With the increased availability of networked computers in every walk of life in different parts of the world, from classroom to public library, from home to workplace or Internet café, the opportunities provided for learning online are expanding exponentially. For language learners, the possibility of accessing authentic L2 materials and "meeting" native speakers of that language in a virtual environment has never been greater—or easier or faster. And yet, for many language learners with access to the appropriate technology, these opportunities are not being optimally exploited. The potential of the Internet to allow learners to communicate not with their machines but with other, real people, both native and non-native speakers of the target language, is too often ignored. Instead, they sit at their computers and use "traditional" CALL packages to interact with machine rather than with flesh-and-blood; if they access the Web at all, it is to participate in similar activities there. The role of the computer is that of an electronic "teacher" rather than as a tool to increase participation in and knowledge of different real (and virtual) communities.

Admittedly, the use of asynchronous, text-based computer-mediated communications (CMC) technologies in the language classroom has become increasingly popular in recent years, and there are many examples of email tandem exchanges being used successfully with both young learners and at university level (Schwienhorst 1997a; Vilmi, 1997; Warschauer, 1998). Bulletin boards and discussion lists, too, have been used as tools to promote reflective aspects of language learning (Goodfellow & Lamy, 1998; Warschauer & Lepeintre, 1997). Less frequently-used for language-learning purposes, however, are synchronous, text-based CMC tools, such as IRC, Webchat and MOO, although researchers have for some time been investigating the nature of the discourse generated within such environments (Haase et al., 1997; Krause, 1997; Negretti 1999), tending to concentrate on the "chat" aspect of synchronous text-based environments rather than on the types of learning activities which might optimally promote autonomous language learning within such virtual learning environments (VLEs).

In contrast, in this paper, we will concentrate on the pedagogical aspects of one type of synchronous text-based tool, MOO (Multi-User Domain, Object-Oriented), examining how it may be employed to support and develop both the principles of constructivism and autonomous (language) learning strategies. We will present examples from our own experience of using MOO professionally and with language learners, and outline the ways in which we are developing (language learning) tasks within a framework designed to support autonomous and collaborative language learning.

Text-based CMC can be broadly divided into two major categories, each of which may be further subdivided according to communication routes and tools available to the user (Table 1).

As already noted, asynchronous tools have been widely used by language learners and language teaching professionals (Vilmi, 1997) and by researchers investigating the potential of such tools to support (reflective) language learning strategies (Goodfellow & Lamy, 1999). The effect on learner confidence and spoken production of both synchronous and asynchronous text exchanges to support language learning has also been investigated on several occasions; Chun (1994) found that language learners who used synchronous text-chat between scheduled face-to-face sessions became more confident about speaking the target language and Stevens and Hewer (1998) claim that confidence in spoken production L2 was improved by providing groups of distance language learners with access to email between scheduled telephone conferences. It has also been shown that shy learners tend to make greater contributions to discussions and learning events in the anonymous environment of text-based CMC than in face-to-face situations (Turbee 1995, 1996, 1997; Turner, 1996; Warschauer et al, 1994) and that they report increased confidence in using the target language in such environments (Healy Beauvois, 1998).

Clearly, then, CMC can be of benefit to language learners in several ways; asynchronous tools promote metacognitive learning strategies and synchronous tools cognitive learning strategies, while learner confidence in L2 production appears to increase with the use of both synchronous and asynchronous CMC tools. Among such VLEs, MOO offers the following features (see Table 2 for details):

· All in-MOO activity may be recorded or "logged". Learners can access such logs later, reflecting on the content in terms of vocabulary, structures and communication strategies, for example.

· Users can finish a comment/post without being interrupted, a benefit also present in asynchronous fora.

· Empowering learners by allowing them to make contributions in the knowledge that they will not be interrupted can lead to increased confidence both socially and in terms of using L2.

· Scrolling back through text on the screen allows learners are able to consider their responses, even in a synchronous environment.

· Learners can use research tools to find information pertinent to a discussion without interrupting the discourse. In face to face interaction, consulting a dictionary or an encyclopaedia would probably halt conversation; in synchronous, text-based environments, this does not happen, since users become adept at managing multi-threaded discussions and can scroll back through a discussion if they miss a comment).

· Engaging in multi-threaded discussions also lends itself to using metacognitive strategies in real-time: this would be impossible in a face-to-face encounters, since such behaviour would interrupt discourse between interlocutors and be seen as socially inept, but can be carried out in a MOO without invoking either social or practical difficulties. A player can engage in discussion with multiple interlocutors without interrupting the flow of discourse with any of them.

· Text-based discussions tend to be slower than in face-to-face— or even telephone — situations. Contributions to a MOO discussion can, therefore, be more reflective than in an oral exchange. At the same time, synchronous communication does require responses in real time: Learners report that on a MOO there is just a bit more time to compose thoughts than there is when speaking, but not much more, leaving them no time to do any mental translating (Turbee, 1996).

· Real time exchanges provide learners with immediate feedback on their performance in L2 - if the effect of the communication is not what was intended, then the communication was unsuccessful, and the learner will have the opportunity to rephrase that communication. Further, keeping a log of the exchange allows the learner to return to that exchange later and to reflect on why it was/was not a successful communicative event.

· Its relative anonymity. Shyer learners to contribute to discussions on an equal footing with their more extrovert peers.

· While Groves (1997) reports that some learners believe that "he who types fastest, shouts loudest", it is probably true to say that, in theory at least, all MOO users can make themselves "heard" because all contributions—even if these are made simultaneously—appear on screen in a linear fashion.

· In spite of concerns about fast typists dominating the discussion, long "speeches" take time to enter via a keyboard. Synchronous exchanges, therefore, tend to comprise short entries. Learners are encouraged to reflect on their contribution before making it.

· Learners with disabilities are empowered by virtue of the anonymous environment. Judgements—whether conscious or unconscious, about participants cannot be made as a result of their appearance. Further, visually-impaired learners can participate fully in all exchanges by using "screen-reading" software, and aurally-impaired learners are equally empowered by being able to read text on the screen as it appears. Finally, speech impairments are not noticeable in the text-based environment of MOO.

· Depending on the theme of the MOO, or even on the popularity of the MOO itself, the chance of meeting another user is quite high, since the VLE is available 24 hours a day, 7 days a week. If no other user is online, learners can even interact with the environment.

At this point, we should, perhaps, clarify that by "autonomous learning", we do not mean "teacherless learning". Like Little (1991), we believe that to equate learner autonomy with self-instruction is a common fallacy. In fact, as Little remarks, learner autonomy is "…a capacity – for detachment, critical reflection, decision-making and independent action." (1991, p. 4), and Schwienhorst (1998) points out that autonomous learning "… contains the idea that learning arises essentially from supported performance…" while Dam (1990, p. 17) emphasises the "socially responsible" nature of the autonomous learner. Autonomous learners, then, take responsibility for their own learning and for group learning. They negotiate with each other and with their tutors about what is to be learned, when and how. The role of the tutor is that of guide and co-learner, for autonomous learners may choose to construct their own learning in such a way that the tutor may also become a learner. Our view of autonomous learning is closely related to constructivist educational theory, where learners construct meaning and take responsibility for their own learning.

Like other CMC environments, MOO allows learners to login at times convenient to themselves. Such "24/7" technologies are excellent vehicles to promote autonomous learning strategies since they offer learners the opportunity to connect at any time, whether or not their tutor is available. As we have seen, however, MOO combines both synchronous and asynchronous tools within a single VLE and provides users with the opportunity to extend and fashion the environment according to their own desires, thus allowing them to participate in learning experiences which are meaningful to them and so motivating and engaging (both Shneiderman,1997, and Schwienhorst,1997b, point out the importance of these factors to a successful learning experience). MOO also allows learners to work individually or colla-boratively in customisable virtual workspaces which exist even when the occupiers are not online.

With all these points in mind, then, we have developed learning activities to take advantage of the features of MOO which promote constructivist principles and autonomous learning strategies.

We have already been involved in developing MOO-based language learning activities within a framework intended to promote autonomous language learning strategies, and on the basis of what we have learned from our experience, we are currently designing further learning activities using the MOO environment as a base for learners. In the remainder of this paper, we will outline the results of our first venture into MOO-activity design, describing the successes and pitfalls. We will show how the lessons we have learned have fed into our current MOO-based work and outline future learning activities.

The Bamburgh project (named after the area in the North-East of England represented by the MOOspace developed for the project) attempted to address the principles of constructivist learning summarised by Brooks and Brooks (1993). The ways in which this was done are illustrated in Table 3:

Table 3-1

Table 3-2

The outcomes of our initial investigations: some reflections and guidelines.
Table 4

The Bamburgh project was run twice in 1998 with two sets of participants (see Table 4 above). For both trials, participants negotiated to meet at a scheduled time once a week. Learners would be able to meet with each other at any other time they felt to be appropriate. It was made clear to all participants that there were no lesson plans and that they should negotiate among themselves about how they wished to use the project area to achieve the objective of writing a collaboratively researched, designed and produced web-page about any aspect of the project.

It soon became apparent that the first trial was not progressing as had been intended. Discussions with participants revealed that although the designers had attempted to develop a supportive and non-threatening VLE which included synchronous and asynchronous tools, links to a supporting website and accessible, in-MOO support materials, several other factors, which had not been taken into account, had a negative effect on the trial.

Bearing these factors in mind, a second trial was run with learners from Australia and Japan, whose timezones were only one hour apart. This time, learners experienced few technical difficulties and socialised with each other between scheduled meetings, thus building their own learning communities. Although, initially, each group of learners had concerns about working with the other group (see Table 3) they found that they were able to help each other with different aspects of the project during its course, and, so, were well-motivated to succeed.

As a result of our initial experiences of designing MOO-based projects, we have been able to developed a set of guidelines to aid such projects (see Table 5 below).

Table 5

We are currently developing projects for use within two different MOO environments. The first of these will be based at GrassRoots MOO and will involve groups of learners from around the world. In order to avoid the problems experienced during the "Bamburgh" project as a result from participants' lack of MOO-experience, before starting the main project activity, learners will be involved in two "warm-up" activities in order to a) provide them with the skills necessary to function effectively in a MOO environment and b) foster a sense of community.

The main project activity will involve learners in working collaboratively to extend the MOO to incorporate artefacts of interest to their group. In order to do this, they will use the MOO as an environment in which to a) negotiate with each other, b) research the topic they choose, c) build objects, d) draft and redraft descriptions , and e) design a supporting website.

A second investigation will take place at Dreistadt MOO, a German language learning MOO, where we propose to run two separate projects.

The first will involve L2 learners of German in Brazil and native German speakers in Germany. There are considerable remnants of German immigrant communities in southern Brazil. The fifth generation of these immigrants is now attempting to recover its cultural roots and language. Many of these immigrant communities originally came from a region in Germany called Hunsrück. The intended activity is to pair classes of L2 learners of German in Brazil with German students in the appropriate area of Germany. These tandem groups will research and document their common heritage together. The MOO will be not only their meeting place, but they will also build examples of their culture and heritage in MOO, thus creating a "virtual Hunsrück". There will also be Web pages to include multimedia elements. Dreistadt MOO has a web interface that allows to direct

connection to web-based graphics, sounds or videos, as well as to web-based texts. Anticipated outcomes are that the German participants will learn about their past and why their parents had to emigrate, the Turkish immigrant children in the German classes will input their experience in a way similar to the experience of the German immigrants in Brazil, and participants will be involved not only in linguistic development but also in sharing and learning about each others' culture and experiences as immigrants.

In the second project, distance learners of Deutsch als Fremdsprache (DaF) who live throughout Europe will be paired with DaF learners based at a German university. The distance learners spend one week at this university for the residential component of their course. The activity will involve groups of DaF learners from Germany and learners of German at a distance working together to identify an aspect of the university city which interests them. They will then build a virtual representation of that aspect in Dreistadt MOO, using the multimedia capabilities offered by the environment. The expected outcomes are that the final stage of this project will see the distance learners visiting the city, meeting the DaF learners based in Germany and being able to compare the virtual representation they have built with the real thing. This experience will add an extra dimension to the reflection and review aspect of the project.

Collaboration on joint projects of mutual interest is a powerful force for learning new things and for sharing insights and experience with others. MOO offers the opportunity for learners to collaborate, in real-time and asynchronously, on projects which are meaningful to them. By approaching learning activities within MOO from a constructivist point of view, we can design a mutually supportive learning environment. This type of environment encourages the development of autonomous learning strategies through the provision of a flexible but firm activity framework which reflects "real-world" tasks. In addition, such an approach provides learners with a meaningful, thus motivating, purpose for participating in the learning event.

Engaging, MOO-based learning events provide an important shift of focus. They move from target language as main object to working together with native speakers in a setting close to realistic (or "authentic" interaction). L2 becomes a mere means of communication while the achievement of common goals in social interaction emerges as the main priority. Learners become involved in the activity to the point that they forget they are trying to learn a new language. It is at these moments when they probably actually learn most. What users can do or create together in their own imaginations then, would seem to be the only limitation encountered in using a MOO-based VLE for engaging and motivating learners. Careful design and planning of MOO-based activities can stimulate and reignite this imagination.

Dreistadt MOO

telnet 129.177.24.214:7777

Web Interface http://cmc.uib.no:7001/

Homepage http://cmc.uib.no/dreistadt/

telnet health.acor.org:888

Web Interface http://health.acor.org/grassroots/java/

Homepage http://www.enabling.org/grassroots/

telnet schmooze.hunter.cuny.edu:8888

Web Interface http://schmooze.hunter.cuny/test.html

Homepage http://schmooze.hunter.cuny.edu:8888