Social Interaction in Shared Virtual Environments

We're now in the second of today's presentations by Ralph Schroeder from the Oxford Internet Institute, hosted by the Centre for Creative Industries and Innovation (they're certainly getting their money's worth!). This one shifts our focus considerably, to virtual environments.

Perhaps the obvious question here is how people interact in such environments. Ralph suggests that from current environments, it's already possible to forecast what shape future environments will take; he has created a model called the Connected Presence Cube to describe such environments.

Shared virtual environments combine presence, interaction, and co-presence (the sensory experience of being in a place other the one users are physically in, being able to interact with it, and being there with others - in other words, being there together). Ralph suggests that there are two distinct forms or end-states of VE technology here: fully immersive, virtual, computer-generated spaces on the one hand, and the holodeck on the other (where the image and real environment of the user is captured and transferred into the virtual environment).

Both these end-states, and technologies which work towards them (mobile phones, 2D spaces, etc.), can be theorised from a general point of view. In doing this, it's important to realise that many preconceptions about 'being there together' are misleading (for example, that none of them can be as good as face-to-face contact). Further, we already know these end-states of virtual environment technologies, and it is therefore useful for technology development in the interim to work backwards from these intended end states.

On this basis, Ralph suggests a theoretical model which he describes as the Connected Presence Cube. The three dimensions of this cube are presence, co-presence, and interaction (the landline telephone, for example, rate highly in co-presence only, single-person simulators rate highly in presence, etc.). A Swiss immersive teleconferencing system called Blue-C, for example, could be described as the holodeck-style, video-captured, fully immersive end-state; CAVE-style systems are the end-state of the computer-generated branch of shared virtual environment development - both would rate highly on all three axes of the Connected Presence Cube. As further examples, Ralph now runs us through a number of examples of virtual environments, from Activeworlds to Virtual Traveler and through to World of Warcraft.

It's possible to measure how much people feel that they're really there in the environment (have a sense of presence) - and immersive CAVE-style environments usually rate better than those only displayed on a computer screen. At the same time, how they interact with the environment also affects their sense of presence - which makes the key factors determining a sense of presence for any one space more difficult to determine. Co-presence can similarly be measured (and in hybrid situations, immersed users usually take leadership roles when compared to users who are accessing the space through desktop PCs, for example - technology affects leadership relations). What is interesting to observe here is also whether face-to-face interaction conventions are being followed (for example, do avatars walk through one another or not, and if they do, how do they react to this? does real-world body language still apply, even if the physical reasons for it disappear?).

A variety of media forms can be understood as intermediate forms of shared virtual environments at a distance from the eventual end-points of development. Ralph now runs through a number of examples from recent studies into such media forms, including instant messaging (which users often describe as a space they enter or leave), mobile telephony (whose users still gesticulate while talking), etc. All of these combine to form our overall mediasphere, of course, adding to and complementing an overall multimodal connectedness. In this environment, less media richness does not mean poorer interpersonal relationships or task performance - but it does lead to different forms of relating and collaborating (but more research on larger-scale groups needs to be done here).

So, what kinds of avatars, environments, and interactions are required here? What is the best modality (text, voice, haptic, etc.) and level and type of realism? Context matters, of course, but further work must also advance beyond it - and ultimately, are the key issues in design, or in what kind of connected presence in society we want?

Today, videoconferencing is proliferating, but also merging with other technologies. Online spaces support spatial interaction, the development of social norms, and content that engages users. Social networking offers 'always-on' togetherness, and expresses identity and social availability and awareness. Technological problems into the future are solvable, and users will adapt to modality and self-representation and suitability fo conditions to the point that a convergence of modalities will take place. The implications of this are a linking of copresence with engagement, a renewed focus on the affordances of co-presence systems, a need to maintain connectedness and mutual awareness, and the potential for a continuity from high-end to low-end uses, from video to computer-generated environments, and from small to large groups.

Technorati : co-presence, interaction, presence, virtual environments
Del.icio.us : co-presence, interaction, presence, ual environments