Modelling the Dynamics of Cooperative Work Arrangements

Abstract

Many computer systems are developed using object oriented methods. Many words are written about object oriented methods. Less is reported about evaluations of the actual usages of these methods in development projects, and about the correspondence between the methods and the characteristics of the various problem domains. This paper speculates on some modelling issues of object orientation from a usage-related, yet theoretical point of view, one of the points being the identification of areas where larger, empirical, studies would be meaningful. Cooperative work arrangements are the type of problem domain in focus. The need for efficient computer support for the coordination of complex and distributed tasks poses requirements beyond those of more traditional application areas where the information contents of the system to be has been the focus. Coordination means that the usage of the system must be controlled, guided, provoked, signalled, or restrained, and, crucially, that the usage of interest is not that by a number of independent persons, but by and amongst several cooperating persons. Three object oriented modelling techniques are studied, paying particular attention to such requirements and to the system context views in which the techniques are seen, and the inherent dynamics of coordination mechanisms are discussed in the light of the distinction between event and function. Among the considered modelling techniques are behaviour diagrams, interaction diagrams, and data flow diagrams, each corresponding to one or more well known object oriented methods. On the basis of a specific theory of coordination work and coordination mechanisms, the strengths and weaknesses of these approaches are discussed. It is concluded that state of the art object orientation does not provide for the necessary techniques, but that they do possess a potential for expansion in this direction.