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

Article2_BakerBernardDumez-Froc2012.pdf

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

Integrating computer-supported collaborative learning into the classroom: the anatomy of a failurejcal_435 161..176

M. Baker,* F.-X. Bernard** & I. Dumez-Féroc*** *Centre National de la Recherche Scientifique and Telecom ParisTech, France **Paris Descartes University, France ***Poitiers University, France

Abstract We present an analysis of a longitudinal case study whose aim was to understand the processes of integration of a face-to-face and networked collaborative learning technology and pedagogy into a secondary school history-geography classroom. Students carried out a sequence of argu- mentative tasks relating to sustainable development, including argument generation, sharing and elaboration, debate using a computer-mediated communication, and organization of argu- ments in a shared diagram. Students’ interactions and diagrams were analysed in terms of degree and quality of argumentativity, as well as catachresis (‘getting round’ the software to perform a non-prescribed task). Results run counter to positive systems of ideas and values con- cerning collaborative learning and its technological mediation in that the scenario did not meet its pedagogical aims, having to be abandoned before its planned end. We discuss possible explanations for this ‘failure story’ in terms of the articulation between everyday, technology- related and educational discourse genres, with their associated social milieux, as well as the social structure of the classroom. The relevance of these aspects for future attempts to integrate such technologies is discussed. In conclusion, we discuss a vision of learning that takes into account students who do not accept to play the educational game.

Keywords adolescent culture, argumentation, computer-supported collaborative learning, digital natives, discourse genres, educational technology appropriation.

Introduction

A current tendency in educational research involves going beyond the study of learning effects of introduc- ing educational technologies, in attempting to under- stand the processes by which they are appropriated, in relation to the transformation of educational practices (e.g. Overdijk & van Diggelen 2008, Ludvigsen et al. 2011). Research on the integration of educational tech-

nologies into the classroom has a history spanning several decades (Baron & Bruillard 2007). Because computer-supported collaborative learning (‘CSCL’) technologies have only emerged relatively recently, with the increasingly widespread use of Internet in society, there have been relatively few studies of their appropriation in the classroom (cf. Chen et al. 2010).

Given that Internet-based and mobile technologies now pervade life in economically developed countries, they could be seen as highly relevant tools for education (e.g. Stahl 2005; Kemp 2011) that are adapted to the characteristics of ‘digital natives’ (Bennett et al. 2008; Crook in press). Moreover, although collaborative

Accepted: 6 May 2011 Correspondence: Michael Baker, LTCI, CNRS–Telecom ParisTech, 46 rue Barrault, 75634 Paris Cedex 13, France. Email: michael.baker@ telecom-paristech.fr

doi: 10.1111/j.1365-2729.2011.00435.x

Original article

learning (Dillenbourg et al. 1996) is not always effec- tive (Barron 2003), the very ideas of ‘collaboration’and ‘dialogue’ are associated with strongly positive societal values, as vehicles of socialization, democratization, distant teamwork, and conflict resolution. A significant body of educational research is thus concerned with identifying ‘success stories’, the conditions under which educational innovations have led to improved learning. We propose that something can also be learned about the relevant conditions for educational integration of CSCL technologies from detailed analysis of a failed attempt to do this.

We analyse a long-term attempt to integrate a particu- lar networked collaborative technology [Dialogical Reasoning Educational Webtool (‘drew’) – Corbel et al. 2003; http://drew.emse.fr] into a secondary school history-geography classroom for debates on sustainable development. This work was carried out within the European Union (EU)-funded ‘LEAD’ project,1 the overall aim of which was to develop pedagogical and software tools for small-group discussions (such as debates, group reflection or explanation) in the class- room, where students could discuss at the same time face-to-face and via the network. Students were asked to debate questions such as ‘should production of geneti- cally modified organisms be allowed in France?’, in small groups or in a whole-class-moderated discussion, to make argument diagrams and to write texts together, using shared software over a local network, while dis- cussing face-to-face.

This attempted CSCL technology integration failed in that pedagogical objective of the teaching sequence (elaborating knowledge of societal debates) was never achieved. The planned teaching sequence had to be con- tinuously modified in the light of failure of preceding phases. The study was abandoned before its planned end, on the teacher’s decision, when the classroom degenerated into chaos and threats of physical violence between some students.

In the final sections of this paper, we discuss possible explanations for this failure in terms of adolescents’ subcultures, articulation between discourse genres, and the social structure of the classroom. The relevance of these issues for CSCL technology redesign and educa- tional integration is discussed, together with implica- tions for theories of learning, once we include in the ‘sample’ students who do not accept to try to solve the problems set in schools.

Research and development background

Technology appropriation and argumentation in education

The introduction of technologies into educational situations can be seen as an ‘expansive’ process (Engeström 1999) that stimulates the elaboration of new practices rather than as a matter of adding to and rendering more efficient a pre-formed and unchanging activity. Technology appropriation is the set of pro- cesses by which an artefact becomes an instrument for subjects (Rabardel 1995), a hybrid technico-cognitive entity comprising the tangible features and affordances of the artefact with its associated usage schemas. Instrumentalization can also include ways in which users of artefacts can try to ‘get round’ creatively their uses intended by designers, termed catachresis.2 In col- laborative learning situations, appropriation of arte- facts has been described as the set of transitions among individual mastery of the tools (knowing how to use the interface), personal utilization (using the interface to perform educationally relevant tasks), and collective utilization (learning to work together on the task; Overdijk & van Diggelen 2008).

The initial aims of our work can therefore be described as attempting to favour the appropriation, by the system comprising the school, the teacher and the students, of a pedagogical approach termed computer-supported collaborative argumentation- based learning (Andriessen et al. 2003), with the peda- gogical aim of enabling students to ‘broaden and deepen’ their understanding of a ‘space of debate’ (Baker et al. 2003). There is now a substantial research literature on the role of argumentation in collaborative learning (e.g. Andriessen & Coirier 1999; Andriessen 2006; Muller Mirza & Perret-Clermont 2009). An important issue here is the necessity to find school- based topics that are debateable (Golder 1996; Baker 2009). Another is the necessity for students to argue across a variety of argumentative tasks (teacher-led debates, argumentative text and diagram writing, and small-group discussions) to favour deep and stable argumentative knowledge reconstruction (Schwarz et al. 2002). There are two main interactive processes of knowledge creation associated with argumentative interactions: justification (the production of arguments or counter-arguments) and negotiation of meaning

162 M. Baker et al.

(Golder 1996; Leitão 2000). On a conceptual plane, argumentative interactions can function as processes of negotiation of meaning of the key concepts underlying the topic debated (e.g. ‘nature’ and ‘freedom’), involv- ing new dissociations or associations between concepts (Baker 2002).

On the side of educational practice, debates are pres- ently strongly encouraged within the French national curriculum (MJENR 2002), whether at primary or at secondary school level. They are seen not only as methods for learning in specific disciplines (e.g. native language, citizenship, history-geography) but also as means for developing cross-disciplinary skills of argu- mentation and communication. It has, however, been claimed (Guerrini & Majcherczak 1999) that many teachers avoid organizing debates, principally because it is not clear exactly what should be learned from them, whether this is an efficient method of learning, and how such learning should be evaluated.

The CSCL situation

In the educational situations designed in the LEAD project (see endnote 1), secondary school students worked with CSCL software in small groups in the classroom using concurrently both the communication and the collaborative work software via the local network (partitioned for each group) and communicat- ing (verbally and non-verbally) face-to-face. The guiding principle of the project as a whole was that stu- dents would be able to coordinate their interface actions more easily using speech (that can be produced in paral- lel to interface actions), and thus focus on collaborating to perform the educational task.

Within a design–implementation–evaluation itera- tive cycle, during the first year of the LEAD project, we decided to experiment with existing software, similar in functionalities to what would be developed subse- quently, to inform ongoing software design. The soft- ware that was used in our study was called drew (Corbel et al. 2003; see http://drew.emse.fr); it was designed to support collaborative argumentation-based activities across Intranet or Internet.

Figure 1 shows one of the configurations of the shared workspace of drew, with a chat in the top-left corner, a collaborative writing tool in the top right, a whiteboard bottom left, and an argumentation diagram tool in the bottom right.

The argumentation diagram tool (bottom right in Fig 1) comprises boxes for arguments/theses and only two types of argumentative links (‘+’ and ‘-’, or else undefined: ‘?’). Different colours are automatically assigned to each participant (up to 12). Opinions (for or against) can be expressed with respect to each element of the diagram. As a means of focusing students’ atten- tion on conflicts of opinion, when for and against opin- ions are expressed by different users with respect to a given argument box, it becomes ‘scrunched’. The ratio- nale of these features was to provide a simple tool in terms of types of links and boxes, which would thus enable students to focus on the domain of discourse, rather than the representational format (Suthers et al. 1997), and to reify the difficult argument/opinion distinction.

The Internet version of drew was reprogrammed to function in a local network, with individual chat rooms for each subgroup in the class (see Fig 2). Once the main application is launched, students input their names and enter chat rooms corresponding to their groups, as defined by the teacher.

The teacher cannot control the information input by the students; the software does not prevent changing from one chat room to another and performing inter- face actions within it. drew allows the researcher to recuperate an automatic trace of all interface actions (including local server time of actions).

Method

The appropriation of tools is a long-term process (months or years), especially when their integration into the classroom also requires the elaboration of new learning and teaching practices. We thus carried out a longitudinal field study, in close collaboration with teachers and students. As with pedagogical design experiments (Brown 1992), our approach was interven- tionist (which distinguishes it partly from naturalistic observation) in that researchers collaborated with the teacher (and the students) in designing and in putting into practice the situation, within an iterative process (design–implement–evaluate) in the classroom itself.

Educational context

The study was carried out over a period of 6 months (from September 2006 to March 2007) in a ‘general,

CSCL in the classroom: a failure story 163

technological, and professional’ secondary school (here named ‘Lycée Renaud’) on the southern perimeter of Paris, which mostly prepares students for a profession- ally oriented baccalauréat. History-geography was chosen as the subject matter in which to carry out the study, given that the organization of debates is pre- scribed in the curriculum. In this baccalauréat, history- geography is an obligatory yet ancillary subject (the marks for it count for little in the baccalauréat examina- tions). Students receive 4 h of history-geography teach- ing per week.

In sociological terms, the school had a student popu- lation that was ‘socio-economically mixed’ (Pasquier 2005), comprising a majority of students from families who were recent immigrants and socio-economically disadvantaged, with some students (usually boys) who

were from more well off families but who had engaged in educational training that was lower in societal status than that of their parents (e.g. the son of an engineer and a doctor who trains to be a technician).

Participants

The study was carried out in collaboration with a (female) teacher of history-geography who had fol- lowed a special training course on the use of media in education. The teacher was experienced and regularly took charge of trainees for teaching practice in her class- room. The classes took place in a ‘language laboratory’, equipped with 20 computers, as well as a teacher- computer at the front of the class, installed into desks around the room.

Fig 1 Screen dump of the main shared workspace interface of Dialogical Reasoning Educational Webtool.

164 M. Baker et al.

A class of 24 students took part in the study, 21 boys and 3 girls, aged 16–17 years old, with the exception of 3 boys who were 18 years old. These three students were older because they had been obliged to repeat the previ- ous school year, having failed its end-of-year exams. In some cases (described below), the class was divided into two half-classes of 12 students. Students worked in ‘friendship’ groups of two, three, or four (decided by themselves), depending on the task, using adjacent computers and tables, organized in an ‘L’ so that they could see and talk with each other while communicating and working across the partitioned local network. Written permission was obtained from the parents of all students to participate in the study, and for the data to be collected and used exclusively for research purposes.

Procedure

Pedagogical design The study began with two meetings with the teacher, during which we explained the objectives of the project, together with the types of activities that we wanted to carry out and the software to be used. The teacher pro- posed sustainable development as a suitable subject matter for debates, given that she would be teaching this topic within the timing of our study.

We assisted as observers in two ‘regular’ lessons (i.e. with the teacher standing at the front of the class using

the blackboard and the students sitting in rows), each of 1 h duration. After presenting the EU project to the class, we asked for (and received) the students’ partici- pation then placed ourselves (two researchers) at the back of the classroom to observe, without further inter- vention on our part. During the first lesson, the teacher asked the students to read the textbook on the topic of dealing with water shortages in the world then to write, in pairs, a summary text. During the second lesson, the teacher returned their texts and commented on them to the class. During both lessons, we observed many ‘control interventions’ from the teacher, of the type ‘be quiet please!’, ‘listen to me!’, ‘put your mobile phone back into your bag and switch it off’. The students gen- erally disputed the teacher’s interventions. As well as generally chatting among themselves, several of the stu- dents were overtly using their mobile phones or else lis- tening to mp3 recording devices with headphones (which is forbidden by the school rules).

In a subsequent meeting with the teacher, she stated that she thought that because this particular class was very difficult for her to control in a regular teaching, perhaps working in groups with computers would be motivating for the students, and that it would be easier to control the class by supervision of group work.

It was decided to try to reuse and adapt a canonical teaching sequence involving the use of drew, which had already proved successful in French classrooms at

Fig 2 The startup window of Dialogical Reasoning Educational Webtool.

CSCL in the classroom: a failure story 165

the same level (Baker et al. 2003; Lund et al. 2007), as follows:

• Training. Familiarization with the software and with what it means to argue.

• Preparation for debate. Studying teaching materials (texts, videos, and websites) on specific debates to identify the main social actors and (counter-) arguments.

• Debate. In small groups, the students debate the ques- tion using the chat (2.1) and organize arguments using the argumentation-diagram tools (2.2).3

• Consolidation. Students co-write a synthesis of the debate.

The pedagogical aims of this teaching sequence are that the students acquire and activate potential arguments, that they confront and refine them during interaction, and that they summarize and personally internalize what they have learned from the debate, with respect to their own prior ideas.

The teaching sequence actually implemented In practice, we had to modify continually the planned pedagogical sequence, in collaboration with the teacher, in the light of the students’ work. For example, because the students’ small-group chat debates were very poor in arguments (see results below), it was not feasible to ask them to organize subsequently those sparse argu- ments in a diagram: a moderated whole-class debate was tried instead. The actual teaching sequence that was implemented is shown in Table 1.

Overall, the implemented teaching sequence com- prised 11 h and 5 m of time spent with the students and their teacher in the classroom. Details of the phases of the implemented teaching sequence are as follows:

• Phase 1 (training). Students were introduced to the drew software, focusing mainly on the argumenta- tion diagram tool (the students had no problems at all with the chat and collaborative writing tools). They were guided in making a diagram corresponding to an

Table 1. The actual teaching sequence implemented.

Phase Title Subclass1 Students (N) Groups2 Duration (min) Activities

Phase 1 Training 2 11 1 ¥ 3 4 ¥ 2

60 Familiarization with DREW software; guided creation of argument diagram

1 11 1 ¥ 3 4 ¥ 2

Phase 2 Small-group CHAT

debates 2 9 3 ¥ 3 60 Topic of GMOs previously taught in

class. Debate in subgroups 1 13 3 ¥ 3

1 ¥ 4 60

Phase 3 Whole subclass- moderated CHAT debates

1 13 1 ¥ 13 60 Whole class connected to same CHAT room, teacher moderation, debate on GMOs

2 11 1 ¥ 11 60 Phase 4 Retraining 2 11 Individually 90 Students given rules on how to

debate and fictional debate to represent in argumentation diagram

1 13 Individually 90 Phase 5 Small-group

argumentation diagrams

1 and 2 24 (2 students on each computer)

6 ¥ 4 90 (abandoned after 35 min)

Debate on nuclear energy in France, using argument diagram tool, with supporting documents and videos

1In most sessions, the class of 24 students was divided into two subclasses: subclass 1 contained 13 students, 1 female, 12 male; subclass 2 contained 11 students, 2 female and 9 male. During certain sessions, there could be fewer students present in each (sub)class due to absences (e.g. illness and absenteeism). 2‘1 ¥ 3; 4 ¥ 2’ means 1 group of 3 students, and 4 groups of 2 students (etc.). GMO, genetically modified organisms.

166 M. Baker et al.

argumentative text. The teacher prepared the students for debating the topic of genetically modified organ- isms (GMOs) by regular classroom teaching during two previous lessons.

• Phase 2 (small-group CHAT debates). Students, in small groups, were asked to debate the question (e.g. ‘Should France’s nuclear energy policy be pursued?’) via the chat of drew. They were instructed to each present their points of view, with arguments for (possibly) against it, to listen and think about what the others said, and to argue similarly with respect to others’ points of view. They were asked to try to come to a common statement, on what they (dis) agreed on, towards the end of the debate (during the last 10 min).

• Phase 3 (whole subclass-moderated CHAT debates). Students were presented with an example of a chat debate on the same topic that had been produced in another school (same level and type of class), which corresponded more closely to the pedagogical goals. The group debate was carried out with students all seated in a circle at their individual computers. The debates were moderated by the teacher.

• Phase 4 (retraining). Because 3 months had elapsed since the first training on argument diagrams (phase 1), the students were retrained with the argument diagram tool (making a diagram, in groups, corre- sponding to a short written debate on the pros and cons of forbidding smoking in public places).

• Phase 5 (small-group argumentation diagrams). The teacher had previously prepared the students in the classroom teaching on the topic of nuclear energy in France. During a first lesson of 2-h duration, the stu- dents read additional material on the debate concern- ing nuclear energy and watched two videos on the Web, one arguing in favour of nuclear energy and the other against. Working individually, they were asked to identify the main arguments for and against. Stu- dents were then asked to create an argument diagram in small groups to represent as fully as possible the space of arguments for and against nuclear energy. We had planned that this activity would be a preparation for co-writing a textual summary of the debate during a subsequent lesson (abandoned).

Data and analysis method The data analysed comprises principally trace logs of chat interactions and shared argumentation diagram

creation. Given the classroom situation, with often several groups discussing at the same time in the same room, it was not possible for us to video or to audio record spoken interaction. Our results and discussion also draw on participant observation in the classroom and planning/review meetings with the teacher.

As stated in the introduction, our initial research aim was to create a situation whereby the educational tech- nology and its attendant pedagogy would be appropri- ated and to focus analysis on this aspect. Throughout the study, it became apparent that appropriation of the tech- nology – at least for pedagogical ends – was not the dominant activity in which the students were engaged. We therefore focused analysis primarily on the extent to which the students pursued the pedagogical goals, i.e. engaging in argumentation on the topic. Secondarily, we describe one aspect of appropriation, concerning ways in which the students ‘got round’ the software (‘catachresis’: see above) for ends that were other than educational.

In pedagogical terms, we analysed the chat debates of all seven subgroups using the ‘rainbow method’ (Baker et al. 2007) that was specifically designed for understanding computer-mediated pedagogically ori- ented debates. This method distinguishes functional categories of communicative interventions, concerned with the debating task (expressing opinion, argument; negotiation of meaning of what is being debated) from those that relate to interaction management. For our purposes here, we only used the following five ‘rainbow’ categories:

• category 123: off-task, social talk, and interaction management

• category 4: task management (topic shifts of the debate)

• category 5: expression of opinions concerning the debate

• category 6: arguments for or against a thesis • category 7: broadening and deepening the debate

(discussing the meaning of related concepts, producing arguments on arguments, and discussing the links between arguments).

All data were analysed by the three authors, who resolved any disagreements. The students’ interactions were so little elaborated in terms of argumentation

CSCL in the classroom: a failure story 167

that there was little space for disagreement between analysts.

Students’ argumentation diagrams were analysed with respect to their quality, in terms of the number of arguments, their degree of elaboration (number of propositions), number of links [together with their (in)correctness] and depth of the diagram (mean number of sublinks and arguments moving outwards from the main thesis).

Results

Results focus on the students’ tasks of phases 2, 3, and 5 of the teaching sequence (Table 1), as phases 1 and 4 involved (successful) training.

Phase 2 (small-group chat debates)

This phase comprised 120 min of time in the class- room with the students and their teacher. The general characteristics of all seven chat debates are shown in Figs 3 and 4, for each half-class, as percentages of the interactions in terms rainbow categories (Baker et al. 2007).

The students’ interactions were dominated by non- argumentative categories of interaction (category 123: 64%–79% on average). This is general socio-relational or interaction management talk, not directly related to the topic being debated. The second most frequent com- munication type was expression of opinions (category

5) concerning the topic debated (12%–19% of commu- nication, on average). Argumentation itself (the sum of categories 6 and 7) only constituted 5%–8% of the stu- dents’ chats. In sum, the students’ interactions resembled more informal ‘chats’ and exchanges of opinion rather than genuine argumentative debates centred on the topic.

In previous work (Baker et al. 2003), we had studied the same chat debate (on GMOs) in a different school, with students of the same age. In that case, the percent- ages of different interaction categories were as follows: categories 123 (off-task, social talk, and interaction management), 41%; task management, 9%; opinion, 17%; argumentation, 19%; and broaden and deepen, 14%. In that previous study, the ‘123’ category was approximately half of that found above, and the argu- mentation categories comprise approximately double those observed in the present study.

Table 2 shows an extract from near the begin- ning of the chat debate of one group in the present study.

Of the 22 messages in the extract shown in Table 2, only three mention the topic of debate, and these messages are limited to expression of opinion rather than argument. The principal function of this exchange of messages can only be described as maintaining or as establishing the social relations in the group, including playful (?) insults. The messages were written princi- pally in ‘short message service (SMS) language’ (SMSese).

Fig 3 Categories of interaction in CHAT

debate (subclass 1).

168 M. Baker et al.

From the teacher’s point of view, the students’ debates were also very poor in quality of expression and in terms of their coverage of the essential points; she did not consider their debates to be a suitable starting point for analysis in the form of diagrams and as a basis on which to deepen their understanding, as initially planned.

And yet low frequencies of arguments do not necessarily mean that the arguments themselves that were in fact expressed were never relevant. In other chat-groups’ interactions, there did appear to be argumentative potential in the students’ messages, albeit in a slang and SMS abbreviated form, for example:

31 <Yorick> (. . .) they’ve done tests in maize fields 32 <Yorick> in the phillipines without authorisation 33 <tmoy> and? 34 <tmoy> wotta th results? 35 <Yorick> n’if ts not dangerous why did they do it

completely illegally 36 <tmoy> maybe they did it illegally but that dsnt

mean thts bad

Such an extract reveals the students’ competences in argumentation and reasoning that could be ‘shaped’ to more elaborate forms.

Fig 4 Categories of interaction in CHAT

debate (subclass 2).

Table 2. Extract from initial small-group CHAT debate.1

Line Student Message

16 <Yiamine> what the boloss 17 <oblanchard> clear off 18 <oblanchard> syg 19 <Yiamine> yer missed me 20 <oblanchard> fat . 21 <vchambreuil> m tellin yer 22 <vchambreuil> clear off 23 <Yiamine> julie tells you to clear off 24 <oblanchard> cool your life 25 <Yiamine> yeh yvseen 26 <vchambreuil> ts serious yacine. 27 <oblanchard> well 28 <vchambreuil> yes 29 <oblanchard> let’s talk about GMO.^^ 30 <vchambreuil> yeah f yer lyke ^^ 31 <obertrand> vincent for or against GMO 32 <oblanchard> ? 33 <vchambreuil> me rther for 34 <vchambreuil> yeah but gotta debate 35 <oblanchard> and you julie? 36 <Yiamine> dnt like GMO

1The students’ names have been changed. The extract has been translated while trying to transliterate students’ slang, spelling errors and abbreviations, preserving roughly equivalent levels of language. ‘Syg’ is a transliteration of ‘tg’ (‘ta gueule’ = ‘shut your gob’). ‘Boloss’ is a slang term of abuse. ‘Cool your life’ was written in English in the original.

CSCL in the classroom: a failure story 169

Phase 3 (whole subclass-moderated chat debates)

This phase comprised 120 min of time in the classroom with the students and their teacher. Figure 5 shows rainbow interaction categories for both half-classes in the teacher-moderated debate.

In each debate, approximately three students largely dominated the interaction in terms of numbers of mes- sages (14–18 messages). Some students did not inter- vene at all.

From Fig 5, it can be seen that in each half-class, cat- egory 123 (off-task, social talk, and interaction manage- ment) largely dominated the debate, as with the small- group debates. However, in the present case, this is largely explained by the fact that 12 and 9 participants had to log on to the chat, and thus make greetings that were replied to, as well as to log off and close the inter- action. Therefore, contrary to the small-group debates, in the moderated class debates, there were almost no ‘123’ (off-task) interventions once the debate had actu- ally begun.

It is clear, therefore, that these teacher-moderated debates were much more argumentative than the small- group debates. In the latter, ‘123’ interventions could comprise up to 90% of the interaction (Fig 3). In the half-class-moderated debates, this was generally almost halved, and, correspondingly, the argumentative cat- egories were approximately doubled. However, the stu- dents’ arguments were, again, very little elaborated: in sum, they were very short (in accordance with the first

‘s’ of ‘SMS’, ‘short message service’) and were largely still written in slang.

Table 3 shows an extract from one of the moderated group debates.

This teacher-led debate sequence seemed to keep the students on the topic, while not always changing the slang abbreviated form in which they wrote. In each sub-debate, the teacher made very few interventions, mostly asking the students to start debating and asking probing questions. It seems that teacher moderation in a whole-class discussion made the students remain at least on the topic to be debated, although it did not lead the students to really elaborate their arguments nor to frame them in school-preferred language.

Fig 5 Rainbow categories in students’ CHAT messages for moderated group debate.

Table 3. Extract from half-class one-teacher-moderated debate.

Line Interlocutor Message

107 (teacher) Should one forbid the production of gmos?

108 (vcharbonnier) i think not 109 (atruong) Pernhaps 110 (oboissy) No 111 (gfelix) shunt necessarily stop em bt

studi em 112 (teacher) Why not? 113 (vcharbonnier) cos otherwise there wont be enuff

food for people in the world* 114 (srabhi) i also think nt 115 (tmoy) at present i think not because we

dont really know effects of gmos

170 M. Baker et al.

Phase 5 (small-group argumentation diagrams)

This phase was planned to comprise 180 min of time in the classroom with the students and their teacher. Because of interruption of the study (see below), it finally only lasted 125 min of classroom time. During this phase, students were supposed to work in six groups of four. However, log-file analysis showed that in each chat room, there were between 7 and 12 logged-on participants (10 on average). There were seven ‘nomadic’ participants who each logged on to at least four of the six chat rooms, with either humoristic pseudonyms (e.g. ‘the click in the shadows’), random characters (e.g. ‘??????????’), or else recognizable ini- tials and names. In each chat room, only one or two stu- dents out of the four actually created the argument diagram. In two groups’ chat rooms, the diagram was deleted just before the end of the session, in whole or in part.

Figure 6 shows the argument diagrams created in the chat room ‘hall’, before (above) and after deletion (below).

The students’ argument diagrams contained few arguments (3–8 in number, five on average, including the statement of the thesis); they were generally very shallow (only arguments directly connected to the thesis) and were little elaborated (usually a single short proposition). The degree of correctness of argumenta- tive links (direction and polarity) was very variable from completely correct (four out of four) to not correct at all.

This argument diagram session was interrupted and abandoned before its planned end. One group of four students – the ‘hall’ group whose whole diagram had been deleted by intruders – became angry and shouted at other students on the opposite side of the room (‘It’s him, it’s him! Get out, get out!’). The situa- tion degenerated into threat of a fight. The teacher decided to stop the session, to take control over the class, and to abandon the planned subsequent textual synthesis phase: in several cases, there was no diagram produced that could have served as a basis for the text.

Analysis of the students’ diagrams from the log files enabled us to identify the intruders who had put insults in argument boxes of other groups then deleted their diagrams, as the three older boys of the class (see ‘par- ticipants’ above).

Discussion

One possible explanation for the failed outcome of this study would be that it was not specifically related to the precise characteristics of the particular CSCL technol- ogy and its attendant pedagogical scenario. This was generally an unruly class that the teacher had difficulty controlling; any change at all from the ‘talk-and-chalk’ mode might have led to disruptive behaviour. Our results do suggest, in fact, that the degree of teacher control over the class was an important factor here (e.g. the teacher-moderated class debates were much more argumentative than the autonomous small-group debates). Another possible line of explanation would be in terms of the fact that history-geography counted little towards the students’ overall mark in the baccalauréat. However, other results and considerations lend cre- dence to roles of specific characteristics of collabora- tion and social interaction using computer-mediated communication as explanatory factors, notably relating to multiple discourse genres, adolescent subcultures, and the social structure of the class.

Learning in schools can be seen as a process of appropriating an educational discourse genre (Bakhtine 1929/1977; Wertsch 1991) and being able to flexibly articulate this with others, according to the social situation. A first distinction can be made between school genres and everyday genres. The school genre in general is associated with the ability to write and speak standardized language in a way that is considered by the institution to be appropriate to the communicative situation. It also includes general appropriation of explicit and implicit classroom rules (Brousseau 1998), and more general cultural rules (e.g. it is against the law in France for any discourse inciting racism to be produced in any state arena, which includes schools). More specifically, in this study, stu- dents were expected to appropriate general ‘ground rules’ of communication in groups (Mercer 1995), as well as the argumentative genre of epistemic debate (Barth & Krabbe 1982).

There are of course multiple ‘everyday genres’ relat- ing to the diversity of social spheres.4 A relevant sphere of activity in this case is French adolescents’ subculture (Pasquier 2005), an integral part of which comprises information and communications technology (ICT)- related practices of ‘digital natives’(Bennett et al. 2008, Crook in press). For the students in the school where our

CSCL in the classroom: a failure story 171

study took place, possessing and being able to use all the latest technological devices is just as important to group identity as hairstyles, clothing, and musical taste (Pas- quier ibid.). The students’ chat interactions reveal

specific aspects of their subcultures. For example, one student’s greeting was ‘Wesh le poto!’ (roughly, ‘Yeah matey!’). This utterance refers, in France, to ‘wesh- wesh’ culture (Gonçalves 2010), a strong aspect of

Denfense consolidated

It brings money

Increase unemplayment

Competitive cost

No CO2 given off

Nuclear energy solution for the future

No CO2 given off Abundant source

of energy

! +!

Fig 6 Argument diagrams of chat room ‘Hall’ before deletion (above) and after deletion (below).

172 M. Baker et al.

group identity, inspired by ‘gangsta rap’ ways of speaking and dressing, and principally associated with tough outskirts neighbourhoods.

Discourse genres, cultures social-communicative situations can interrelate in several different ways. Stu- dents may have different degrees of mastery of a given genre (and thus one may dominate over another), and they may see genres as more or less appropriate to the situation. One possibility would thus be that the stu- dents had poor skills in standardized French and in argu- mentation, and so wrote in their everyday genres. There are reasons that weigh against such an explanation. First, recent research on mobiles and literacy (Thurlow 2003; Kemp 2011) shows that, despite cultural differ- ences, there is presently no compelling evidence that SMSese fluency is necessarily associated with poor standardized literacy. Second, during meetings with the teacher, we examined with the students’ (hand)written essays from regular teaching; she stated that their level in standard French was average for students of their age. Third, with respect to the argumentative genre, we saw that students, in moderated debates, did possess the communicative–argumentative skills of people their age (see Stein & Bernas 1999).

We therefore argue that a more plausible explanation for the students’ discourse, as produced during our study, would be in terms of a strong coupling between communication technology, adolescent culture, and the SMSese genre. In other words, because the students use SMSese via chat and mobile devices in their everyday lives, and they were asked to use chat in the classroom, they therefore saw it as appropriate in that situation (cf. Pilkington & Walker 2003) and even nec- essary to preserve their group identities and interper- sonal relations.

This explanation is also plausible with respect to wesh-wesh culture. Clearly, collaborative learning, carried out without teacher intervention, requires the students to interact with and relate to each other: for the students, relating to each other is done (more or less) in the wesh-wesh mode. When students debated with teacher moderation, they were not primarily relating to each other (this was mediated by the teacher), and so there was some degree of change in their discourse. This suggests that for these students, the educational genre is for interacting with the teacher, and not for interacting between themselves, even when their interactions occur in school.

Finally, we need to explain why the students did or could not simultaneously manage their everyday genres with the educational genres. We propose two possible explanations, in terms of difficulties in relating to each other while relating to the educational task.

The first refers to the ‘distance’ between everyday and educational genres, depending on the social milieux of students. With students for whom in- and out-of-school genres are very different, as is the case with those who have appropriated wesh-wesh culture, the problem of discursive articulation is more acute. It would seem that in this case, the students in our study did not accomplish this discursive balancing act: they fell down on the side of preserving their interpersonal relations and everyday ways of speaking to each other.

The second (related) reason why students might not have been able to mix their everyday discourse with the school-prescribed one in their autonomous interactions relates to the general social structure of the classroom, which can also be invoked in explaining the final degeneration into violence during the last phase of our study. The students who had created extra logons and crept into other groups’ chat rooms, to proffer insults and delete other students’ work, belonged to the group of three older boys in the class (see the ‘participants’ section above). These three boys had stated that they could not debate because they were ‘bad students! [laughs]’. From participant observation and review meetings with the teacher, it emerged that these three boys were the ‘bosses’ of the class (or rather, one ‘boss’ with his two ‘assistants’). Whenever other students spoke in the class, in a way that complied with the teacher, the ‘boss’ students silenced them with threat, mockery, or sarcasm. It is therefore possible that many students did not dare to speak the educational genre to each other because of the structure of power and domi- nance in the group class (Glaser 1996). In creating small autonomous groups of students, this might have allowed some students to escape the power of the bosses, which could also relate to the social identities of the boss-students as ‘bad students’: either by con- firming those identities (in the case where other stu- dents were allowed to succeed in the educational task) or else by threatening them, in the case where the bosses found themselves (‘despite themselves’) succeeding in the task (how could ‘bad’ students succeed?).

CSCL in the classroom: a failure story 173

Concluding remarks

While this analysis of a single longitudinal case study authorizes no generalizations, it could neverthe- less limit generalization of experimental results on (computer-supported) collaborative learning, given that (sub-)cultural differences become relevant factors and also indicate aspects to be considered in further work on integration of CSCL technologies in schools. Close attention may need to be paid to the students’ everyday milieux, discourses, communicative practices, and sub- cultures, as well as the social structure of the class, precisely because these will be at stake in computer- mediated social interaction. Depending on these charac- teristics, groups may need to be created and monitored to neutralize power structures; and a longer period of practice, with explicit teaching of different genres, may be required to lead students to decouple their digital native practices and subcultures from ICT so that the technology may be used for educational ends. In some situations, it may be necessary to engage in very long- term work indeed, on an institutional level, to create the premises for an appropriate pedagogical culture of (computer-mediated) debate.

Within the LEAD project, this work contributed to software redesign and the implementation of the Col- laborative Face-to-Face Educational Environment soft- ware (http://www.coffee-soft.org). Additional aspects relating to teacher control of groups in the classroom were added, to prevent, if the teacher wished, additional logons beyond a predefined list, to ‘freeze’all screens so that the teacher could address the class and to dynami- cally reconfigure groups during sessions. Providing such possibilities for tight teacher control does not oblige teachers to use them; it is possible that teachers could gradually relax control, as the pedagogical approach is appropriated.

Finally, this study raises questions about theories of learning in schools. It goes without saying in much edu- cational psychology research that students who do not genuinely attempt to solve the problems set to them will be ‘excluded from the sample’, given the aim of study- ing effective problem-solving and learning processes. What would the theory of learning in schools be like, which also takes into account students who do not accept to ‘play the educational game’? Such a theory might foreground social relations and identities, social milieux, and discourse genres, seeing knowledge elabo-

ration as a social, cognitive, and discursive practice to be appropriated and articulated with others.

Acknowledgements

This research was partly financed by the EU Informa- tion Society Technologies (IST) programme (LEAD project, number IST-028027). We would like to thank the director of Lycée Renaud (Paris), the teacher and her students who participated in the study, and the members of the LEAD project. We thank Éléonore Baker for translations into standard French of some of the students’ messages. This paper owes much to dis- cussions with Charles Crook. We thank the two anony- mous reviewers of this paper for their insightful remarks, the taking into account of which has enabled us to improve the organization and the argumentation of this paper.

Notes

1The LEAD project (technology-enhanced learning and problem-solving dis- cussions: networked learning environments in the classroom; http:// www.lead2learning.org) was funded from 2005 to 2008 under the EU IST programme, 6th Framework Programme (project number IST-028027). 2The word ‘catachresis’means ‘the use of a word in a way that is not correct, for example, the use of mitigate for militate’ (Oxford English Dictionary). Within Rabardel’s (op. cit.) theory, the notion of ‘incorrect use’ is extended from lan- guage use to artefactually mediated action in situations where it is possible to define prescribed tasks. 3It had shown that students who debated with the chat then attempted to repre- sent their own chat debate using the argumentation diagram tool had superior learning effects to students who used both tools concurrently to debate (Lund et al. 2007). 4Bakhtin (1986, p. 6) writes that ‘[t]he wealth and diversity of speech genres are boundless because the various possibilities of human activity are inexhaustible, and because each sphere of activity contains an entire repertoire of speech genres that differentiate and grow as the particular sphere develops and becomes more complex’.

References

Andriessen J. (2006) Arguing to learn. In The Cambridge Handbook of the Learning Sciences (ed. R.K. Sawyer), pp. 443–460. Cambridge University Press, NewYork.

Andriessen J. & Coirier P. (1999) Foundations of Argumenta- tive Text Processing. University of Amsterdam Press, Amsterdam.

Andriessen J. Baker M. & Suthers D. (2003) Arguing to Learn: Confronting Cognitions in Computer-Supported Collaborative Learning Environments. Kluwer, Dordrecht.

Baker M.J. (2002) Argumentative interactions, discursive operations and learning to model in science. In The Role of

174 M. Baker et al.

Communication in Learning to Model (eds P. Brna, M. Baker, K. Stenning & A. Tiberghien), pp. 303–324. Lawrence Erlbaum, Mahwah, NJ.

Baker M.J. (2009) Argumentative interactions and the social construction of knowledge. In Argumentation and Educa- tion: Theoretical Foundations and Practices (eds N. Muller Mirza & A.-N. Perret-Clermont), pp. 127–144. Springer, Dordrecht.

Baker M., Andriessen J., Lund K., vanAmelsvoort M. & Quig- nard M. (2007) Rainbow: a framework for analysing computer-mediated pedagogical debates. International Journal of Computer Supported Collaborative Learning 2, 315–357.

Baker M.J., Quignard M., Lund K. & Séjourné A. (2003) Computer-supported collaborative learning in the space of debate. In Designing for Change in Networked Learning Environments: Proceedings of the International Confer- ence on Computer Support for Collaborative Learning 2003 (eds B. Wasson, S. Ludvigsen & U. Hoppe), pp. 11–20. Kluwer Academic, Dordrecht.

Bakhtine M. (1929/1977) Le Marxisme et la Philosophie du Langage [Marxism and the Philosophy of Language]. Éditions du Minuit, Paris.

Bakhtin M.M. (1986) The problem of speech genres. In Speech Genres and Other Late Essays (eds C. Emerson & M. Holquist), pp. 60–102. University of Texas Press, Austin, TX.

Barron B. (2003) When smart groups fail. The Journal of the Learning Sciences 12, 307–359.

Baron G.-L. & Bruillard E. (2007) ICT, educational technol- ogy and educational instruments. Will what has worked work again elsewhere in the future? Education and Infor- mation Technologies 12, 71–81.

Barth E.M. & Krabbe E.C.W. (1982) From Axiom to Dia- logue: A Philosophical Study of Logics and Argumentation. Walter de Gruyter, Berlin.

Bennett S., Maton K. & Kervin L. (2008) The ‘digital natives’ debate: a critical review of the evidence. British Journal of Educational Technology 39, 775–786.

Brousseau G. (1998) Théorie des situations didactiques [Theory of Didactical Situations]. La Pensée Sauvage, Grenoble.

Brown A. (1992) Design experiments: theoretical and meth- odological challenges in creating complex interventions in classroom settings. The Journal of the Learning Sciences 2, 141–178.

Chen W., Looi C. & Tan S. (2010) What do students do in a F2F CSCL classroom? The optimization of multiple communi- cations modes. Computers in Education 55, 1159–1170.

Corbel A., Jaillon P., Serpaggi X., Baker M., Quignard M., Lund K. & Séjourné A. (2003) DREW: Un outil Internet

pour créer des situations d’apprentissage coopérant. [DREW: an internet tool for creating cooperative learning situations]. In EIAH 2003 (eds C. Desmoulins, P. Marquet & D. Bouhineau), pp. 109–113. INRP Presses, Paris.

Crook C. (in press) The digital native in context: tensions associated with importing Web 2.0 practices into the school setting. Oxford Review of Education doi: 10.1080/ 03054985.2011.577946.

Dillenbourg P., Baker M.J., Blaye A. & O’Malley C. (1996) The evolution of research on collaborative learning. In Learning in Humans and Machines: Towards An Interdisci- plinary Learning Science (eds P. Reimann & H. Spada), pp. 189–211. Pergamon, Oxford.

Engeström Y. (1999) Activity theory and individual and social transformation. In Perspectives on Activity Theory (eds Y. Engeström, R. Miettinen & R. Punamäki), pp. 19–38. Cambridge University Press, Cambridge, MA.

Glaser H.F. (1996) Structure and struggle in egalitarian groups. Dimensions of power relations. Small Group Research 27, 551–571.

Golder C. (1996) Le développement des discours argumenta- tifs [The Development of Argumentative Discourses]. Delachaux et Niestlé, Lausanne.

Gonçalves C. (2010) Le langage wesh-wesh, une forme de résistance ? [Wesh-wesh language: a form of resistance?]. Amnis, 9. Put online, 1 May 2010. Available at: http:// amnis.revues.org/277 (last accessed 13 February 2010).

Guerrini J.-C. & Majcherczak E. (1999) L’argumentation au pluriel: polyphonie, valeurs, points de vue [Argumentation in the Plural: Polyphony, Values and Points of View]. Presses Universitaires de Lyon, Lyon.

Kemp N., ed. (2011) Special section on mobiles and literacy. Journal of Computer Assisted Learning 27, 1– 75.

Leitão S. (2000) The potential of argument in knowledge building. Human Development 43, 332–360.

Ludvigsen S., Lund A., Rasmussen I. & Säljö R. (2011) Learning Across Sites: New Tools, Infrastructures and Practices. Routledge, London.

Lund K., Molinari G., Séjourné A. & Baker M. (2007) How do argumentation diagrams compare when student pairs use them as a means for debate or as a way for representing their debate? International Journal of Computer Supported Collaborative Learning 2, 273–295.

Mercer N. (1995) The Guided Construction of Knowledge: Talk Amongst Teachers and Learners. Multilingual Matters, Clevedon.

MJENR (2002) Ministère de la jeunesse, de l’éducation nationale et de la recherche [Ministry of youth, national education and research publications] Education civique,

CSCL in the classroom: a failure story 175

juridique et sociale en classe de seconde générale et tech- nologique. Bulletin Officiel Hors série n°6 du 29 août. MJENR, Paris.

Muller Mirza N. & Perret-Clermont A.-N. (2009) Argumenta- tion and Education: Theoretical Foundations and Prac- tices. Springer, Dordrecht.

Overdijk M. & van Diggelen W. (2008) Appropriation of a shared workspace: organizing principles and their applica- tion. International Journal of Computer-Supported Col- laborative Learning 3, 165–192.

Pasquier D. (2005) Cultures Lycéennes. La tyrannie de la majorité [High-School Students’ Cultures: the Tyranny of the Majority]. Autrement, Paris.

Pilkington R. & Walker A. (2003) Using CMC to develop argumentation skills in children with a ‘literacy deficit’. In Arguing to Learn: Confronting Cognitions in Computer- Supported Collaborative Learning Environments (eds J. Andriessen, M. Baker & D. Suthers), pp. 143–175. Kluwer, Dordrecht.

Rabardel P. (1995) Les Hommes et les Technologies: Approches cognitives des instruments contemporains [Humans and Technologies: Cognitive Approaches to Con- temporary Instruments]. Armand Colin, Paris.

Schwarz B., Neuman Y., Gil Y. & Ilya M. (2002) Construction of collective and individual knowledge in argumentative

activities: an experimental study. Journal of the Learning Sciences 12, 219–256.

Stahl G. (2005) Group cognition in computer-assisted col- laborative learning. Journal of Computer Assisted Learning 21, 79–90.

Stein N.L. & Bernas R. (1999) The early emergence of argu- mentative knowledge and skill. In Foundations of Argu- mentative Text Processing (eds J. Andriessen & P. Coirier), pp. 97–116. Amsterdam University Press, Amsterdam.

Suthers D., Toth E. & Weiner A. (1997) An integrated approach to implementing collaborative inquiry in the classroom. In Proceedings of the Second International Conference on Computer Support for Collaborative Learn- ing (eds R. Hall, N. Miyake & N. Enyedy), pp. 272–279. Toronto University Press, Toronto.

Thurlow C. (2003) Generation txt? The sociolinguistics of young people’s text-messaging. Discourse Analysis Online. Downloaded from the Internet. Available at: http:// extra.shu.ac.uk/daol/articles/v1/n1/a3/thurlow2002003- paper.html (last accessed 13 February 2011).

Wertsch J.V. (1991) Voices of the Mind. A Sociocultural Approach to Mediated Action. Harvester Wheatsheaf, London.

176 M. Baker et al.