Information systems as a social science: an alternative to the FRISCO formalism

The FRISCO Report has initiated an important debate on the discipline of information. Chapters 1, 2 and 6 contribute ideas for all shades of opinion, a Broad View contrasting with Chapters 3, 4 and 5, “the core of the report”. This paper presents a Social View as an alternative to this Core View, which it criticises in two ways. Firstly, it limits the scope of information systems by treating it as an adjunct to software engineering, having no place for key properties of information in organisations and society. Secondly, it does not satisfactorily link its framework of concepts to the empirical world, but relies on the mentalistic notions of ‘perceptions’ and ‘conceptions’ in the mind of some, unspecified interpreter. The strength of the Core View lies in its formal precision. But a Social View can lead to an alternative formal framework that can place a much wider range of information systems concepts on a firm empirical basis. Thus provides a foundation for information systems as a social science, rather than a branch of applied mathematics. The Core View includes the belief in an objective reality to which we have direct access via the supposedly transparent languages of words, numbers and diagrams, whereas the Social View, without rejecting that position, requires us to explain how we come to construct such knowledge. It also compels us to examine all the other functions we perform with information, especially our construction of social reality. Forced to examine these questions, we have to address many elusive problems, such as the creation of meanings, the role of intentional communication, the construction of time, and the systems of norms we call ‘organisation’. Finally, the paper re-works the Japan Wines case study used in the FRISCO Report. This shows that the Social View leads to a more detailed analysis of the business problems than the Core View, with as much formal precision, which lends itself to computer interpretation, while also yielding models that are easier for users to interpret