1 Introduction
In recent years, computer-mediated communication (CMC) has been adopted as one of a number of language teaching tools, with email one of the most commonly used forms. A large body of studies on second language (L2) learning via email has established that email interaction enables learners to enhance their language learning experiences (Aitsiselmi, Reference Aitsiselmi1999; Florez-Estrada, Reference Florez-Estrada1995; Ioannou-Georgiou, Reference Ioannou-Georgiou1999; John & Cash, Reference John and Cash1995; Knight, Reference Knight2005; Rooks, Reference Rooks2008; Stockwell, Reference Stockwell2003; Stockwell & Stockwell, Reference Stockwell and Stockwell2003; Ushioda, Reference Ushioda2000; Woodin, Reference Woodin1997). Email has been found to serve as a means for students to access authentic language and learn about culture (Gray & Stockwell, Reference Gray and Stockwell1998), and Itakura and Nakajima (Reference Itakura and Nakajima2001) state that email removes some of the stress associated with communicating in an L2, and provides a record from which to monitor one's own learning processes. Email projects may also encourage students to draft and edit their work, with Tella (Reference Tella1992) finding that students drafted their emails in a word processor before compiling the revised draft into the email program, despite their teacher's leniency towards spelling mistakes and similar errors in the electronic medium. Yet, while email or similar CMC activities have been introduced into many L2 classes, what happens beyond the class-based project is an area that has been underexplored. This paper explores students’ email communication with native speakers (NSs) of Japanese after the conclusion of a university email project as one example, but has implications that are applicable to many forms of CMC.
According to Yoshinari (Reference Yoshinari1998), email can provide an enjoyable way to make use of one's L2 as a means of communication, and Ishida (Reference Ishida1995) agrees that it can be a powerful motivator for students. Other benefits include amending stereotypes (Itakura, Reference Itakura2004; Yoshimura & Miyazoe-Wong, Reference Yoshimura and Miyazoe-Wong2005), and giving learners exposure to language variation in the form of popular grammar, slang, and regional dialects (Kano, Reference Kano2004) which may be difficult to achieve in a traditional classroom setting. In another study carried out by Itakura and Nakajima (Reference Itakura and Nakajima2001), students reported attaining information about cultural differences, improving their Japanese, and their research skills. A number of benefits were found in Torii-Williams’ (Reference Torii-Williams2004) email project in a Japanese language program, including students correcting each other's electronic letters, learning more Japanese characters, and noticing gaps. Stockwell and Levy (Reference Stockwell and Levy2001) also found evidence to suggest that the level of participation in an email project was linked to improvement in L2 output, and this higher production was related to moving beyond teacher-assigned topics to their own areas of interest. A further study carried out by Stockwell and Harrington (Reference Stockwell and Harrington2003) found that learners of Japanese involved in a five-week email project with NSs made consistent improvements over a number of measures, including a reduction in errors. However, Stockwell (Reference Stockwell2004) observes that while there are benefits for learners, there are problems that teachers and students need to be aware of also. Such problems include communication breakdown, which may occur because of a variety of lexical, syntactic, or pragmatic errors.
Although recent trends indicate that email may no longer be the preferred CMC medium of choice for university students (Thorne, as early as 2003, for example, found that email constituted a constraining variable in the intercultural communication of one student who would have preferred Instant Messaging (IM) to participate in telecollaboration), email remains an important communication tool, largely because it is via email that most people gain access to other modes of communication. Pasfield-Neofitou (Reference Pasfield-Neofitou2012) outlines the key role that email plays as a “passport” between modes of communication, highlighting the fact that the vast majority of forms of CMC require an email address to sign up, and many relationships either transition from the “real” world to the “virtual” world via an exchange of email addresses or Facebook names (which can then be used to search for IM contacts, Facebook and other Social Networking Site (SNS) profiles, and so on), or transition from the “virtual” world, starting out on an SNS for example, and then shifting towards email for private communication, in preparation for a face-to-face meeting. While other forms of communication are largely transitory (with IM and MySpace, for example, being largely supplanted by Facebook in recent years), email has remained in constant use throughout the history of the internet. There are also other advantages to the use of email in comparison with a tool like Facebook for educational purposes, such as the fact that students do not need to register for a service they may not already use (contrary to popular reports, not all students have or want Facebook pages!); unlike Facebook's wall, the default setting for emails is private, which may be appealing to less confident students, and importantly, there also need be no awkwardness regarding the removal of a contact as a “friend” from one's list if interactions are not to be continued. Most students are able to use their institutions’ email without needing to agree to an End User License Agreement (EULA) of an external company such as Facebook. Despite these considerations, we believe there may be an argument for flexibility, rather than the mandating of a single tool, as will be further explored in section 3 below.
In addition to the ubiquity of email, with a Radicati (Reference Radicati2011) report estimating that the number of email accounts will increase from 3.1 billion in 2011 to nearly 4.1 billion in 2015, and the key role email plays in facilitating other forms of CMC, another important aspect when considering the use of email is the popularity of mobile phones in Japan, and the global increase in smartphone use. Cell or mobile phone email has been in standard use (as opposed to SMS) for over a decade in Japan, and the use of email applications on smart phones is steadily increasing elsewhere. In Japan, where home and office space is often limited, and as a result, laptop or mobile phone use is very popular, it is quite likely that at least some of the Japanese contacts of participants in the current study were using their mobile phones. Indeed, Lucas’ contact Hisayo indicated that she communicated with him via her mobile phone. As emails composed on a mobile phone, which are often shorter and more informal in nature, become more common, and are viewed either on another mobile or on a computer, it is likely that perceptions of email (such as Thorne's Reference Thorne2003 report that email is unsuitable for age-peer relationships as it was primarily used for communication with teachers or parents) will change.
Finally, Pasfield-Neofitou (Reference Pasfield-Neofitou2012) also indicates the important role that email communication has for language learners as they transition from university to the workplace. Email activities such as that currently under investigation may provide opportunities for students to not only develop their L2, but to develop their professional communication skills. As the same Radicati (Reference Radicati2011) report indicates, the average corporate email user in 2011 sent and received 105 emails per day, and while the rate of growth is expected to slow over the coming years as IM and social networking use increases, the number of emails both sent and received is predicted to continue to increase.
Gonglewski, Meloni and Brant (Reference Gonglewski, Meloni and Brant2001) outline a convincing rationale for the use of email in L2 learning, stating that it provides a context for authentic communication by promoting student-centred language learning, encouraging participation, and connecting speakers quickly and cheaply. Importantly in the context of the present paper, they argue that email activities expand the topics students can cover beyond those relating to the classroom, and that above all else, these activities may help to extend language learning in terms of both time and place. The authors argue that while email may be considered relatively “low-tech” nowadays, it is still effective in providing benefits to learners. Indeed, numerous recent studies, including Bower and Kawaguchi's (Reference Bower and Kawaguchi2011) analysis of Japanese/English tandems which combined chat and emails, Kabata and Edasawa's (Reference Kabata and Edasawa2011) examination of another Japanese/English exchange, and Conley and Gallego's (Reference Conley and Gallego2012) study of negotiation of meaning in email tandems demonstrate that such activities remain a popular use of CMC in the classroom. However, while the use of email for tandem language exchanges may be described as well-entrenched, little is known about the long-term effects.
Most of the above studies investigated email exchange projects which were set up as a compulsory task as part of a language course curriculum. While such projects have become well-established in recent years on the basis of the kinds of benefits observed above, follow-up analyses of what occurs after a curriculum activity finishes, particularly investigations of what factors promote continued communication, or cause communication breakdown, have received less empirical attention to date. For this reason, we sought to investigate the longer-term effects of an established email project.
1.1 Research questions
This paper aims to explore the factors which promote language learners’ continuation or non-continuation of email interactions after the end of their coursework email exchange task. We utilize the term, ‘non-continuers’, to describe those learners who cease interaction with their partner at the conclusion of the project. To our knowledge, it appears that there has been little research into learners’ language learning opportunities through continued interaction after an exchange task as a part of coursework ends. Furthermore, there is still an absence of research into naturally occurring online language exchange between non-Japanese and Japanese students.
Based on the goals of this research, this paper aims to explore the following research questions:
1. Does an email exchange activity provide long-term opportunities for communication with native speakers beyond the class, and what factors promote the continuation or non-continuation of these email interactions?
2. What benefits might exist in allowing students to communicate with their current Japanese contacts in a course project?
The first research question, which explores the long-term opportunities for the continuation of interaction following an activity, is the primary research question explored in this paper, and will be addressed in relation to class surveys, and the in-depth analysis of five learners. The secondary question, which examines the potential benefits of allowing students to communicate with contacts with whom they already have established relationships, will be addressed mainly in relation to a single case study, outlined in section 2.
1.2 Conceptual framework
As a means of investigating the factors influencing language learners’ continuation or non-continuation of email interaction, we utilize Stockwell and Levy (Reference Stockwell and Levy2001) as a framework. Stockwell and Levy report that there were several factors which appeared to influence the sustainability of email interactions, specifically, learners’ language proficiency, computer skills, in-country experience, number of interlocutors, and the topic of the email messages. Their discussion of these influential factors provides the current study with an instructive frame of reference. To support the findings of this study, Stockwell and Levy's (2001) framework is utilized in conjunction with salient features reported by Ware (Reference Ware2005), which may be broadly categorized as expectations and norms, social and institutional factors, and motivation and use of time.
2 Methodology
This paper draws upon data collected in two separate research projects at the same Australian university. In the main study reported on here (Takahama, Reference Takahama2010), fifteen undergraduate students volunteered to participate. In order to gather in-depth information in relation to their continuation or non-continuation of interaction with their project partners via email, triangulation of data collection procedures was employed. These methods consisted of two questionnaires, the collection of email messages, and a face-to-face interview with participants. Firstly, a 2009 questionnaire undertaken in Japanese 3 (J3), a lower-intermediate level of Japanese, was collected from the coordinator in order to learn how the students evaluated the “J3 email project”. Secondly, in 2010, a hard copy questionnaire was distributed to learners (who had taken part in the J3 email project the previous year) in the J5 lecture. This questionnaire was designed to obtain information regarding the continuation or non-continuation of learners’ email exchange as well as to recruit volunteers for the current study. The data collected from this questionnaire was categorized according to the students’ motives with regard to their continuation or non-continuation of email interactions with their Japanese partners.
Thirdly, this study utilized two types of email message samples. One sample type was the email messages exchanged during the “J3 email project” submitted by the four Japanese learners and their partners. These email messages were examined by drawing mainly upon Stockwell and Levy's (2001) framework in order to identify what factors promote learners’ continuation or non-continuation of email interaction. Further, to shed light on the intentions and motives of the participants, the students completed a brief face-to-face interview with the researcher. The second sample of emails collected was the messages exchanged after the J3 email project ended. All interviews were audio-recorded and transcribed immediately after each interview.
2.1 Participants
The participants were all enrolled in the higher-intermediate level of Japanese, also known as Japanese 5 (J5). These learners were chosen because approximately 40% of J5 students previously participated in a compulsory email exchange activity in J3 in semester one in the previous year. In order to evaluate students’ continuation/non-continuation of these exchanges, it was necessary to recruit participants one year on. More information regarding the J3 email exchange activity is given in the following section. In total, fifteen participants submitted responses to the questionnaire, and four of these participants volunteered to submit their email messages exchanged with their email partners and participate in a face-to-face interview. They and their partners have been assigned the pseudonyms of Adam, Beck, Cindy, Dina, Takashi, Tomomi, Takako, Kanako and Yuki. The four participants’ background information is presented in Table 1.
Participants’ Background Details
The findings related to these four students and their partners are contrasted with a case study from Pasfield-Neofitou (Reference Pasfield-Neofitou2012), which examined the email project-related and other communication of a student who also participated in the email project under investigation, with the pseudonym Lucas. While Pasfield-Neofitou (Reference Pasfield-Neofitou2012) concentrates mainly on Lucas’ naturalistic communication, the present paper draws largely on other data collected during the study to explore his participation in the formal email project. Like the students in Takahama's (Reference Takahama2010) study, participants in Pasfield-Neofitou's (Reference Pasfield-Neofitou2012) research submitted their CMC exchanges and took part in interviews.
Lucas took part in the J3 email exchange in 2008, under the same conditions as those students described above. However, Lucas’ experience makes an interesting comparison to that of either Adam, Beck, Cindy or Dina, in that rather than communicating with a partner assigned by the teacher, he chose to participate in the J3 email project with his friend Hisayo, with whom he had an established relationship. We have chosen to compare Lucas’ case with the other four students in order to examine the effects of a structured email exchange on students with and without established relationships, in terms of continuation of communication, as outlined in the first research question, and to discover whether such an established relationship might afford the learner any benefits in completing this task, as outlined in the second research question. Lucas’ details are provided in the final column of Table 1 below.
2.2 Nature of the email project
The email exchange activity was a compulsory assignment, and took place between learners of Japanese in Australia and students at a Japanese university between May and the beginning of June. Those students who already had established relationships with Japanese NSs had the option of inviting their current epal (or 
$$$$
merutomo) to take part in the project, while those who did not were asked to send their teacher an email requesting matching with one of the Japanese students. Lucas was excited about the opportunity to communicate with his friend Hisayo, whom he had met through an exchange program in high school, for the project. Being able to communicate with Hisayo for his academic work was a major motivation for Lucas (Pasfield-Neofitou, Reference Pasfield-Neofitou2012), while all of the participants interviewed by Takahama (Reference Takahama2010) opted to be paired by their teacher with a student from the Japanese university.
The instructions given to the five participating students were to maintain at least three exchanges over the period of approximately two months. The assigned topics for the email exchange were “youth culture (Japanese, Australian and/or of students’ home country including shopping, food and travel)” and “books students have recently read and/or films that they have recently seen”. The students in Australia had to write email messages in Japanese, and Japanese students had to reply in English. Each message was to be approximately 300 words/characters in length.
A unit evaluation questionnaire was undertaken in Japanese 3 classes at the end of semester one, 2009. In this, the students were asked to provide their opinions of the “J3 email project”. The set responses were as follows: the “J3 email project” was (1) very enjoyable and useful, (2) enjoyable and useful, (3) no opinion, (4) neither enjoyable nor useful, and (5) neither enjoyable nor useful at all. Approximately 81% of the students showed a positive response to the activity. They commented that the assessment task was “very educational”, “motivational”, “a good opportunity to establish a peer relationship/friendship with a student in Japan” and “to practice Japanese in a practical way”, “helpful to improve” their target language, “useful to realize Japanese people's perspective” regarding their youth culture, and so forth. In contrast, only 19% of the students judged the email activity negatively. The negative evaluation by the learners appeared to be mainly due to the lack of Japanese partners’ involvement in the activity and limited time to exchange emails by the submission date. As a result, the activity seemed to be stressful for some learners. Nevertheless, despite the above negative feedback, the majority of the students perceived the “J3 email project” as useful and enjoyable.
In semester one, 2010, a further questionnaire was undertaken in the Japanese 5 lecture. The purpose of this questionnaire was to gain insight into the continuation or non-continuation of learners’ email exchange with the aim of better understanding what contributed to the maintenance of such email contact. The number of questionnaires taken by the learners was 62, and the number returned was 15, or 24% of the total. Surprisingly, only two out of the 15 respondents maintained email interaction with their Japanese partners, the rest of the students being non-continuers. The relatively small number of continuers suggests conflicting evidence as to the positive outcomes of the “J3 email project”, which will be explored below.
3 Continuation/non-continuation factors
In the questionnaire, participants described their reasons for the continuation or non-continuation of their email exchange with their Japanese partners. The fifteen students’ reasons for continuing or not continuing are presented in Table 2 below. The main case-studies focussed upon here are referred to by pseudonym, while other respondents have been assigned letters.
Students’ Reasons for Continuation or Non-Continuation
The two continuers Cindy and Dina mentioned “socialization”, “friendship”, and “just for fun” in their comments. These words imply that a personal connection in the dyads could be one of the factors for promoting sustainability of email interaction. On the other hand, the thirteen non-continuers considered the reasons for the cessation of their interactions were, for instance, lack of interest in the dyads, a sudden cessation of interaction from their partner, delay in their partners’ responses, the assessed nature of the email exchange activity and their existence of Japanese friends prior to the “J3 email project”. Interestingly, no participant mentioned any issues concerning their target language use or their Japanese study. Even though students seemed to understand email interaction with NSs enables them to develop their Japanese skills, the purpose of improving their target language appears unlikely to be sufficient to maintain their email communication. Findings from the email and interview data are discussed in the following sections to develop more in-depth insights into the factors which contributed to learners’ continuation or non-continuation.
3.1 Differences in participants’ sojourn experience and existing social networks
Analysis of the data showed that a difference in participants’ sojourn experience and their existing social networks are some of the factors affecting continuation of email interactions. Two non-continuers, Adam and Beck, had in-country experience in Japan, whereas the two continuers, Cindy and Dina, did not have any previous sojourn experiences at all. According to Stockwell and Levy (Reference Stockwell and Levy2001), the participants without sojourn experiences had a tendency to create more email exchanges than those with this experience. Furthermore, the participants’ existing Japanese friends prior to the “J3 email project” also appeared to be a contributing factor in maintaining the email interactions. Adam, for instance, had hosted two Japanese exchange students in 2006 and 2007, and still interacted with them on the internet. He also mentioned that he had met several Japanese people and become friends with them during his stay in Japan, whom he also still communicates with via the social networking site Facebook.
Similarly, Beck corresponded with her homestay host family for approximately one year after her two-week stay in Japan in 2004. She also became friends with a Japanese exchange student at her university in semester one, 2008, with whom she keeps in contact on the internet. Further, since she is a member of a Kendo club at her university, she meets Japanese members regularly. In contrast, the continuers seem not to have had a substantial number of contacts with Japanese people in their social networks, either online or offline. For instance, Dina had only one Japanese friend who she had met in high school in Australia when the latter was on an exchange program. Dina still has interaction with the Japanese student on Facebook. Cindy did not have any interactions either online or offline with Japanese nationals prior to the email project.
Lucas’ case is interesting to observe in that, like the two non-continuers, Adam and Beck, he had in-country experience in Japan, and his established friendships similarly made him uninterested in being paired with a student from the Japanese university. However, unlike Adam and Beck, Lucas chose to communicate as part of the project with one of his friends with whom he had an established relationship. Yet, once the project was over, he and Hisayo immediately ceased emailing and returned to their favoured medium of communication – Facebook.
Ware (Reference Ware2005) argues that learners with more opportunities to engage with internet communication might perceive email partnerships as an expansion of their current mundane activities. In contrast, those with fewer experiences view email as a beneficial tool for interacting with NSs and building connections with them. This discussion can be applied to language learners’ social networks in general, with or without existing relationships with NSs. The non-continuers in this study already had interactions with NSs prior to the “J3 email project”. This might have encouraged them to view their partnership in the email activity as an ordinary activity or additional workload. On the other hand, due to their lack of interaction with Japanese people in their existing social networks, the continuers considered the project as a significant opportunity to communicate with NSs. This suggests that the existence of Japanese nationals in students’ established social networks also appeared to be an influential factor in encouraging email interactions. Hence, the differences in participants’ sojourn experiences and their existing social networks can be considered factors for the promotion or otherwise of language learners’ email interactions.
Like Adam and Beck, Lucas too had sojourn experience in Japan, and an established network of Japanese contacts with whom he communicated via the internet. However, as he opted to participate in the project with Hisayo, one of his established contacts, his participation in the email project did not represent any additional workload for him. Prior to participating in the email project, Lucas composed most of his emails to Hisayo mainly in English (6/9).The three emails collected that were written mostly in Japanese, as displayed in the table below, were considerably shorter than the three emails he composed during the email project (approximately half the length):
Lucas’ Japanese emails to Hisayo before and during email project
When asked if he noticed any differences between the emails he generally wrote to Hisayo, and those written specifically for the project, Lucas commented “it was quite a challenge to get to 300 characters, because I'd never written so much before. It was actually good, because my exam's coming up, and I assume we're gonna have to write a lot, so [it was] good practice”. Writing these more extended pieces took Lucas between 30 minutes and an hour at a time, particularly because he used an online dictionary and his textbook index to look up grammar points more frequently than he would have if he was communicating for non-assessed purposes. Lucas stated that participating in the project was a “bittersweet” experience, because while it was “good catching up with Hisayo again”, he did not enjoy “the fact that we were forced to do it, and the 300 word [count]”. Lucas said,
I enjoy catching up with Hisayo in emails, but just for the sake of leisure. Instead, we were forced to do it for assessment. With this email project, I sort of felt pressured to ask her questions and all, which is not something I'm comfortable with. I had to look things up and stuff, but if I was emailing her on a casual, non-assessment task basis, I would have felt better.
3.2 Participants’ different perceptions of the curriculum task
The continuers and non-continuers in this study were likely to perceive the “J3 email project” differently. Even though the non-continuers considered the email activity a beneficial experience for language learning, they also developed a certain degree of stress, as even Lucas expressed above, as a result of this task. This is because the email exchange activity was a compulsory task which had to be submitted to their teachers for assessment purposes. However, the continuers viewed the email exchange task as more of an opportunity to meet Japanese people and to make friends. Similarly, Ware (Reference Ware2005) argues that some of her participants were overwhelmingly concerned with the assignment aspects of the email activity, rather than building friendships with Japanese speakers, while other students showed no interest in marks, but in constructing interpersonal relationships, findings that are reflected in the present study also.
Moreover, Ware (Reference Ware2005) also argues that a difference in participants’ perceptions can also be prompted by social factors. Her American participants, for instance, explained their motivation for German study was simply for their intrinsic enjoyment. On the other hand, her German participants viewed English as necessary for obtaining a better job. Similar to Ware's findings, in this research, Dina, in particular, appeared to study Japanese to pursue her private interests. She showed a strong interest or willingness to work in Japan in several of her email messages. Her ultimate purpose for studying Japanese also appeared to encourage email interactions with her partner. These examples suggest that a difference in their perceptions towards the curriculum task as well as their target language itself can be influential in the students’ maintenance of their email interactions.
Finally, Lucas and Hisayo appear to have shared the perception of their communication during this period as primarily fulfilling a curriculum task for Lucas’ assessment, and drafting and feedback were an important part of their communication for the duration of the email project. As previously described, Lucas stated that he viewed the project as primarily “for assessment”, and made more frequent use of online dictionaries and his textbook than he would have otherwise. Furthermore, he and Hisayo engaged in peer feedback, editing one another's emails and sending them back, although Lucas decided not to incorporate any of Hisayo's feedback into the emails he submitted for assessment, as he viewed this as “cheating”. Lucas’ view of his communication with Hisayo during the project as primarily for the purposes of assessment is evidenced by his careful drafting, his switch to the teacher-prescribed tool (email rather than Facebook) and even his avoidance of utilising Hisayo's feedback due to his perception that it would constitute academic misconduct. Other evidence which suggests that Lucas viewed his communication with Hisayo during this period as primarily a curriculum task includes the explicit references that he made to the project, including adding comments such as PS 
$$$$
(PS this email is for the project!), or even instructing Hisayo on how to respond to emails. In one case, Lucas prefaced an email with “Please pretend this was sent on your birthday! Sumimasen!” (sorry), further highlighting the artificiality of their communication. He had already carried out his own, personal, congratulations to Hisayo on the occasion of her birthday, sending her a card through the mail, however decided to email Hisayo on this topic in order to fulfil one of the requirements of the assignment, which required students to ask their Japanese ePals about an aspect of Japanese culture. Lucas made use of this opportunity to ask Hisayo what Japanese people do to celebrate birthdays. Stockwell and Levy (Reference Stockwell and Levy2001) highlight the importance of the topic of emails as a key factor influencing the sustainability of such exchanges, and while suggested topics may indeed facilitate communication between previously unknown interlocutors, in the case of contacts with an established relationship, such assessment requirements may actually decrease the authenticity of interaction. Finally, at the conclusion of the project, Lucas and Hisayo, who already had an established relationship of two years, continued their frequent communication, yet as previously mentioned, significantly, they chose to do so through the medium of Facebook, not email.
3.3 Participants’ different attitudes towards their email exchange partners
Ware (Reference Ware2005) argues that a difference in interactional purpose can influence learners’ engagement in email communication with their language partners. In Ware's study, some participants perceived their online partners more as language tutors who could provide feedback on their target language, not as potential online friends. That is, they tried to engage in the type of activity that Lucas and Hisayo did, without first establishing a friendship.
Adam viewed his partner as “just a guy from a project”. He commented that because the partnership in the email activity was randomly set up and the interaction was only through email, it was difficult for him to develop further connections with the Japanese student. Dina also perceived her partner, Yuki, with whom she did not continue to interact, as “more of a project partner”. Dina found developing discussion beside the assigned topics was difficult because her partner's email messages appeared to lack enthusiasm, which made Dina feel uncomfortable to discuss anything outside the set project topics. As previously mentioned, Stockwell and Levy (Reference Stockwell and Levy2001) have identified the importance of email topics in the continuation of communication, as will be further discussed below; however, rather than seeking to stimulate interest, the non-continuers attempted to be as polite as possible so as not to offend their partners. As a result, they discussed only safe topics which made their discussion non-personal. On the other hand, the continuers viewed the NSs as friends rather than just project partners. Cindy, for instance, commented that she was more concerned with building a friendship than having a personal language tutor. She ventured to share her thoughts and the events which happened in her life with her partner. Dina also mentioned that Kanako, her continuing partner, is more sociable. This encouraged her to develop conversations outside the assigned discussion topics. Overall, the non-continuers in this research were not able to perceive their partners as possible friends whereas the continuers sought to connect personally with the Japanese students via email. Such attitudes might have impacted the building of friendships between the Japanese learners and their partners.
3.4 Topics and shared interests in dyads
How language learners are able to broaden discussion topics appeared to be one of the key factors for sustaining their email interactions. Stockwell and Levy (Reference Stockwell and Levy2001) argued that their participants who discussed topics outside the set topics sustained their email interactions for an extended period, while those who developed a smaller range of topics were less likely to sustain their email interactions. In the present study, the continuers sought to develop two-way communication with their partners by sharing various things that happened in their everyday life. Because they explored many kinds of individualized topics, they could discover shared interests and establish more connections with their partners. Cindy and Dina commented that they had no difficulty in finding topics as their partners made a mutual effort to develop conversations by suggesting topics. Cindy also stated that she intentionally “tried to find something in common” with her partner by asking questions and introducing new topics. However, the non-continuers were less likely to establish individualized topics in their email interactions. Adam intentionally avoided asking questions which might offend his partner. This is because he viewed his partner as “just a project partner” and did not know the partner's opinions. Thus, he did not explore new topics further in email interactions but stayed in the safer area of the set topics.
Moreover, although sometimes the non-continuers did introduce new topics, their partners did not share their interests in these topics or vice versa. For instance, Adam introduced an English idiom, “hop on the band wagon”, in his email message; however, his partner only said in the reply that he did not know the saying and did not ask Adam for further explanation or, from Adam's perspective, show any interest. In another case, Adam's partner mentioned his interest in baseball, but Adam showed no interest either. He simply replied he neither played baseball nor knew the rules as the sport is not popular in Australia. Then he quickly changed the topic. Thus, the non-continuers’ low expectations of the partnership and their lack of interest prevented them from enhancing their email communication. These examples indicate that sharing individualized topics and common interests plays a significant role in sustaining email interactions.
3.5 Time issues
A difference in the learners’ use of time can also be considered as an influential factor. The participants in Ware's study (2005) stated that reading and replying to their partners’ messages was time-consuming, a factor which may be related to the learner's language proficiency and/or computer skills, two key considerations highlighted by Stockwell and Levy (Reference Stockwell and Levy2001). Students thus perceived these reading and writing activities in their target language as an academic assignment rather than communicative behaviours. As a result, such an ‘unpleasant’ task demotivated learners’ interactions with their partners. Likewise, the non-continuers in this research found the email activity time-consuming. Adam and Beck commented that they had to spend one or two hours composing each email message due to their limited knowledge of Japanese. Hence, when they became busier with their private life or workload at university, they felt stressed under the time pressure.
Lucas, who participated in the project with his Japanese friend Hisayo, was used to communicating over the internet with her in Japanese; however, he too found participation in the project time consuming. While the average length of Lucas’ usual emails to Hisayo was 233 characters, the minimum required for the project (300 characters), as well as the fact that it was an assessed task (and hence Lucas spent more time drafting, using online dictionaries and drawing on his textbook more frequently, as seen above) meant that the time commitment required to write a project email was substantial. He reported that it took between half an hour to an hour to compose each email for the project, a factor which may explain his switch back to their norm of Facebook communications with Hisayo, which were generally short in nature, after the completion of the project.
On the other hand, Cindy and Dina spent only 20 or 30 minutes to compose each email message. This time saving enabled them to exchange email messages more often. Cindy and Dina explained that they usually exchanged emails with their partners once a week, but in case of urgency (e.g., when the continuers were asked to answer questions by their partners for interview assignments), they replied within a day or two. The continuers had more active interactions with their partners which consequently promoted the continuation of their email communication.
3.6 A difference in the participants’ desires to continue their email interactions
The continuation and non-continuation of email correspondence may most obviously be related to differences in the participants’ and their partners’ desire to maintain their communication. In Cindy and Dina's email message for the last exchange in the “J3 email project”, they showed their explicit desire to maintain their email exchanges. For instance, Cindy wrote 
$$$$
$$$$
(Let's continue to exchange email messages sometimes if it is okay with you.) However, Adam and Beck did not mention their willingness to continue to correspond. Adam's partner, Takashi, showed his gratitude to Adam for the correspondence and stated that he wished he could meet Adam one day in the future. Takashi then closed his email with stating “bye until then”, which Adam considered the end of their email interaction. This may indicate that both sides of the pair may have made a polite final salutation to perhaps see each other in the future, but it was agreed upon to give up correspondence. Conversely, Beck stated that her partner just stopped writing back to her without any signs of closure.
For Lucas, participating in the email project represented somewhat of an anomaly in his and Hisayo's usual communication patterns. In the past, the pair had communicated primarily via email, with occasional MSN chat conversations, for almost two years. However, around six months prior to taking part in the email project, they began using Facebook as their primary means of communication. Thus, their participation in the email project was a break from this pattern of use. During the email project, their communication on Facebook ceased; however, as soon as Lucas had completed the requirements for his assignment, they returned to Facebook communication.
3.7 An additional factor for non-continuation: face threatening behaviour
An additional factor which appeared to affect non-continuation of email exchange is face threatening behaviour. During the course of email exchange, Dina asked Yuki, her non-continuing partner, to answer the questions for her Japanese assignment. Because Dina did not receive a reply from Yuki, she had to ask a different person to help with her task.
After Dina completed her assignment, she finally received an email message from Yuki with an apology for the delay. Dina replied in Japanese with a sincere apology for she had already found another person to help with her assignment, and gratitude for Yuki's help. Extract 1 shows Dina's bilingual email message. In the email message it can also be seen, however, that Dina clearly stated in Japanese that her partner's reply was late. In the Japanese context, this could be considered overly direct and might offend or upset a Japanese reader. Yuki may possibly have felt that her sense of face was threatened. Even though Dina showed her willingness to maintain the contact in the English part of the message, Yuki never replied to Dina. This suggests that email messages may include face threatening acts which might cause the cessation of email interactions.
Extract 1: Dina's bilingual email to Yuki
$$$$
$$$$
$$$$
(Dear Yuki, it's alright, but your reply was a little late)
$$$$
(I had another person for the interview)
$$$$
(Sorry but thank you for your reply)
$$$$
$$$$
(I am happy. From Dina)
Hi Yuki, thanks for your reply even though it was a little late, I interviewed another person already sorry… but I'm really grateful for your email. Thank you a lot^^, hope your studies and English is going well. Hope we continue to communicate as well, if you ever need help with English just ask me :) Dina
3.8 Additional factors for continuation
A final factor for the continuation of email exchanges appeared to be the participants’ efforts to present email messages that were enjoyable to read. From a relatively early stage of interaction, the continuers attempted to convey a friendlier and lighter mood to their readers by utilising code-switches, innovative language and informal speech. Cindy commented that she wanted to depict herself as a friendly and generous person in the message, and to show how she was excited to exchange email messages with her partner. These rapport building moves in the presentation of email messages are likely to help the participants sustain their email interactions. Another factor for sustaining email communication was the visit by the participant's partner. The communication between Cindy and her partner increased significantly in frequency when she learnt that her partner would be visiting Melbourne. They continued to maintain their frequent email exchange and made plans to meet in Melbourne. Therefore, it can be seen that if partners plan to visit each other, email correspondence is more likely to be continued.
4 Conclusion
This study sought to determine whether or not language learners continued interacting with their partner after a curriculum activity finished, and to identify what factors are influential to the continuation or non-continuation of learners’ interactions. While this study took email as an example, the findings may be applicable to projects using other forms of CMC. The results suggest that although the learners understood the email exchange with NSs as beneficial, the number of continuers was relatively small. Further, the study identified a number of factors which appeared to have an effect on the learners’ continuation or non-continuation of these interactions. These are a difference in learners’ sojourn experience and existing social network, expectations of the email curriculum task, attitude towards their partner, topics and shared interests, time issues, explicitly mentioning a desire to continue, a face-threatening issue, and the presentation of messages. Thus, this study clearly showed that the purpose of improving the learners’ target language is likely to be insufficient for them to continue interaction. In the same way that language learners need skills to manage face-to-face interactions, they need to be equipped with social skills for intercultural CMC interactions.
4.1 Limitations
Due to the nature of the focus of our analysis (communication outside of the classroom, after the completion of a formal email project), our sample size was limited, and as a result, many of the findings in the present study should be viewed as tentative. Further research exploring the long-term effects of email and other CMC projects needs to be undertaken in order to determine the potential benefits of such activities beyond their immediate application for classwork and assessment purposes.
4.2 Implications
In spite of the above limitations, we feel that in order to enhance and sustain students’ learning opportunities and experiences through email, the present research can suggest an instructive framework of reference for conducting an email exchange activity. Kötter (Reference Kötter2003) points out that guidelines which demand participants utilize only their second language prevent them from increasing opportunities to receive meaningful input in their target language. Therefore, the instructions should allow the learners and their email partners to negotiate what language to use and when to utilize it in consideration of the relevant balance as a whole in each email message. Further, the empirical studies by John and Cash (Reference John and Cash1995), Ushioda (Reference Ushioda2000), Absalom and Marden (Reference Absalom and Marden2004) and Marden (Reference Marden2007) have shown that their participants improve their target language by monitoring and reusing the NSs’ natural and practical use of the language, rather than receiving corrective feedback from their partners. This fact supports the above-mentioned importance of writing and receiving email messages in a combination of the students’ target and native language to make the interaction advantageous to both interlocutors in a dyad. In addition, this copy and paste strategy might help learners to promote the frequency of their email interactions. This study has found that non-continuers spent one or two hours to compose an email message which appeared to discourage them from maintaining their interactions. However, if the learners are equipped with this strategy, they could possibly save time composing written discourse. Consequently, this could help learners to sustain their email interactions.
In addition to addressing the balance of language selection, project guidelines should provide learners with the opportunity to consider the sociolinguistic aspects of their target language by letting students negotiate individual decisions as to what style of language to adopt, and when and where to use it. This allows them to utilize not only classroom language but also informal speech in their second language. This might be challenging for learners but can provide them with opportunities to learn a different aspect of the language usage and could possibly make their communication more interpersonal and enjoyable. In the present study, the lack of personal connection in dyads seems to be a crucial factor which is associated with sustaining email interactions. In order to help learners establish connections with their partner, applications of multiple CMCs, such as Skype, chat, and social networking sites, could possibly trigger active interactions between the interlocutors in the dyads. For example, after the first few email exchanges, they virtually meet face-to-face on Skype for discussion. The case of Lucas and Hisayo, who switched back to using Facebook and MSN to communicate as soon as the project was complete, demonstrates the importance of allowing students to utilize their preferred communication media in order to maintain motivation. This utilisation of different forms of CMC might also support learners in building stronger personal connections with their partners. In the case of students like Lucas, allowing those students with established friendships to continue to communicate with that friend via their established medium(s) may prove beneficial.
Aitsiselmi (Reference Aitsiselmi1999) suggests that the teachers should consider their role as facilitator or adviser in order to support learners’ participation of the email activity during the course of the event. Hence, before the learners commence their email interactions, the teachers could invite the learners to discuss various features with regard to intercultural email interactions. For instance, teachers could introduce students to some successful or unsuccessful cases of email interactions in the past. The case studies could help students how to develop individualized topics, avoid face-threatening issues, present messages in an enjoyable way and so forth in order to promote their interaction. Lucas and Hisayo's case, for example, shows how topics assigned by a teacher may be helpful if one is communicating with a previously unknown interlocutor, but may feel artificial if there is an established relationship. Also, teachers need to monitor students’ activity and give guidance to those in need during the course of the exchange. The above list of possible implications is far from exhaustive. However, what is essential is that language learners and teachers be aware of the factors for promoting continuation of interactions, in order to enhance their CMC partnerships.
