An emergent, multi-level model of the systems development process

While systems development projects have become associated with increasingly sophisticated technologies and organisational changes, the models underlying the process have not kept pace. The Legislation Systems Project (LSP) is an example of a recent project where computerised technology was employed to improve the creation of and access to legislation. This project was substantially delayed due to unforeseen technical complexity and user acceptance was an ongoing concern. It will de demonstrated that the underlying model of process, which broadly reflects the dominant model of current best practice, is strongly linked with these problems. This dissertation introduces and describes a multi-level process model which provides an in-depth expose of the systems development process. This model reflects the closely intertwined nature of the content, context and process of change (Pettigrew 1985) and the interconnected relationship between the micro, macro and institutional levels of change. As illustrated by descriptions of the LSP, the process can be viewed as the negotiation of meaning and interest at a micro level. This involves active participation, the creation and sustainment of coalitions of commitment and the application of authority and expertise. This analysis builds on the work of Walsham and others who have suggested the process of systems development should be viewed primarily as social interaction. However, there are other levels of analysis which should be considered. Process implies change over time and Giddens' (1979) three levels of temporality provide a framework for a multi-level model of the systems development process. At an institutional level, Giddens' structuration theory provides a useful framework for describing the manner in which change is deeply embedded in its organisational context and facilitates the creation of new social institutions. Both this and the microlevel processes can be described as emergent, both over time (chronologically) and in terms of social relationships (hierarchically). This dissertation argues that the commonly used model of the process at a macro level, the systems development lifecycle, is inadequate as it ignores the emergent nature of the process at the micro and institutional level, the ongoing influence of the organisational context and the socially negotiated nature of change. However, there are existing alternative macro-level process models which do reflect both chronological and hierarchical emergence and this dissertation suggests some alternative ways of viewing system development initiatives such as the LSP.