Interaction

The new technologies provide opportunities for learners to interact with native speakers at a distance through a variety of different online tools such as e-mail, chat, and discussion boards. The literature overall supports the use of CMC with language learners, an exception being a survey conducted by Corbel and Taylor (2003), in which a majority (56%) of teachers of adult immigrants felt that e-mail had low educational cost effectiveness. However, in a nondirected self-access situation, adult immigrants themselves chose e-mail in preference to all other applications, but their choice of language was primarily their home language (Lever, 2004), including accessing news from their home country. Left to their own devices, these students did not choose educational packages to increase their English language learning. Their choice mirrors computer use in noneducational settings, where e-mail has become the “killer app” (application).

Most commentators and researchers have noted the advantage of asynchronous CMC modes such as e-mail and discussion lists because they allow learners to interact in their own time and place. Savignon and Roithmeier (2004), for example, demonstrated how two classes, one a U.S. class learning German, the other a German class learning English, collaborated to produce a jointly-constructed CMC conversation on a bulletin board. These learners used “the potential of CMC to engage … in the interpretation, expression, and negotiation of meaning essential for the development of communicative competence” (p. 284). Similarly, Schwienhorst (2004) found that in NS/NNS interactions in MOO (text-based synchronous interaction), there was no significant difference in topic initiation between the two groups. However, as in much of the extant literature, studies in different contexts have produced differing results, as is often the case when research is context-dependent and not large scale. In an interchange via e-mail between students of Japanese in Australia and students of intercultural communication in Japan, students seldom tried to deal with breakdowns in communication (Stockwell, 2004).

However, synchronous CMC such as chat and teleconferencing have also been found to be effective in language teaching and learning, although because of the synchronous nature of this medium, there are overlapping turns, with little time to compose messages reflectively (see for example, Sengupta, 2001). Asynchronous CMC such as e-mail, because of its time delay, gives learners the opportunity to produce more syntactically complex language (Sotillo, 2000). Although, as Biesenbach-Lucas and Weasonforth (2001) have found, whereas e-mail and word processed texts shared similar cohesive features, e-mail texts were shorter and provided less initial orientation for the reader.

Because negotiation of meaning has been found to be necessary for language acquisition (see for example, Pica, 1994), several studies have examined whether in fact negotiation of meaning occurs in CMC and whether learner output increases. Most of this research has examined various forms of chat—either restricted chat, available only to the learners and their instructors, or chat in public chat rooms where learners can interact with native speakers. Sotillo (2000) found similar discourse features occurring in internet relay chat as in face-to-face interactions, with rapid interactions and negotiations among their ESL students; however, learners were often unfocussed, discussing issues not pertinent to the task. Learners of Japanese, interacting with native speakers via chat in a virtual world, on encountering difficulties in understanding each other, negotiated meaning (Toyoda & Harrison, 2002), as did learners of English (Smith, 2003). Similarly, learners of Italian, participating in public chat rooms “negotiate[d] for meaning and modif[ied] their interlanguage when engaged in open ended conversational tasks with unfamiliar interlocutors” (Tudini, 2003 p. 141). Most negotiations were the result of lexical or syntactic problems. In general then, the research on synchronous CMC has found it provides learners with opportunities for informal conversational practice and facilitates fluency, rather than accuracy. However, Smith (2003) also found that task type affected the extent of learner negotiations and that this effect was different from what has been identified as used in face-to-face interactions, prompting him to develop a new model for describing negotiated interaction via CMC.

Although most CMC research has examined written versions, recent developments in the technology have made voice CMC possible for language learning. Researching a distance education context, Wang (2004) found that using NetMeeting (a proprietary desktop videoconferencing system) supported learners' oral and visual interaction. With increasing availability of broadband in some areas and of the convergence of visual, audio, and text, research into technology use and L2 learning will need to investigate these new channels of communication.

Identity formation

“In the presentation of self online, one is not recognized by one's physical appearance, but through one's textual behaviors” (Wood & Smith, 2001, p. 55). In the online context, identity is formed through interaction, by how people create identities through CMC in their home language and the target language. As Turkle (1995) has noted, the Internet provides opportunities for multiple subjective identity, opportunities that Warschauer (2000) identified in his study of a Hawaiian language class. He demonstrated how several of the learners, through synchronous discussions, e-mail and the writing of projects on web pages, found symbolic value in Hawaiian language use on the Internet, “allowing them to say to themselves and to the world that they are Hawaiian and proud of it” (p. 11). For one of these learners, the online activities allowed her to express her Hawaiianness, which had been difficult in face-to-face encounters, not only because of her shyness, but also because of her fair coloring. Examining a Usenet forum popular with Chinese students in the United States, Bloch (2004) found them liberated in using English as they adopted a traditional Chinese rhetorical style, finding their collective argumentative voice in English. In contrast, Warschauer, El Said, and Zohry (2002) express concern that the use of English among a group of Egyptian professionals was encroaching on the more traditional use of classical Arabic for formal written communication. They found that a diglossic situation developed: the Egyptian professionals used English for formal CMC, but Egyptian Arabic for more informal uses, thus creating both a global and local identity. In a recent groundbreaking study of informal, noninstructed chat use, Lam (2004) found that the two teenage Chinese immigrants she studied developed new identities as bilingual speakers of English and Cantonese. In this chat room, participants used both languages, code-switching between them, using a variety of English constructed by the group that differed from standard American English. Her focal learners were more comfortable using English in the chat room than in face-to-face situations.

What these studies have in common is how the interaction of global and local identities give rise to hybridized language varieties in these increasingly glocal discourse contexts (Koutsogiannis & Mitsikopoulou, 2004). Just as these language learners choose among their available languages to communicate and indeed, develop hybridized forms, so too might learners want to choose among the available technology modes to interact. Almost two decades ago, with the much more primitive CMC tools then available, Murray (1988) found that highly technologically savvy workers chose among the available modes depending on the context. Recent research (Thorne, 2003b) on American students learning French and being required to use e-mail with a French keypal found that for some learners, e-mail was not a mode of choice for developing interpersonal relations with age-peers, which had been the purpose for the teacher assigning the keypal interactions. Some learners did not participate in e-mail, even though the e-mail part of the course was graded. For these learners, identity with age-peers is established through immediate, synchronous modes such as instant messaging (IM) or short message service (SMS), the latter of which has yet to be exploited in instructed language learning.

Collaboration

Second language acquisition research has shown that collaboration among learners facilitates language acquisition. Such advantages of collaboration have also been noted in online collaborative activities, such as project-based learning (Debski, 2000) or other task-based activities that require collaboration (Shield, Weininger, & Davies, 1999). Most supporting research has been of case studies in natural, not experimental, settings. Recently, however, a quasi-empirical study (Beatty & Nunan, 2004) of two groups of students, one working with an interface that was based on constructivist principles, the other on one that followed a behaviorist model, found that the constructivist interface did not lead to more collaboration, as hypothesized. The researchers note that, although hypertext provides learners with more choices and is said to lead to greater learner autonomy, in fact, learners may not have the necessary skills to navigate the choices productively, a finding similar to that of Toyoda (2001) for project-oriented CALL. Schwienhorst's analysis (2003) suggests that in keypal activities, in addition to computer literacy skills, activities need to be set up so that they are embedded in coursework, teachers need to explicitly teach learners to reflect, learners need to participate in the design of the activities, and time needs to be provided for reflection. In other words, collaboration is not a necessary consequence of CMC, but may need to be contextualized and scaffolded for learners.

Telecollaboration has been identified as “the application of global computer networks to foreign (and second) language learning and teaching in institutionalized settings” (Belz, 2003a, p. 2) and involves learners using CMC for a variety of interactions (for example, debate) with other learners distributed around the world. Within this broader theme, a focus on the sociocultural dimensions of such interactions has arisen, in addition to the more descriptive linguistic analyses of the language of interaction, the argument being that, not only does the distributed communication network of CMC facilitate access to native speaker language models and interaction, but to native culture. These studies (Belz, 2003b; Thorne, 2003a) have examined intercultural competence and found, among other things, that the medium provides an opportunity for cultures-of-use to co-evolve, just as other researchers have identified the evolution of hybrid discourses (see, for example, Lam, 2004).