Embedded reference: Translocating gestures in video-mediated interaction

Audio-visual technologies can enable informal communication akin to face-to-face interaction. However, they prove less successful when deployed to support work and organisational activities. This is, in part, due to the limited ways such technologies provide access to the materials, objects, documents and the like, that are critical for supporting work activities as they emerge and unfold. What is often neglected in these new technologies is a consideration of how objects are referred to, manipulated and transformed within and through interactions between colleagues. In this paper, we consider an advanced prototype system called t-Room that seeks to provide geographically dispersed participants with rich and varied access to physical and digital documents. This technology has been designed to support flexible collaborative activities with and around objects, as if remote participants and materials in their local environment were co-present within a common space. By undertaking quasinaturalistic experiments in this prototype environment we reveal that at times participants could unproblematically refer to detailed features of the environments and when there were difficulties resolve them. We notice, however, that at other times participants had great difficulties in assessing the relationships between themselves, their remote colleagues and objects in the environment; the very flexibility of the technology introducing instabilities into the sequential accomplishment of referential activities. By considering examples of this technology in use, we suggest that these limitations may reflect wider issues concerned with our understanding of how interactional activities are embedded within the local environment