Doctor of Philosophy
Department of Information Systems
Underwood, Jim, Meaning in IS development: understanding system requirements and use with actor-network theory, Doctor of Philosophy thesis, Department of Information Systems, University of Wollongong, 2001. http://ro.uow.edu.au/theses/1454
Design is a process of matching technical possibilities to user needs. Informatlon systems development methodologies usually rely on achieving this match through the development of mutual understanding between technical experts and users. With traditional life-cycle methodologies this understanding is sought via increasingly detailed agreements as development passes from stage to stage, while with prototyping understanding is continuously refined through interaction with models of the proposed system. If mutual understanding is not achieved it seems likely that this will lead to systems failure.
System understanding is represented by a variety of texts - documents, diagrams, conversations and other system models. Meanings of these texts are realised differently for different participants in the development process - they are interpreted in the discourses of disciplines such as computer science, accounting or marketing.
Our research surveyed a wide range of theories of meaning, linguistic, semiotic and social, and their application to information systems development. The most promising theories for understanding meaning in IS development were found to be Foucault's discourse theory, which describes the formation of the serious language of disciplines and the disciplining of people by these languages, and actor-network theory, which follows the evolution of projects via the propagation of appropriate scripts through a population of potential supporters. These theories in combination were applied to a case study of an internet based flexible learning project in an Australian university.
In our case study we found clear examples of the stages of project development suggested by actor-network theory, of scripts being translated as they passed among actors, of alliances based on acceptance of these translations, and of actors who were humans, activities, concepts, projects or computer programs, all interpreting scripts through their own preferred discourses. We found that misunderstandings were not regarded as a major problem, and that different and incompatible interpretations of scripts were allowed to coexist in order to keep stakeholders with diverse interests committed to the project.
Our findings confirm that ambiguity can play a positive role in IS development, and that project success may depend on keeping meanings open to question during both system development and system operation.