Culture and Control in a Media Space

Media spaces integrate audio, video and computer networking technology in order to provide a rich communicative environment for collaboration. The connectivity which they provide brings with it important concerns regarding privacy, protection and control. In order to derive the fullest benefit from this technology, it is essential that these issues be addressed. As part of our investigation of media space systems, we developed a computational infrastructure addressing these problems our own working environment. A key aspect of this work is the relationship between two aspects of this control system—the technological components which determine how the system will behave, and the social components which determine acceptable use and behaviour. This paper discusses our experiences with the privacy and control aspects of our RAVE media space environment, specifically with regard to connection management, and compares them to the experiences of other research groups. We discuss the nature of the relationship between technological and social elements in using this technology, and discuss the consequences for the design of such systems. Introduction: Media Spaces One focus for research into workgroup communication and collaboration is the investigation of media space technology. A media space is formed by the combination of audio, video and computer networking technologies to provide a flexible, dynamic interconnection environment for collaborative work groups. Our “RAVE” media space (Buxton and Moran (1990); Gaver et al (1991)) is one of a number of systems which have been used in investigations into issues of workgroup support and collaboration (e.g. Root (1988); Stults (1989); Abel et al (1990); Mantei et al (1991)). At its most basic, the media space provides a means for setting up multi-media communication channels between individuals and groups; beyond that, it supports collaborations, both formal and informal, providing a communication infrastructure which is amenable to rapid reconfiguration and connection. This element of flexible and dynamic control by individuals is at the heart of what we mean by media space; a media space is more than simply point-to-point audio and video connections. While fixed video links can provide communication between remote individuals, the easy reconfiguration and control of connections is critical in exploring the means by which the technology can create a space in itself, rather than being simply a restructuring of the physical space. Clearly, the provision of this sort of media infrastructure is a useful means towards contact and direct collaboration between individuals who are not colocated in time or space. The media space is not just a desktop video-conferencing system; our research (e.g. Dourish and Bly (1992)) has pointed to other, less formal mechanisms which can be built in media space environments to support a variety of work groups. On the other hand, it is equally clear that the introduction and use of media space technology raises a host of issues concerning the protection of individual privacy and access control. This is particularly true in environments such as RAVE which are used not only by research staff involved in the development of the technology, but also by other office workers throughout our lab. Even within our own environment, then, it was critical that these issues be addressed if our work was to be successful. The result is a software component called “Godard” (Dourish, 1991) which provides inhabitants of the media space with flexible control over the degree of access they grant to others, and dynamic information on connections and connectivity. However, control and accessibility in the RAVE environment is not purely based on a technological infrastructure, but also on the culture which has developed around this system. We would claim that any technological control system like Godard sits within a culture which determines aspects of its use; and that this relationship between technological and social control must be considered as part of the design process. In essence, the relation between the technology and its use is coadaptive, and must be analysed as such (Mackay, 1990). This relationship is the central focus of this paper. In the rest of this paper, we will first discuss aspects of the Godard system in more detail, and describe the way in which its introduction changed aspects of media space use within RAVE. We will then consider the design of some other media space environments and aspects of privacy control and usage culture in those systems. Finally, we will talk more generally about the relationship between technological and social control and its importance in designing such systems. The RAVE Media Space Environment The Ravenscroft Audio-Video Environment is a media space being developed and used at Rank Xerox EuroPARC. It comprises switchable analogue video and audio networking technology, workstation software to control connections among individuals and public spaces, and other systems which provide facilities for coordination, informal interaction and focussed collaboration. RAVE extends these with computer-based systems which augment the power of the media space for collaboration. Multi-user drawing surfaces (e.g. ShrEdit (McGuffin and Olson, 1992) and shdr (Dourish, to appear)), a shared active calendar (Lövstrand, 1991) and other tools for collaboration in a number of computational domains are all important parts of the system as a whole. One critical aspect of the RAVE system is its ubiquity within our work place. Rather than being restricted to a small group of individuals who are involved in the development of the technology, RAVE connections are provided to all members of the lab, including research, technical and administrative staff. Although there is no obligation to make use of the technology, the hardware and associated software systems are made available by default. Furthermore, users are encouraged to “adopt” the technology and make it their own. For instance, there is no obligation to have cameras pointed directly at work areas. Pointing the camera out of a window for privacy during a meeting or conversation is quite common. Clearly, the ubiquity of our system raises a number of issues concerning privacy and protection for individuals. Being deployed through our lab, the RAVE system does not require the “sign-up” process observed in some other environments, where an individual must explicitly request membership of the media space, and accept whichever norms govern its use. Hence it was extremely important for us to explicitly address issues of privacy and control in the development of our system. It was critical to the success of our experimentation that the participants could maintain the level of privacy they desired, maintain control over the ways in which others could connect to their offices, and have flexible control over feedback indicating the state of the media space and their visibility within it. At the same time, it was also important that we maximise the flexibility and utility of our media space, in order to investigate the range of ways it could support collaborative work in our environment. RAVE is, after all, a research tool. These requirements led us to develop a solution based on people’s styles of interaction in the real world (Anderson and Moran, 1990). This is realised primarily in a system component called Godard.