This paper reports some of the findings of a distinctive innovative use of audio-conferencing involving a population (campus-based advanced learners) and a type of application (task-based language learning) that have received little attention to date: the use of Wimba Voice Tools to provide additional opportunities for spoken interactions between advanced learners of French. The experiment had a dual aim: (a) to examine the suitability of Wimba Voice Tools as an environment for sustained interactive talk, and (b) to study the nature of interactions between advanced learners, with particular reference to the processes supporting collaborative activity.

After a brief summary of the rationale and main characteristics of the experiment, the paper focuses on the strategies used by three non-native speaker (NNS) dyads to resolve language problems as they worked on a set of four tasks. Extending the classical model of negotiation for meaning to cover other instances of language-related episodes identified through discourse analysis of the empirical data, the study offers a detailed account of the incidence and nature of negotiated interaction and collaboration between partners. This leads to a discussion covering the impact of functionalities, scaffolding and task effects. The paper ends with some suggestions for future research.

This paper has a dual aim: to situate functionalities among the complex of factors that help shape online interactions and to explore the heterogeneity of audio conferencing and its implications. Following a critical discursive synthesis of the treatment of variables in the literature, attention focuses on the distinctive and diverse characteristics of synchronous audio environments. The existence of a synchronous-asynchronous dichotomy is questioned and the effects of a number of differences are highlighted. As a further illustration, the paper next undertakes a comparison of Lyceum and Wimba Voice Direct. The main features of the two environments are introduced and a detailed analysis of their text and voice chat facilities is carried out. In the last section the benefits of text-based CMC are revisited through the lens of voice chat technology. It is concluded that the benefits of CMC appear to have been overstated and that no general claims can be made about voice CMC, but that collectively, well documented studies could provide a useful source of information on the level of contingency of particular findings.