1 Communication emerging? On simulating Structural Coupling in Multiple Contingency Manfred Füllsack Department of Philosophy at the University of Vienna manfred.fuellsack@univie.ac.at Paper, to be presented at the conference „Luhmann in Action” Dubrovnik, 11-15.4.2011 Abstract: The paper presents a Multi-Agent-Model of a population of agents which „learn to communicate” by reciprocally constraining their possibility spaces. Agents are conceived as „closed systems”, meaning that agents are operating solely on the base of their onboard means. They do not have any concept of an „external world”. Therewith, it cannot be said that agents „intend” to communicate or to reach a „consensus” of some sort. However, as they are unspecified in respect to their actions they co-evolutionary aggregate probabilities on how to cope with their environment. In this way, from the structural coupling of multiple unlikely actions (or operations) an „Eigenbehavior” emerges which an observer eventually might interpret as „communication”. Introductory remark on the empirics of this presentation As this paper is presented on a conference subtitled „Empirical Studies of Structural Coupling”, it might need an explanatory remark on its methodology. The empirics of this research-work are self-generated, meaning that the data from which here theoretical conclusions about the coherence of the conception of Niklas Luhmann are drawn is obtained from a computer-generated Agent-based-Model (ABM, see to this technology a.o.: Axelrod 2002, Epstein 2006). The model has been generated in the course of an ongoing endeavor to reconcile, integrate or just to compare the Luhmannian Theory of Systems to the Anglo- American debates currently lead under titles like Dynamical Systems Theory, Theory of Complex Adaptive Systems, or just Complexity Theory (Füllsack 2010, 2011). The data used in this paper thus is not empirical in a classical sense. But it is not purely theoretical either. What has been said about the methodology of ABM in general (Axelrod 2002), thus can be said about the data and the results of this paper: they are generative and therewith implying a „third way” of scientific research. 1. Introduction The list of attempts to simulate aspects of a Theory of Social Systems as suggested by Niklas Luhmann (a.o.1984) is getting longer (cf: Kron 2002, Fleischmann 2005, Barber/Blanchard/Buchinger/Cessac/Streit 2006, Leydesdorff 2008). Repeatedly, a focus of these attempts has been communication, which Luhmann conceives as a process of three consecutive selections - the selection of information, of message and of understanding. According to Luhmann communication takes place whenever an information can be understood as message. (cf. Luhmann 1984: 498) This conception stands in a more or less striking contrast to some basic assumptions of earlier theories of communication, therewith irritating linguists, social scientists and philosophers alike, and probably also impeding the reception of Luhmann in non-German speaking scientific communities. One of the main targets of Luhmann’s argumentation has been the naturalistic or realistic premises on which classical ontology tried to base its attempts to analytically grasp communication. Especially,