This study examined the local organisation of interaction in chat rooms from an ethnomethodological perspective in order to theorise about the organisation of chat room interaction in a way alternate to cyber-culture and critical cyber-culture studies. Ethnomethodology, conversation analysis and membership categorisation analysis were selected as methodologies ideally suited to the analytic task of unveiling the situated and achieved nature of social order and identity in chat rooms. This study's focus on the situated features of order production addressed the main limitations of prior studies by detailing the quiddity of chat room talk, using categories reflectively (as distinct from the unreflective use), and by reconceptualising the phenomenon of social order in chat rooms.
Data were collected by 'logging' chat room messages (saved as text files) produced in the chat rooms joined. 'Logs' (or log files) contained everything seen on the computer screen by chat room users except for the interface that included a user list showing all the users in the chat room and their access status. Data were collected daily for one or two hours over a two-year period resulting in over 3000 pages of textbased conversation. One chat room, #IRCbar was recorded daily over the two year period while several other chat rooms were visited once or twice. While a variety of chat clients and servers were sampled, the data examined were collected using the MIRC client and only one or two chat servers. These servers are not named here in order to preserve the anonymity of participants. In some data the researcher is a participant in the chat room talk, while in other data the researcher is absent in talk although 'lurking' in the chat room. Access to highly accurate records of participants' chat room talk across a range of different chat rooms and occasions provided for the detailed observation of members' organisation of their talk and variation across instances.
Analysis of these data revealed methods by which chat room members organised their interaction in situated ways by orienting to, and thus accomplishing, the local relevance of particular versions of social order on each occasion. In this way, social order was seen to be not a property of chat rooms but, rather, the achievement of chat room members on each occasion of interaction. Analysis also revealed that versions of social order were accomplished in different ways on occasions, and that the version of social order invoked on an occasion of interaction could also vary. Membership categorisation analysis identified four membership categorisation devices recurrently oriented to by members in organising their chat room interaction. These devices were ownership, hospitality, access status and stage of chat room life.
The analytic findings of the study were significant in that they contradicted prior studies of chat room interactional organisation that treated chat rooms as a uniform interactional setting and that assumed social order was a stable phenomenon across chat rooms. Detailed analysis of talk and description and attention to the reflective use of categories revealed possibilities of practice in organising chat room interaction that provide a new focus - outside of structuralist and post-modem paradigms - for further observational studies of chat room interaction. The findings also demonstrate the analytic rewards of reflective category use by the researcher and of embedding the analysis of categorisation within an analysis of sequential organisation.