and 45 as personal interventions that conveyed his opinion. For her, his summary in Turn 47 looks the same. For example, these are her comments on Ahmad's interaction with Judith: One argues with the other, between Ahmad and Judith, and at the end he writes HIS summary. Interestingly, Fatima concludes her comment on Turn 47 by expressing an opposition to Ahmad status rather than to his opinion: His summary relies on our opinions, maybe on mine, but what we say is that we should have reached this conclusion by ourselves. Therefore, the opposition, and the silence originates from what Fatima seems as a violation in Ahmad's role of moderator. The context of the discussion was, for Fatima, peer argumentation without any authoritative guidance: I transmit you what I have and you transmit what you have and maybe something different will come out of that. In her interview, Fatima was clear that if the authority conveys his opinion, the discussion is destroyed: If the teacher intervenes during discussion, students feel ashamed to express their opinion or to argue in front of the teacher. […] the teacher can tell you that your idea is not correct, […].

The appropriation of the opinions of Rim and Fatima in Ahmad's summary, which was a way for Ahmad to promote a further discussion, is understood very negatively:He took ourwords as if theywere his!Anatural consequence of this feeling in Turn 48—when Ahmad requests agreement or alternative summaries, is her refusal to resume participation: He began to sum up, he summarized what

13B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

hewanted andwhy to putmy reaction there? There's no need!…He should have shown interest inwhatwewrote. Her comment on Judith's intervention on Turn 49 (I agree) completes this very negative appraisal: This is what he wants to hear, so she gives it.

At the end of the auto-confrontation, Fatima assessed the discussion and her participation. She was not satisfied. She was surprised that she depended on the moderator and that she did not contributed enough. Fatima admitted that she had mental blocks she was not aware of before. For example, she admitted that the activeness of the moderator caused her to lose any interest in the discussion. She also admits that she should have beenmuchmore active although at the end it was really boring. She thought that she should have sent signs of anxiety: It is true, maybe, that I should have said something…to get him to pay attention to two participants and that he should have taken us back to do something. These reflections sound like a kind of confession which surprised us. This episode showed— like the two first episodes, deep divergences of interpretation. These were not disagreements but prolonged misunderstandings enabled by a communication mode that gave the opportunity to maintain them. Fatima who seemed absent, was quite loquacious about her silence in her auto-confrontation, as if this absence reflected a parallel inner lifewhich is there, invisible but palpable, rather than a desertion or even the kind of passive attendance which is so frequent in face-to-face teacher-led discussions.

For Fatima, the emptiness of the map — emptiness of her, was a sign of distress, and of pretended boredom and of anger. For Ahmad, Fatima's and Rim's absence in the map is a sign of his failure as a moderator to think that they deserted the discussion. For Judith, the final part of themapwas a sign of brightmoves ofmoderation, not taken into consideration byweak students. Therefore, the different participants in the discussion have nothing to share in order to disagree: they do not knowwhat the other believes andwhat is at stake. Nothing to share? Not exactly! The discussionmap— a persistent inscription of previous discursivemoves, is shared by all, and serves as a conduit for communication, even in the case of prolonged misunderstandings. Why communication can be maintained in this case? What is special in this synchronous mode in comparison with face-to-face modality for which misunderstandings avoid communication?

Conclusion: co-alienation mediated by a common representation as a new manifestation of maintenance of communication

The idea of intersubjectivity has been developed to explain how people maintain communication in various social practices. Matusov has discerned several kinds of intersubjectivity that can be classified as (a) overlapping of subjectivities, (b) sharing of (divergences of) meaning, (c) coordination of participant's contribution in joint activity and (d) human agency. By definition, all kinds of intersubjectivity describe diverse forms of states of subjectivity shared among several individuals. These instances of intersubjectivity were recognized as being of the highest importance for learning and for the development of productive guided participation (Rogoff, 1990). The particular setting of this research has uncovered what we consider as not being classifiable to any of these categories: the interview uncovered very different beliefs on learning, on the role of discussions for fostering learning, on the role of moderators, and on whether Digalo helps discussants and the moderator. We anticipated that these beliefs would come to the surface in the discussion. We expected that disagreement would arise and would be fueled by the common recognition of divergences. The confrontation of the discussants with their previous discussion in the cued retrospective reconstituted, step by step, with the help of the replay function of the Digalo tool, showed a different picture. Ahmad, Judith, and Fatima had divergent interpretations of their synchronous discussions, but this divergence was not shared: what was meant to facilitate construction of knowledge from the part of Ahmad was interpreted as brutal interference by Fatima;What was meant to be an exchange of ideas (by Fatima and probably by Rim) was seen as a shallow discussion that does not lead to learning by Judith and Ahmad; or Fatima's silence which was interpreted by Judith as a sign of weakness, but was forcefully justified by Fatima as a determined resistance to participate. There is even not any clear cut between the presence and the absence of discussants: Ahmad's absence in the first episode for technical reasons was interpreted as a welcome ‘presence’, a tacit agreement to give students the opportunity to talk about the issue at stake without interfering. Clearly, actions are not co-ordinated.

What is shared is extremely general: Following the same university course, shared background of knowledge of articles they were asked to read. Except for these generalities, specific beliefs about the discussion (e.g., disagreements) are not shared and the order of actions is quite whimsical. How, with all those unshared divergences and the absence of coordination of actions, communication could be maintained?

Of course, one may argue that the question is not a real one, since during the first and the third episodes there is no real communication among all discussants. However, ‘not-reacting-to-a-request’ or ‘not-participating-actively-for-a-while’ are two behaviors that are inherent to synchronous discussions. Moreover, these behaviors are communicative in the sense that they are posted on a shared object, the argumentative map, and each of the participants intends to convey a message. Even in the third episode duringwhich Fatima and Rim remained silent, they are not really out: in her auto-confrontation, Fatimamakes clear that she wanted her silence to be posted on Ahmad's face! Fatima is ‘in’, attentive to Ahmad's moves and eager to convey her anger. So, what kept the group together in this weird communication?

We saw that the nature of tools for synchronous discussion enables to communicate differently from in face-to-face communication. The discussionmap shows all previousmoves, their authors, aswell as the interlocutors to which theywere directed. Discussants use themap to reason, as well as to communicate. For example, Ahmad selected two contributions by Rim, one recent and one remote, to point at some inconsistency. And this map is always present, even for participants who are silent.

The growingmapmediates the interpretations of the discussants in awaywhich is radically different from face-to-face settings. In face-to-face interaction, interpretations are updated at any moment and adjacent interventions influence more interlocutors than non-adjacent ones (Felton & Kuhn, 2001). In contrast, at the time an actor intervenes in synchronous discussions, he and his interlocutors can see the traces of a whole history. Their interpretations at any time of the discussion are sometimes mediated by

14 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

previous misunderstandings inscribed in the map, and the discussants cannot instantaneously dissipate misunderstandings as easily as in face-to-face interactions through facial expressions or intonation. The big discrepancies in emotional states— satisfaction against anger, disappointment or boredom against interest, suggest that the map hosted unshared cohabitating misunderstandings. Each participant seemed isolated, the presence of the other being transient, as the traces inscribed represent actions that already happened, and those actions may not be backed by what the other does or does not do here and now.

This observation adds up to the analysis of the cued retrospective reporting to affirm that the e-discussion developed without the establishment of any state of intersubjectivity in a sense of coordination of contributions. Grounding which is so important for coordinating contributions in joint activity is clearly not nurtured: the typical triplets of turns in which one posts a message, the other acknowledges his/her attendance, and the first continues, are largely inexistent.

In the absence of proper grounding and of state of intersubjectivity, we considered different traditions that explain the propensity people have in participating in dialogs. According to a cognitive psychology perspective, dialogs demand a lot from discussants. However, people manage to participate in dialogs. Garrod and Pickering (2004) have proposed the idea of interactive alignment to explain this propensity: This is a process by which people align their representations at different linguistic levels at the same time. They do this by making use of each others' choices of words, sounds (in f2f communication), grammatical forms, and meanings. Interactive alignment ensures that interlocutors operate on common representations. So in speaking, each partner generates his utterance on the basis of what he has just heard from the other and can leave out redundant information without the risk of misunderstanding. Similarly in listening, aligned representations at the levels of the situation model, semantic interpretation, and syntactic formenable the listener to fill in the gaps at these levels. As interactive alignment predicts, speakers reuse the structures that they have just interpreted as listeners when formulating their response. Tomake utterances appropriate for the addressees, as a conversation proceeds, interactive alignment predicts that interlocutors build up a body of aligned representations, whichGarrod and Pickering call implicit common ground. The interactive nature of dialog supports interactive alignment of linguistic representations. In turn the alignment of representations has the effect of distributing the processing load between the interlocutors because each reuses information computed by the other.

Is the idea of interactive alignment applicable to synchronous e-discussions in general? Not exactly, but it is useful! But it provides some inspiration about for an alternative idea: Instead of aligning representations at different linguistic levels at the same time on the basis of what has just been heard, discussants interact with a growing map. This map has two contradictory characteristics. On the one hand, it changes instantly; messages arrive often at a hectic pace, sometimes simultaneously from different interlocutors. On the other hand, the map is stable; it mostly remains unchanged, with an accumulating history. The first characteristic seems to invite discussants to align their representations. However, discussants are not obliged to react to these messages. The second one leads discussants to rely on past persistent interpretations perpetuated by the map. There can then be different ways to maintain communication in synchronous discussions. It is reasonable to think that an implicit common ground fuels many synchronous discussions. But our case is different. For some discussants, the map provides an imaginary implicit common ground based on recent contributions. For others, the whole history is relevant, and every new contribution cannot be detached from past ones. Anyway, the high pace of communication gives the illusion to each discussant that he/she and his/her peers posted their beliefs about teaching, learning and moderating and interacted with them, and makes clear his/her (dis)agreements. However, what happened was a co-alienation— the juxtaposition of incompatible alignments of representations through a common external representation.

We need additional examples to understand more the phenomenon of co-alienation. In particular, it is not clear to us whether Bakhtinian ideas about dialogism can help to understand co-alienation: Of course, Bakhtin asserts that gaps between interpretations and understandings always exist, and somehow fuel dialogic meaning making. But as notices Matusov (2011), Bakhtin's gaps are positive as people continue their dialogs because of the expectation “of their mutual surprise from each other, of theirmutual wonder about each other, of their mutual interest in each other, of their mutual respect of one another's agency of decision-making – this interest and respect is not instrumental but rather goal- and value-defining – iswhat I call dialogic interadressivity” (p. 103). This is not what we feel in the discussion we analyzed. More generally, our subjective experience as users of social networks is that when we interact with various electronic tools we have at the same time an impression of closeness and togetherness but also of solitude: we think that we understand the other and communicate with her, but it is sometimes an illusion. How are these impressions related to co-alienation? This is an issue to be pursued in further research. We only showed in this paper that co-alienation is possible in synchronous discussions. But, of course, communication is not governed by co-alienation in all synchronous discussions, althoughwe conjecture that it is frequent.

At that stage, it is premature to discuss the educational relevance of co-alienation and to decide whether its emergence is a priori welcome or should be avoided. But the question whether co-alienation is utterly bad or whether it can lead to learning, seems to us a wrong question. What is certain is that it does not fit any of the four senses of intersubjectivity recognized so far. These senses were useful to document different forms of guided participation or of peer collaboration towards learning gains. Since co-alienation exists, the right question is how people can learn from such communication. And indeed we can ask whether the discussion between Ahmad, Judith, Fatima and Rim lead to any kind of learning. On the one hand, the discussion in itself seemed quite shallow, scattered with persistent misunderstandings. However, Ahmad and Judith earned invaluable insights during this lopsided discussion. True, this is their auto-confrontation that demonstrated clear gains but it is impossible to know whether this reflective activity revealed or promoted those gains. Anyway, synchronous discussions should be considered in their larger educational contexts. They rarely happen as isolated activities but rather belong to a series of activities. The cued retrospective/ auto-confrontation was arranged in the present study for experimental purposes, but it resembles common educational settings in which synchronous discussions are reflected on. Therefore, Judith's understanding that discussions do not only help deciding who is

15B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

right but help co-constructing new ideas, Ahmad's confession that his plan to facilitate without interfering was too simplistic and is a very complex endeavor, Fatima's recognition that she should have overcome her anger to be more active in her interventions… all these lessons derivate from critical moments that occurred during the discussion. We should say more: the very breakdowns that burst during the synchronous discussion provided invaluable learning opportunities. It seems then that synchronous discussions provide unique opportunities to learn what othermodes cannot provide because, among other reasons, they yield persistent displays to reflect on and because they enable actions on previous moves. The precariousness of communication in synchronous discussions, the state of co-alienation we described, can be dangerous. However, it can serve, with appropriate activities, to improve interactions in discussions among learners.

References

Asterhan, C. S. C., & Schwarz, B. B. (2009). Peer argumentation that fosters concept learning: Maintaining the balance between critical dialogue and interpersonal harmony. In B. B. Schwarz, T. Dreyfus, & R. Hershkowitz (Eds.), Transformation of knowledge through classroom interaction (pp. 159–172). Routledge: New Perspectives on Learning and Instruction.

Anderson, T., Rourke, L., Archer, W., & Garrison, R. (2001). Assessing teaching presence in computer conferencing transcripts. Journal of the Asynchronous Learning Network, 5(2).

Baker, M., Hansen, T., Joiner, R., & Traum, D. (1998). “Grounding” for intersubjectivity and learning (Rep. No. CR-15/98). Lyon, France: Centre National de la Recherche Scientifique.

Baker, M., Hansen, T., Joiner, R., & Traum, D. (1999). The role of grounding in collaborative learning tasks. In P. Dillenbourg (Ed.), Collaborative learning: Cognitive and computational approaches. : Elsevier Science Publishers.

Cazden, C. B. (1995). Play with language and meta-linguistic awareness: One dimension of language experience. In J. S. Bruner (Ed.), Play—Its role in development and evolution (pp. 603–608). New York: Penguin.

Clark, H. H. (1994). Discourse in production. In M. A. Gernsbacher (Ed.), Handbook of psycholinguistics. San Diego: Academic Press. Clark, H. H., & Schaefer, E. F. (1989). Contributing to discourse. Cognitive Science, 13, 259–294. Cobb, P., Stephan, M., McClain, K., & Gravemeijer, K. (2001). Participating in classroom mathematical practices. The Journal of the Learning Sciences, 10(1&2),

113–163. Condon, S. L., & Cech, C. G. (1996). Discourse management strategies in face-to-face and computer-mediated decision making interactions. Electronic Journal of

Communication/La revue électronique de communication, 6(3). Felton, M., & Kuhn, D. (2001). The development of argumentive discourse skills. Discourse Processes, 32, 135–153. Garrod, S., & Pickering, M. J. (2004). Why conversation is so easy. Trends in Cognitive Science, 8(1), 8–11. Herring, S. (2001). Computer-mediated discourse. In D. Schiffrin, D. Tannen, & H. Hamilton (Eds.), The handbook of discourse analysis (pp. 612–634). Oxford:

Blackwell Publishers. Howe, C. (2009). Expert support for group work in elementary science: The role of consensus. In B. B. Schwarz, T. Dreyfus, & R. Hershkowitz (Eds.), Transformation

of knowledge through classroom interaction (pp. 159–172). Routledge: New Perspectives on Learning and Instruction. Marvin, L. -E. (1995). Spoof, spam, lurk and lag: The aesthetics of text-based virtual realities. Journal of Computer-Mediated Communication, 1(2). Matusov, E. (1996). Intersubjectivity without agreement. Mind, Culture, and Activity, 3, 25–45. Matusov, E. (2001). Intersubjectivity as a way of informing teaching design for a community of learners classroom. Teacher and Teaching Education, 17, 383–402. Matusov, E. (2011). Irreconcilable differences in Vygotsky's and Bakhtin's approaches to the social and the individual: An educational perspective. Culture &

Psychology, 17(1), 99–119. Mollo, V., & Falzon, P. (2004). Auto- and allo-confrontation as tools for reflective activities. Applied Ergonomics, 35, 531–540. Murray, D. E. (1989). When the medium determines turns: Turn-taking in computer conversation. In H. Coleman (Ed.), Working with language (pp. 251–266).

New York: Mouton de Gruyter. Rogoff, B. (1990). Apprenticeship in thinking — Cognitive development in social context. Oxford University Press. Rommetveit, R. (1979). On the architecture of intersubjectivity. In R. Rommetveit, & R. M. Blaker (Eds.), Studies of language, thought, and verbal communication.

New York: Academic Press. Rommetveit, R. (1985). Language acquisition as increasing linguistic structuring of experience and symbolic behavior control. In J. V. Wertsch (Ed.), Culture

communication, and cognition: Vygotskian perspectives. Cambridge: Cambridge University Press. Rourke, L., Anderson, T., Archer, W., & Garrison, R. (1999). Assessing social presence in asynchronous computer conferencing transcripts. Journal of Distance

Education, 14(2), 50–71. Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language, 50, 696–735. Schwarz, B. B., & Asterhan, C. S. C. (2011). E-moderation of synchronous discussions in educational settings: A nascent practice. The Journal of the Learning Sciences,

20(3), 395–442. Schwarz, B. B., & De Groot, R. (2007). Argumentation in a changing world. International Journal of Computer-Supported Collaborative Learning, 2(2–3), 297–313. Schwarz, B. B., & Glassner, A. (2007). The role of floor control and of ontology in argumentative activities with discussion-based tools. International Journal of

Computer-Supported Collaborative Learning, 2(4), 449–478. Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. Toronto, ON: Wiley. Trevarthen, C. (1993). The self born in intersubjectivity: An infant communicating. In U. Neisser (Ed.), The perceived self: Ecological and interpersonal sources of

self-knowledge (pp. 121–173). New York: Cambridge University Press. Van Gog, T., Kester, L., Nievelstein, F., Giesbers, B., & Paas, F. (2009). Uncovering cognitive processes: Different techniques that can contribute to cognitive load

research and instruction. Computers in Human Behavior, 25, 325–331. Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal and hyperpersonal interaction. Communication Research, 23(1), 3–43.

16 B.B. Schwarz et al. / Learning, Culture and Social Interaction xxx (2012) xxx–xxx

http://dx.doi.org/10.1016/j.lcsi.2012.09.002

‘Co-alienation’ mediated by common representations in synchronous e-discussions

Introduction: Four senses of intersubjectivity to account for maintenance of communication
Features of synchronous discussions: New turn-taking rules, social (lack of) presence, complexity, persistence and attracti...
Description of the research: Studying maintenance of communication in synchronous discussions in a university course

Participants
Procedure

Analysis of the discussion: three incompatible but cohabitating stories by three discussants
1st episode: starting the discussion without moderator — accumulating similar contributions

Ahmad: discussants do not really refer to each other; they agree instead of thinking in depth
Judith: it is impossible to learn from this discussion
Fatima: the discussion is successful because people express themselves and the moderator is not intrusive
Interim summary

Second episode: Ahmad tries to change the flow and the quality of the discussion

Ahmad: efforts to organize the discussion, and to foster collaborative learning
Reaction to Ahmad in Ahmad's eyes: the discussants do not agree with his description of the situation
Judith: the moderator was not noticeable
Fatima: the intrusion of the moderator turns the discussion to uninteresting

The end of the discussion: a common summary or the moderator's aggressive coercion?

Ahmad: a last attempt to create some shared knowledge
Judith: a surprising episode in which the moderator promotes a high-quality discussion
Fatima: the moderator destroyed the discussion

Conclusion: co-alienation mediated by a common representation as a new manifestation of maintenance of communication
References