Can we really make them work ?

Besides being a distribution medium for up-to-date information, the Internet provides professional communities of users with an infrastructure for collaborative work. An important problem, however, is that groups working over the Internet often fail to accomplish their goals. In this paper our focus is on one category of professional communities, namely research networks. These are networks of professionals collaborating to produce joint artifacts, such as groups reports or journal publications. Although it is often assumed that Internet information tools greatly improve the productivity of research groups, in practice many applications are not very successful. The key questions which we concentrate on in this paper are: what are possible explanations for failures of Internet-mediated research collaboration and, just as important, can something be done to improve on this? One hypothesis, which is the basis of the RENISYS specification method for research network information systems, is that the user-drive co-evolution of requirements and tools will lead to more adequate network information systems, which in turn should facilitate better network collaboration. In this paper we reflect on this assumption using contemporary ideas from organizational sociology. Instead of seeing a research network as a static form of organization, it should be approached as a process of organizing which is continuous and never complete or finished. We conclude the paper by drawing attention to the continuous legitimacy of the structure of a network information system as a crucial condition for its success.