Two Failures in Computer-mediated Textcommunication

Abstract

The low uptake of computer text conferencing, despite its unique functionality, suggests that developers of this technology have paid insufficient attention to human factors. In Ergonomics (Human Factors Engineering) the study of untoward events such as user errors, systems failures and disasters is a widely-used and fruitful research strategy. Anomalous and unexpected events and actions likewise constitute an important source of data and insights for sociological, ethnographic and linguistic studies of human communication. In the present paper we combine these traditions, applying qualitative failure analysis to computer-mediated text conferencing. The data is drawn from two episodes of communication failure involving geographically dispersed groups engaged in different types of task. In one episode a consultative discussion aimed at the development of professional expertise collapsed in rancour. In the other, a design error was propagated, through misunderstanding, into the implementation stage of a project. The analysis has substantive implications for user training in new types of writing and reading skill and for the functional and interface design of textcommunication systems. There are also methodological implications for studies of computer mediated communications. It is argued that a surface analysis of the episodes is insufficient for full understanding of the communication failures: the record must be interpreted in the context of the practices, institutions and structures that exist in the social world outside the computerised textcommunication system.