A Simulation of Conceptual Collaboration in Science
by Petra Ahrweiler
The contemporary forms of science and technology production find their place in a heterogeneous cultural environment. Many different actors offer varied worldviews in terms of which the goals and contents of innovation must be negotiated. Diverging interests struggle over the primacy of their interpretations which attempt to define what particular innovations should look like. These negotiations take place in innovation networks and have to deal with conceptual contributions which often are only interpretable and intelligible against the experiential background of the respective contributors. The arranging and integrating performances of negotiation networks have to feed the strategic perspectives of the individual participants into the complex innovation processes.The computer model presented in this paper simulates the different carriers of negotiation networks and the performance of their interactions.
The model combines two streams of formal studies on knowledge production: on the one hand the language-based approaches to reconstruct the logical relations within notion systems (mental models, theories, etc.) and on the other hand the works on evolutionary dynamics within knowledge systems (memetics, kene-approach, etc.). The former produced inference machines fit for all kinds of internal consistency testing and inter-theoretical comparisons. However, these studies did not show conceptual systems at work or - even more ambitious - conceptual systems in interaction. Such dynamics were, however, shown by the latter approach. While simulating the emergence of knowledge structures these studies dealt with conceptual units as black boxes owning ascribed attributes and capabilities. Their representations were not language-based: why and how exactly conceptual systems could join up was no matter of the internal performance within these models.
The simulation model SISiFOS (Simulating the Integration of Studies on the internal Formation and the Organisation of Science) brings these two approaches together: it simulates the emergence and interaction of knowledge structures using a language-based representation of the participating conceptual units. The SISiFOS simulation generates conceptual networks built up from different knowledge sources (i.e. actors in innovation networks) by two integrative mechanisms. The first mechanism (strong integration) produces a self-organising conceptual ecology based on (re-)combination of different knowledge sources. Here, autonomous knowledge sources react to their respective environment consisting of an object area and all other participating knowledge sources. In case of the incompatibility of two knowledge sources the second integrative mechanism (weak integration) builds networks of loose coupling that allow the switching of the SISiFOS observer between the different conceptual offers of incompatible knowledge sources via conceptual boundary elements and their special reference structures. Some results of SISiFOS are presented from applying the model to an innovation discourse about the future shape of the European car distribution system. This show case suggests applications of the computer simulation to problems of mediation and communication management (round table engineering) in the social practice of technology production.
