Dynamics of Chat Spaces

My first session on this second day of the Association of Internet Researchers conference 2005 in Chicago is on the 'Internal Dynamics of Online Spaces'.

Janet Armentor-Cota: Uses of Web Chat

Janet Armentor-Cota from Syracuse University is the first speaker, presenting on the dynamics of a Web chat community. The paper she presents here is coming out of her dissertation, and looks at a Northeast (U.S.) romance chat room. Web chat, of course, is usually real time, multi-participant, and consists of messages of short length, with almost constant traffic around the clock. Web chat is also a multimedia phenomenon and can incorporate images and audio and video streams. So, how do processes and structures of multimedia technologies organise the chat space, and what processes occur here?

The key interest here is an understanding of community in relation to space - in terms of interface space (the visibility of communications and interactions, in a shared pace; the social space, constituted through community interactions; and the metaphorical space, the sum of the perceived spatiality. The research proceeded, then, through virtual ethnography from the point of view of a lurker in the Web chat space, as well as through a discourse analysis of logged conversations both in text and audio formats.

Some of the findings: the space is a place for excitement - users access the space as a place for fun; the use of voice chat was also interesting (voice utterances tended to be longer and more complete than text utterances); and users tended to divide into drifters and seekers, as well as some very regular participants who have developed a strong sense of community. What is interesting here is that regulars will often move through the networks together, skipping from one chat room to another as a group (and organising that move together). Regular participants also often blurred the boundaries between offline and online, combining meetings in person with meetings in the chat room (and organising offline meetings through the chat room).

Karine Barzilai-Nahon: Gatekeeping in Israeli Discussion Groups

The next presenter is Karine Barzilai-Nahon from the University of Washington. Her approach to virtual communities is one based around gatekeeping, and she begins by noting that control of information remains crucial. The gatekeeping idea comes out of journalism, of course, where editors and journalists as gatekeepers determined what readers would receive. Online, as many have observed, this kind of gatekeeping doesn't apply any more, and alternative models are being developed.

Ultimately, gatekeeping is simply information control; it is a channelling mechanism - and Karine has developed a typology describing a number of new models for gatekeeping (channelling e.g. through search engines, censorship e.g. through filtering and blocking, internationalisation e.g. through localisation and translation, security e.g. through authentication systems, and cost-effect e.g. through paid access restrictions).

These ideas, then, are the foundation for Karine's study of virtual communities. They can be used to ask the questions of whether and why a gatekeeping event (such as the deletion of  a message) occurred in an online community. Karine had access to log and content data from the four biggest ISPs in Israel, and used this to analyse gatekeeping procedures a various levels (from regulators to ISPs themselves, to lower levels within specific communities). All up, some 1.4 million messages were analysed across 717 online communities with some 37,000 members - and some of the gatekeepers were also interviewed.

Some findings of this massive effort: there is a delicate tension between four layers of gatekeepers existing in forums. To begin with, there are the legislative regulators (governments), ISPs, moderators, and community members - and there are various tensions in interest between these levels (ISPs might want to increase the size of groups, while community moderators may want to restrict this in order to keep a sense of community within the group - a question of quality vs. popularity).

Karine also identifies a group of regulars as the community core - these are usually senior users, who most embody the values of a community and thereby also serve as the gatekeepers of the community by contributing to balance the group's regular activities. If this core is strong, the community manager (or moderator) is - and needs to be - little more than a community facilitator; in fact, it is even possible for these core members to talk directly to the ISP to replace the community moderator if necessary, as they hold the real power over the group. At the same time, the senior users were also responsible for a good amount of off-topic posts, as they know one another very well and are more likely to socialise through the forum as well. On the other hand, overlapping loyalties were also notable - users who are members of multiple groups were more likely to have their messages deleted as they are perhaps on the boundaries of all of these communities.

Further, some observations about anonymous guest members: from analysing server logs it was evident that many anonymous users were actually regular participants who were not posting under their own names. In a number of cases they expressed opinions which were contrary to their regular, signed-in persona - perhaps as a way to get around the tendency towards homogeneous group-think in strongly developed, like-minded communities.

Also, there appeared to be a link between the history of users and the likelihood of users' messages to be deleted - even though the community moderators would not be expected to know whether a user's messages had already been deleted in a different community. Moderators, however, appear to accumulate information by rumours and other forms of sharing information - so there is perhaps something of a self-reinforcing spiral where users become ever better known as disruptors, and ever more likely to be deleted quickly.