Understanding and Working with �Failed Communication� in Telecollaborative Exchanges

Abstract:
It is by now well established that telecollaborative exchanges frequently end in �failed communication� and do not automatically bring about successful negotiation of meaning between the learners. Instead, the intended pedagogic and linguistic aims of online interaction are repeatedly missed, and projects may end in low levels of participation, indifference, tension between participants, or a negative evaluation of the partner group or their culture. The reasons offered in the literature are rather diverse in nature, and there has so far been no attempt to offer a comprehensive overview of such areas of dysfunction. Starting from a review of the existing body of research, this paper develops a structured inventory of factors which may lead to cases of failed communication in online exchanges. In sum, 10 different factors are suggested at four different levels: individual, classroom, socioinstitutional, and interaction. Examples of communication failure taken from two exchanges will be used to illustrate how these factors are interconnected and influence each other. It is concluded that a more discriminating perspective of such problem areas, both among the tutors and the students, can help to further increase intercultural awareness and lead to a better understanding of the dynamic nature of online communication.

KEYWORDS

Intercultural Learning, Online Pedagogy, Telecollaboration, Failed Communication, Computer-mediated Communication

Introduction

Telecollaboration refers to the use of online communication tools to bring together language learners in different countries for the development of collaborative project work and intercultural exchange. This type of network-based language teaching (NBLT) covers a wide range of activities and exploits a variety of online

623

communication tools, including email, web-based message boards, and videoconferencing. Apart from the linguistic advantages of engaging learners in authentic language practice with native speakers, telecollaboration is also seen to offer great potential for the development of the skills and attitudes of intercultural competence (Byram, 1997). However, the literature in the area reveals that success in telecollaborative exchanges is far from guaranteed and Kern warns that instead of improving levels of intercultural understanding, “we find here [in online exchanges] that exposure and awareness of difference seem to reinforce, rather than bridge, feelings of difference” (2000, p. 256). For this reason, this paper sets out to identify the reasons why online intercultural communication between language learners in such projects often fails to achieve the intended pedagogical goals. Why does online intercultural contact between language students often end in disagreement and misunderstanding? Why is a common outcome of an exchange the confirmation of negative attitudes and stereotypes towards the target culture, when the development of intercultural understanding is one of the central aims of the activity? By reflecting on the answers to these questions, researchers hope to develop a better understanding of what terms such as online intercultural communication and electronic literacy actually involve, while educators aim to improve the structure of their exchanges so they do not lead students to negatively evaluate their experience of contact with members of the target culture.

In the course of the paper, the umbrella term `failed communication' will be used to refer to cases of telecollaborative interaction which end in low levels of participation, indifference, tension between participants, or negative evaluation of the partner group or their culture. In order to establish an inventory of reasons for failed communication in intercultural online interaction, we will first of all carry out a review of instances of communication breakdown and misunderstanding reported in the literature. Following that, the inventory of reasons for failed communication will be organized into four levels and explained in detail. At this stage, various practical measures are also proposed in order to support teachers in their attempts to deal with problems in their exchanges. Finally, examples taken from a recent telecollaborative exchange will be used to highlight how factors from the different levels of the inventory can contribute to failed communication.

CONFLICT AND MISUNDERSTANDING IN ONLINE INTERCULTURAL EXCHANGE: REPORTS IN THE LITERATURE

Since the pioneering work in online projects carried out in the late 1980s by the Orillas Network (Cummins & Sayers, 1995; Sayers, 1991) and the AT&T Learning Circles (Riel, 1997), the activity of telecollaborative exchange has grown in importance and is considered today one of the main pillars of online language learning. In the literature on NBLT, a new approach to research into students' use of online networks has seen a shift from essentially quantitative research in networked learning in single classrooms to qualitative studies of interaction between groups of learners in different locations (Kern, Ware, & Warschauer, 2004). These studies have looked at a variety of issues related to online intercultural exchanges, including their contribution to the development of learner autonomy and linguistic

624

competence (Belz & Kinginger, 2003; Brammerts & Kleppin, 2001), the development of online literacy in second language learning (Kramsch, A'Ness, & Lam, 2000; Warschauer, 2000), and the pedagogical structure and design of online projects (Müller-Hartmann, 2000a; Meskill & Ranglova, 2000).

Particular attention has been paid to the reasons for failed communication as an area of further research. A review of the literature reveals that dysfunction in online exchange has been attributed to a complex, often confusing, array of factors related to the students and the sociocultural contexts in which they are operating, the organization and structure of the exchange, and the type of interaction which takes place between the groups in the online environments. It will be argued later that it is often impossible to rely on a single factor to explain an exchange's lack of success; reasons usually include a combination of interconnected factors from different areas.

The problematic area which seems to have received the most attention in this field of research has been the influence of differing sociocultural and institutional factors on the development of exchanges. A major contribution to the field has been the work of Belz, writing alone and together with Müller-Hartmann. They have produced an important body of research on how a wide range of social and institutional factors influenced the outcome of telecollaborative exchanges between university-level foreign-language students in Germany and the US. Belz (2001), for example, shows how the lower social and economic value of German in the US compared to English in Germany led to a proficiency mismatch between partners. As a result, a lower level of fluency in the foreign language meant that American students often wrote shorter emails than the German students. This was interpreted by their partners as a lack of friendliness and motivation and thereby led to negative evaluations of the exchange. Other social and institutional factors which the authors found to influence the exchanges included the misalignment of academic calendars, differences in societal norms with respect to technological access (for more information, see Thorne [2003] on different cultures of use of the internet and its effects on telecollaboration), divergent forms of assessment in the respective cultures, and the different physical layout of the universities (Belz, 2001, 2002; Belz & Müller-Hartmann, 2002, 2003).

The work of these authors is complemented by that of Ware (2005), whose qualitative study of an exchange between advanced students of English and German in the US and Germany highlighted the connection between the different uses of linguistic features by both groups in their online interaction, socioinstitutional factors, and low group functionality. Ware found that Americans asked far fewer questions than their partners and also made fewer attempts to establish personal rapport. This type of to-the-point, task-oriented interaction led to what she describes as `missed communication' between the students and left much of the German group dissatisfied with the exchange. Through the examination of in-depth interviews with members of both groups, Ware identified certain social and cultural factors which determined students' online behavior in each case. The first factor was seen to be the different levels of prior experience with online writing. The American group reported being quite used to using online technologies for

625

communication and learning and was therefore more critical of the technologies than their German counterparts who were found to be more enthusiastic about what was, for them, a novel way of language learning. A second influential factor was the differing social contexts in which each group was operating. The Germans were seen to be more motivated in their language learning than the Americans due to the different status of English and German in their respective countries. The Americans were also found to be much more grade focused than their counterparts. As such, their level of interaction was much more limited than that of their partners who were more motivated by an intrinsic desire to improve their English and make international contacts.

Looking at the Spanish-American context, O'Dowd (2005) reported on the socioinstitutional factors which influenced the development of a Spanish-American exchange based on the Cultura model of telecollaboration (Furstenberg, Levet, English, & Maillet, 2001). Apart from identifying the importance of factors already mentioned by Belz and Ware, including different levels of access to technology and differences in the evaluation requirements at the two institutions, O'Dowd also showed that the negative stereotypical images which each group brought to the exchange about the target culture were likely to influence the levels of motivation and participation of the students involved. Since this particular exchange took place during the academic year 2003-2004, the author found that the on-going war in Iraq and the complex political relationship between Spain and the US led many of the Spanish students to react negatively to the idea of an exchange with American students and a focus on American materials during class time.

Other research has revealed how individual students' motivation and intercultural communicative competence can have an important influence on the outcome of online partnerships. In reference to motivation, Ware (2005) identified individual differences in motivation as being an important factor in the low functioning of an exchange. In her study, success in the asynchronous exchange required students to spend a substantial amount of time reading and replying to correspondence, and this often clashed with the amount of time students had put aside for such an academic activity. The importance of individual students' intercultural competence is illustrated in O'Dowd's study (2003) of five Spanish-English email partnerships. He found that the essential difference between the successful and unsuccessful partnerships was whether students had the intercultural competence to develop an interculturally rich relationship with their partners through the creation of effective correspondence. This type of correspondence took into account the sociopragmatic rules of the partner's language, provided the partner with personal opinions, asked him/her questions to encourage feedback, tried to develop a personal relationship with the partner, and was sensitive to his/her needs and questions. Fischer's work on high school email exchanges (1998) also highlighted the importance of the students' level of intercultural competence. In this case, he underscored the need for students to bring tolerance and appropriate attitudes of curiosity and openness to their exchanges.

Although surprisingly little research has looked at the relationship between failed exchanges and the methodological aspects of the activity, elements such as

626

task design, the relationship between the teachers in both classes, and the ways in which students are prepared for the exchange have been acknowledged as having a significant influence on the outcome of telecollaboration. For example, in their discussion of a misunderstanding which developed between a German and English student in a web-platform exchange, Ware and Kramsch (2005) suggested that the teachers of both classes could have done more to avoid the breakdown of communication by ensuring that all messages were peer reviewed before being sent to the partner class and by organizing more tightly focused class discussions which would have given the two students in question an opportunity to analyze the messages with the help of their teachers and their peers. Other researchers have also looked at the important role which teachers can play in avoiding or dealing with low functionality and misunderstandings in telecollaboration. O'Dowd and Eberbach (2004) identified various key tasks which confront teachers who wish to exploit the potential for intercultural learning in online exchanges. These included developing students' awareness of how culture and language are interconnected and training learners how to make effective, culturally rich posts on the project's message board. The authors also underlined the need for both teachers to develop a good online working relationship together in order to co-ordinate and reach agreement on the many aspects of the exchange. Finally, in his work on the role of task design in telecollaborative exchange, Müller-Hartmann (2000a) affirmed the need for teachers to be able to identify problematic or provocative correspondence from the partner group which can then be brought into class and discussed in group or as a whole class. The author suggested that it is vital for exchanges to be integrated within the content of the regular classes so students can learn to reflect on and learn from their partners' correspondence.

A small number of texts in the literature argue that low functionality and misunderstandings in online telecollaboration can be explained by an analysis of students' reactions to cultural differences in online communication style and behavior. Despite many claims that the internet is a `culturally neutral' environment where, in the words of Kramsch and Thorne, many believe that “native and non-native speakers can have access to one another as linguistic entities on a screen, unfettered by historical, geographical, national or institutional identities” (2002, p. 85), a small but significant part of the research on online interaction has revealed that, first of all, the internet itself is based on specific cultural principles and values and that, second, users of the internet bring with them their own culturally specific communicative norms and modes of behavior which may or may not be compatible with those of other online users.

An example of this approach is provided by Belz (2003) who showed that the online interaction style of an American student, described as being uncommitted and self-deprecating, led his German partners to dismiss him as someone who was unwilling to engage in debate and confrontation. Similarly, the American student's interpretation of the German's correspondence, which was characterized by directness and categorical assertions, led him to reject them as being rude and aggressive.

In conclusion, it would appear that while there is a growing body of research

627

which attempts to understand why virtual contact between language learners often fails to achieve the intended pedagogic aims, the reasons offered are very diverse in nature and may not serve to give educators a sufficiently broad and comprehensive overview of problematic areas or to illustrate how these are interrelated. For this reason, the following section proposes a structured inventory of factors which may lead to cases of failed communication in online exchanges. Following that, some examples of this phenomenon taken from a telecollaborative exchange will be used to illustrate how these factors are interconnected and influence each other.

POTENTIAL AREAS OF DYSFUNCTION IN TELECOLLABORATIVE PROJECTS: DEVELOPING AN INVENTORY

The overview of the literature in the previous section reveals that various problems for successful long-distance collaboration have been identified and described in the past. However, those educators who have had experience with online exchanges are likely to acknowledge that the areas mentioned are not an exhaustive list of challenges. For example, it is evident that the thematic content of the tasks chosen for collaboration is also fundamental for the success of an exchange. In addition, to our knowledge, there has been no attempt to classify such problem areas in some didactically substantial way. Such an endeavor would appear both promising and useful because, in Belz's words, “one of the most constructive ways to ascertain the developmental path of a particular phenomenon is to study those instances where it is disrupted, that is, those cases where the system fails” (2003, p. 76). We would argue that by providing a classification of reasons for failed communication, educators can be better prepared for the challenges which await them in their online projects.

Our inventory organizes the reasons for failed communication described in the literature review above (and in our own research) into four different levels. Such a framework is of course a largely idealized display of reality and does not provide a definitive number of reasons. Nevertheless, it is intended to be a step in better understanding the reasons for success or failure in telecollaborative exchanges. Each of the areas in Figure 1 will be briefly commented on, whenever possible with reference to relevant research or to data taken from an exchange between Spanish and German university students of English who used English as a lingua franca in their interactions.¹ More examples and findings from a separate exchange will be provided in the next section in order to illustrate how these areas and levels are interconnected.

Moving from an inner to an outer perspective in the inventory shown in Figure 1, attention first needs to be addressed to an individual level, the learner's psychobiographical and educational background. Naturally, this perspective entails everything the learner brings into the learning process. We suggest focusing on two areas here that may give rise to setbacks. First, it is the learner's current level of intercultural communicative competence (ICC) which should be considered. This construct is well established in the literature and merges skills, attitudes, knowledge, and critical cultural awareness as established by Byram (1997). In

628

the context of online exchanges, learners with little experience in intercultural communication might lack necessary skills of discovery and interaction in order to deal with tasks which involve ethnographic interviewing (see O'Dowd, 2006). Similarly, the simple lack of factual knowledge, also part of Byram's concept of ICC, may lead to misunderstandings in the interpretation of messages from members of the target culture. The example of Katrin,² a German student taking part in the exchange with Spanish students can serve to illustrate this point. She reported taking away a negative impression of the partner group for the following reason:

Some of their utterances were quite provocative, for instance concerning the relationship between the USA and Europe. One of the Spanish side wrote that she did not like the USA. To my mind, if someone has an opinion, he or she should substantiate it.

Figure 1

Inventory of Reasons for Failed Communication in Telecollaborative Projects

However if Katrin had had more knowledge about the political climate in Spain at the time of the exchange (when the question of the participation of Spain in the second Iraq war had heavily influenced the attitudes of the Spanish student population towards the US), she would have perhaps been more understanding of her partner's statement and would not have reacted to it so defensively.

A second key area at the individual level giving rise to special attention could

629

be summarized as the learner's motivation and expectations towards the exchange project. There can be no doubt that the participants' types of motivation and their presuppositions are of great consequence for the success or failure of telecollaboration. In her study, Ware (2005) looked at these two related factors in more detail. One source of tension, she concluded, can be differences in expectations towards a project with regard to grammatical accuracy, message length, or response time. It is perhaps inevitable that there will almost always be differences among learners with respect to how they hope to benefit from an exchange and how much time they are willing to invest in it. However, potential problems can be avoided to a certain extent if both teachers and students are aware of the situation and the expectations of their partners. Extensive contact between teachers before the exchange and introductory writing tasks which involve students telling their distant partners about how they feel about their future work together can go some way to increasing sensitivity to the needs of the other group.

Moving on in the inventory, the classroom (or methodological) level is situated between the individual and the institutional level. Even though such a methodological perspective seems to be the most straightforward field for research, it appears to have received less attention than the other levels. In Figure 1, the areas 3, 4, and 5 have been placed at the interface of the two classrooms because decisions in these fields usually have a direct effect on both classrooms. Other areas (6 and 7) influence rather the local classroom situation and will have only an indirect impact on the partner class.

The importance of a good teacher-teacher relationship has repeatedly been recognized in the literature. O'Dowd and Eberbach (2004) in their account of a German-Irish exchange stressed the necessity of coordinating the many aspects of such a collaborative project in detail, and Müller-Hartmann (2000b) also looked at the problems which arise in teacher-teacher communication. A first step to deal with this source of difficulty is the awareness that telecollaboration is indeed a form of virtual team teaching that requires thorough consensus on all the facets of the cooperation. Contrary to other team-teaching efforts, teachers involved in online exchanges often do not even know each other face to face and shrink from disclosing all their plans to their partner teacher, a fact that is often made more difficult by the virtual nature of their communication.

The second area at classroom level, task design, is one of the domains that involve a profound working relationship between the teachers. Online exchanges are usually arranged around a number of tasks, and their thematic content and sequencing require careful consideration. Teachers are frequently familiar only with the curricular requirements of their local group and have good intuitions what kinds of tasks will lead to positive results in their own teaching contexts. Naturally, there will often be a tendency to try and implement those tasks that are in accordance with their own goals. Finding a compromise in task design that reflects the interests of both groups requires time and openness towards the other side's suggestions. Unfortunately, there is little research to support teachers in the process of answering important questions in the planning of their exchanges: What tasks are particularly suitable for telecollaboration? Are there patterns for

630

sequencing different kinds of tasks that can be recommended? To what extent should student correspondence be written in the target or mother language? (For a description of language use in tandem exchanges, see Appel & Mullen, 2000; for a comparison of language use in the Cultura and tandem exchanges, see O'Dowd, 2005.) To what extent should students have a say in the selection of tasks? It is possible that in many telecollaborative projects teachers may be so preoccupied with organizational or technical issues that thematic content often comes only second within the available time frame.

The clash between student and teacher needs in task design was evident in the German-Spanish exchange mentioned earlier. In the end-of-term feedback, all the negative comments made by the Spanish group in reference to the project were in some way related to aspects of task design. One student complained that the topics chosen by the two teachers (based principally on the discussion of newspaper articles and the comparison of student-led surveys) had hindered the development of the online relationships between the German and Spanish students: “I know that talking about culture could be useful for increasing our knowledge of the world, but neither Germans or Spanish ones will be totally close to one another if we don't talk about hobbies and personal interests.” Another student complained that the fact that the teachers had not corrected language errors before the messages had been posted (another significant issue in task design) had led him to feel he had learned little from the exchange: “We have made mistakes in our messages, but we don't know what they are. This has been a waste of time.” A remedy for problems in task design may lie in a rigorous process of advance planning and consultation between students and teachers. For example, teachers could distribute preexchange questionnaires to students in all the classes asking them for their opinions on topics that they would like to discuss with their partners. To deal with the issue of accuracy in the L2, teachers could also carry out in-class reviews of their students' correspondence before the messages are actually sent to the partner group.

We suggest learner-matching procedures as a third area of potential dysfunction at classroom level. Even though questions of how students can be allocated in virtual pairs or groups seem to be somewhat trivial, it can have a massive effect on the results of the cooperation. Reports in the literature indicate that students are often assigned on a one-to-one basis. As is well known, some of these pairs turn out to be more successful in their interaction than others. The question, then, would be whether to leave these combinations to chance or to consider factors like age, gender, or foreign language proficiency when assigning students. Belz (2001), for example, has shown how a lack of proficiency in the foreign language can lead to students writing shorter correspondence than their partners and that this can consequently be interpreted by their partners as a lack of openness and friendliness. In this context, it is often assumed that telecollaboration takes place between two fairly homogeneous cultures. The literature usually just refers, for example, to a German-Spanish telecollaborative exchange. However, it should not be forgotten that nowadays the composition of groups is usually multicultural. The outcome will be totally different if a student from Portugal, to take an

631

example, happens to spend a term in Germany and is supposed to represent and describe Germany as `his' or `her' country in such an exchange. Sometimes pairs or small groups of students are assigned on each side for cooperation. This may be due to institutional constraints (e.g., class size or the number of available computers) but may also be the result of a didactic desire to support negotiation in the foreign language among the members of the local group. Often, groups on both sides are not evenly distributed—one or two students on one side cooperating with three on the other. It is possible that such setups can have negative effects on the interaction between such groups.

As a fourth challenge at classroom level, we propose local group dynamics. This is closely related to the one before insofar as it directs attention to the atmosphere in the local class, a facet of cooperation that has repeatedly turned out to be an obstruction in our projects. Naturally, in telecollaborative projects, researchers, as well as teachers, tend to focus on the relationship between members of the two different classes. However, learners in pair or group constellations need to negotiate tasks not only with their online partner(s) but also within their local group. In many university classrooms in Germany, for example, participants have met their fellow students only briefly, and need to agree face-to-face with them on given tasks, formats and deadlines, a process which can turn out to be even more difficult than reaching agreements with virtual partners. Thus, the fact that a major part of the learning process takes place within the local context should not be neglected.

The quality of the preexchange briefing, our final aspect at the classroom level, can also be influential on the development and outcome of the exchange. Prior to the actual exchange, learners can be prepared for the stumbling blocks of the cooperation (e.g., technical issues and organizational commitments like response time); factual knowledge about the partners and their background can be provided, or the learners' expectations can be examined and compared to those of the teacher. Preparatory activities and themes should be harmonized to the greatest possible extent to avoid major discrepancies between the expectations and the background expertise of the two groups.

The third level of our inventory in Figure 1 focuses on dysfunctions at the socioinstitutional level. This perspective seems to have attracted the most attention in the literature and covers a diverse collection of factors to be taken into account when researching or planning long-distance cooperation. What all these topics have in common is that they pinpoint external areas, often beyond the participants' direct control. We suggest that the following three areas cover the most essential factors that others have put forward or that we have identified as influential in our own exchanges.

The technology itself—the medium by which long-distance exchanges are carried out—and the degree of access to it have an undeniably strong impact on the course of the exchange. Most commonly used communication tools today are probably still text oriented, while video- and audio-based tools are gaining in importance (see O'Dowd, 2000, 2006 for examples of the use of videoconferencing in telecollaboration). Nowadays, message boards are usually embedded in learning

632

management systems like WebCT, Blackboard, or Moodle,³ allowing teachers to provide easy access to other resources and materials related to the exchange. We have found that even supposedly little differences between such systems, or the way they are set up, can have an effect on the exchange. For example, can members attach photos to their names or attachments to their messages? How comprehensible and user friendly are the general layout and organization of the communication tools and forums? Is the learners' view restricted to their partners, or can they read what their classmates have written? One German student made the following comment about the German-Spanish exchange which had been carried out in a Moodle platform: “At the beginning it was confusing because there were so many introductory mails and the names [of the Spanish students] sounded quite similar. In the next week it already got better although you could still hardly remember to whom you had talked before. Hence, I liked the idea to have little pictures beside each name. I also thought it was nice that we took a class picture.”

Apart from the tools themselves, access to them can also be problematic. How natural is the use of online tools in the students' private lives? Do students have permanent out-of-class access to them? Ware (2005) in her study suggested that this external factor can have a significant influence on students' motivation and expectations. It is easy to imagine how limited access and hence shorter messages or a longer response time can lead to misunderstandings between differently equipped partners.

A second area at the socioinstitutional level, the general organization of the students' course of study, is meant to cover a variety of circumstances that determine the general context of the participants' schooling or studies. Belz and Müller-Hartmann (2003) identified four instances of how socioinstitutional affordances and constraints can affect German-American partnerships: (a) the misalignment of academic calendars, (b) culture-specific assessment patterns, (c) differences in the educational backgrounds of each teacher and the particular aims which they each have for the exchange and, (d) the differences which exist between both countries as regards student contact hours and the physical organization of each university. Such differences, the authors warned, can lead to disagreement on the amount of student work considered appropriate, on the setting of deadlines and on the functionality of cross-cultural group work. From our own experiences, two related facets could be added. First, exchanges often bring together groups following wholly different courses; one group might be enrolled for a class in cultural studies and intend to focus on intercultural themes, another group of teacher trainees is primarily expected to investigate the educational potential of online interaction at a metalevel, while yet another group envisages the collaboration essentially as a means of authentic language practice. It is easy to imagine how such different academic backgrounds and associated objectives can lead to misunderstandings about task design and communicative behavior, not only between teachers but also between students. Second, on many occasions, students as well as teachers act in accordance with a more general teaching and learning philosophy than is implicitly or explicitly set out at a socioinstitutional level. In some

633

institutional contexts, for example, there is a considerable amount of freedom in working through a set syllabus, and it is largely the teacher's responsibility to decide what exactly s/he does in a given time frame. His or her partner teacher, on the other hand, might be under a lot more pressure to fulfill set requirements and justify the decisions made. Similarly, students' attitudes can be partly predetermined by such sociocultural differences and range along a continuum from minimal to total commitment. Again, consequences with regard to reliability or the amount of work that is considered appropriate are evident.

Finally on this level, attention should be drawn to a well-known phenomenon of any intercultural exchange that might cause dysfunction—the differences in prestige values of cultures and languages. Any exchange represents a particular constellation of involved cultures in general and languages in particular, and it is beyond question that these constellations can be a complex source of misunderstandings, feelings of inferiority or resentment. Aspects of cultural stereotyping have been investigated in various facets in the literature. For example, O'Dowd's (2005) description of a Spanish-American exchange, mentioned in the previous section, revealed how anti-American sentiments at the time of the Iraq war in 2004 interfered with the intended negotiation of cultural meaning between the groups.

The final level of the inventory in Figure 1 is the interaction level and refers to the misunderstandings and tension which arise from cultural differences in communicative style and behavior. This aspect of failed communication reflects the focus on intercultural pragmatics of writers such as House who argued that “an emotional reaction [to cultural differences in communication styles] is often the major factor responsible for a deterioration of rapport and for the mutual attribution of negative personal traits which, in turn, effectively prevent any recognition of real differences in cultural values and norms” (2000, p. 147). This means that learners often assume that peculiarities in their interlocutors' way of communicating are due to personal oddities while, in reality, they are part of the target culture's communicative style. Examples of this include the different weighting given to small talk by German and English speakers or different attitudes towards indirectness between Spanish and English speakers. Spencer-Oatey (2000) suggested that cultural differences can emerge in communication in various areas, including illocutionary, discourse, participation, stylistic, and nonverbal domains. Although Spencer-Oatey was referring to traditional face-to-face intercultural interaction, it is possible to extrapolate the following issues related to each of these domains in the context of online collaboration.

1. Illocutionary domain

What is deemed to be the appropriate use of speech acts in the online interaction?

2. Discourse domain

How should the interaction be organized? How should the posts/messages be structured? How much small talk is appropriate? What topics are deemed appropriate for discussion?

634

3. Participation domain

Who should write first and second? How often and how quickly should one reply to the partner's posts? To what extent should explicit reference be made to previous posts?

4. Stylistic domain

What tone is considered suitable in the online interaction? Are irony and humor acceptable? What is seen to be genre-appropriate lexis and syntax?

5. Nonverbal domain

In a text-based medium, how can students make up for the lack of visual and paralinguistic cues? To what extent should emoticons and images be used to express emotions and content?

All of these areas can have a major influence on the outcomes of telecollaboration. In the German-Spanish exchange, different cultural attitudes to what was considered appropriate participatory behavior (i.e., the participation domain) led some German students to make a negative evaluation of their partner group: “I doubt if they took it seriously as we did … they did not always answer our questions and I had the impression that we were the ones who had to make the first move, e.g. creating new topics etc.” The stylistic domain was also seen to be the source of failed communication in this particular exchange. In one group discussion, one Spanish student wrote on the topic of heroes in her country that

… in these days, El Quijote by Cervantes has been considered the most important book of humanity.

Her statement, intended as a report of information which the student had read in a newspaper article, was interpreted by her German partners as an arrogant form of boasting about the home culture: “Is it not quite ignorant, one-sided and rude in respect to other cultures to say this?” Anticipating in advance the exact misunderstandings on the interaction level of an exchange may be impossible, but teachers can take some care to prepare students for these issues by providing them with worksheets containing examples of previous exchanges, showing how misunderstandings occurred due to cultural differences in interaction norms and expectations. Students can be asked, for example, to read through these transcripts and suggest explanations as to what was the intended meaning of the writer and how the text was misinterpreted by the reader.

At this stage, it is important to acknowledge the interrelationship between the four levels of the inventory. For example, it is clear that the learner's motivation and expectations are mutually linked with most of the other fields. Learners' motivation is likely to be high if the chosen technology (e.g., chat) is new and exciting for them in a classroom context, and it is likely to be relatively low if their previous experiences with the technology were ambivalent. To cite a second example, the teacher-teacher relationship and the amount of negotiation between them can have a backwash on the appropriateness of the preexchange briefing, which, in turn, can influence learners' views of what to expect in the forthcoming exchange. Other interrelationships are more subtle, as will be illustrated in the following section.

635

SAMPLE DATA FROM A TELECOLLABORATIVE PROJECT

The asynchronous exchange described in this section was carried out in the summer term 2005 and involved students of English at Essen University in Germany and students of German at Macquarie University in Australia. One of the authors was the teacher of the German group and followed the notion of participant observation and action research.⁴ The project was run on the Moodle platform. Data collection included teacher log books, an archive of all the students' written texts and teachers' messages, written interim and final evaluations by the students, and informal interviews with selected students.

In the 8-week exchange, students worked on four different tasks: (a) an introduction of themselves, (b) the Pope and the media, (c) student life, and (d) typically German, typically Australian? Part of the German students' final assignment was a written evaluation of the whole project, including an appraisal of these four tasks. The students offered generally positive feedback on the tasks, except for the second task (the Pope and the media). Of the 32 students on the German side, 19 were openly dismissive of the task. The intended topic—a comparison of media coverage of a global event—was misinterpreted as an invitation to debate religion.

Students' initial reactions to the second task quickly indicated one source of misconception. German student Sven started his first message as

Hi folks, our task is about the old and the new pope. I am afraid to say I am not very religious, so what can I say? The only thing I do know is that this non-stop Vatican TV is slowly getting on my nerves.

In his final evaluation (which was addressed only to his teacher), he wrote

I truly did not like the second topic. Not everyone is interested in the Pope and/or religion. So our messages about this were very short and superficial. … This disturbed the rather personal beginning of the exchange. It is definitely not a good choice as the second task for people who barely know each other.

A closer inspection of Sven's reactions testifies to various problem areas. On the teachers' side, the task instruction apparently gave rise to misinterpretation, and teachers' preexchange negotiations were not thoughtful enough to prevent the misunderstanding. Sven, on the other hand, created a causal relationship between his lack of interest and the brevity and superficiality of his messages. Of course, this allows a first conclusion to be drawn about his level of intercultural competence (i.e., his attitudes of openness and curiosity towards his partner's views). His implicit suggestion to do this task at a later stage (when people know each other better) was shared by many students. Leaving aside the fact that his proposal somewhat contradicted his declared lack of interest, it raises the question of the best possible sequencing of tasks. What kind of task—beyond an introductory round—is most suitable for “establishing cultural identity” (Müller-Hartmann 2000a, p. 136) and building up a positive working relationship between the exchange partners? The teachers' original intention for the task (media coverage of

636

a global event) had indeed been a gradual start into the exchange. The supposed byproduct, the religious dimension of this task, was clearly underestimated by the teachers and largely impeded the students' interactions. In sum, task design, alongside the students' psychobiographical backgrounds, and the quality of teacher-teacher cooperation led to interaction failures among various members of the groups.

In other cases, though, the analysis is more intricate. The following two examples on the same topic demonstrate that supposedly similar communicative behavior and constellations can result in reasonable success in one case, and open failure in the other. Starting off with an ironic remark about not having heard of the new Pope in the media, Judith and Sandra from Germany then carried on with a fairly blunt declaration that

the church (at least here in Germany) is not very attractive for young people!

Towards the end of their opening message they took up a piece of information from the introductory round and stated

We know that you and your brother are active members of the church and we think that church in Australia is more open and lively, isn't it?

The addressed Australian student replied by describing the different churches in Australia but first gave expression to her accommodating attitude.

First of all I just want to let you guys know that I'm not catholic, so will not be offended by anything you may have to say about the pope.

While here the bluntness of the opening message was somewhat counterbalanced by the partner's forbearance (and maybe the fact that she happened not to be Catholic), in the second example the dialogue developed differently. Amy and Justin from Australia chose a more solemn tone in their opening message, indicating clearly that they considered themselves religious and stressing their active participation in church matters. They then gave thought to the essence of the original task instruction, approaching it in rather academic fashion.

We are questioning why, that in a time when the topic of religion is a taboo, being suppressed in schools and losing significance in people's lives, that the death and change of the pope was such a large media event.

This time the (indirect) question was not taken up by the two German girls who almost exclusively focused on distancing themselves from church in Germany. In addition, they introduced further controversial topics that they related to the task, criticizing the attitude of the Catholic church towards abortion and homosexuality. Having worked off their list of criticisms, they concluded in a conciliatory way.

We don't want to offend or influence anyone with views. Looking forward to hearing/reading your opinion.

Again, contrary to the first example, this time the partners' reaction was not as

637

accommodating. Justin basically ignored the vaguely obliging closing announcement of the German students and directly addressed his apprehension with the partners' criticism—and with the issue of abortion in particular. At this point the original task seemed to have fallen into oblivion, and, indeed, the final message of that thread by Vera, one of the girls of the German group, was written in an overly emotional way.

I, as woman (!!!), do not agree with you. If a woman is raped, she mustn't be forced to bear a child of the criminal! That is what I think! Further more, the woman can be left by the man and then she has got a problem ... Bye, Vera

With reference to our inventory, various explanations for the different outcomes of these two examples can be offered. First, at the interaction level, differences in tone (i.e., stylistic domain) are evident, with a more ironic approach in the first case and a more solemn tone in the latter. Also, the personal and direct question in the first example seems to have been more conducive to interaction than the more indirect, academic approach in the second group. Second, differences in learners' individual attitudes and levels of intercultural awareness become evident. While the first exchange is positively influenced by one of the partners' forbearance, in the second group students rather seem to fall victim to their set beliefs about the topic and their lack of tolerance. This is confirmed by the stereotypical explanation that Vera offers in her final evaluation, stating that “as one could see with the pope task, religion always creates borders and hardly ever does the opposite nowadays.” Third, learner matching happened to bring male and female students together in the second case. It is evident that a topic like abortion which developed out of the original task can lead to gender-related tension. A final reason for failure that is represented in our inventory can be found in the German group constellation. The second girl of that group, who had not made any further contributions to the topic after their first reply, explained in a later interview that she had “felt uneasy with Vera and her rude style.”

To conclude, this range of reactions and variations in students' communicative behavior can be considered typical of misunderstandings in an exchange and illustrates the interrelatedness between the various disruptive features that we outlined in Figure 1. Starting off from a lack of consideration in task design and its thematic content, the majority of the students fall victim to their individual expectations with regard to topic outcome and their insufficient skills in seeing through these conflicts and their genesis. This can be aggravated by local as well as virtual group constellations, which are usually left to chance. Students' specific and culturally influenced verbal behavior, especially when writing in a foreign language, can give rise to further linguistic misinterpretations. All this suggests that to a certain extent one has to comply with the unpredictable and dynamic nature of such exchanges and take the risk of partial interaction failure. However, a more discriminating perspective of potential challenges, both among the teachers and the students, can help to further increase intercultural awareness and the likelihood of satisfying results.

638

CONCLUSION

The aim of this paper has been to identify the reasons for failed communication in online intercultural exchanges, to organize these reasons in a pedagogically useful manner, and then to illustrate how these reasons are closely interconnected and influence each other. It is hoped that such an inventory of problematic areas will provide tutors involved in the area of NBLT with a greater understanding of the many complex factors which come into play when students engage in online collaborative tasks with partner learners in distantiated cultures. However, it is important that this inventory should not simply be seen as proposing a path for avoiding cases of failed communication in telecollaboration. In the words of Belz, “It is very important to understand that these contextually-shaped tensions are not to be viewed as problems that need to be eradicated in order to facilitate smoothly functioning partnerships. … Structural differences frequently constitute precisely these cultural rich-points that we want our students to explore” (2003, p. 87). Educators interested in organizing telecollaborative projects should therefore have an in-depth understanding of the possible reasons for failed communication and should also possess a battery of techniques and practices which they can use in the course of their online exchanges in order for their students to derive maximum benefit from the exchanges. These techniques and practices could include the following: (a) engaging their students in classroom analysis of a collection of examples of failed communication from a previous exchange, (b) a rigorous approach to communicating with their partner teachers to enable both to understand the sociocultural context in which their partner class is operating, and (c) taking an on-going action research approach to their classes which involves collecting and analyzing online interactions and subsequent feedback from their students