Emergent behavior in agent networks: Self-Organization in Wasp and Open Source Communities Sergi Valverde ICREA-Complex Systems Lab, Pompeu Fabra University, Dr. Aiguader 80, 08003 Barcelona, Spain e-mail: svalverde@imim.es Guy Theraulaz Centre de Recherches sur la Cognition Animale, CNRS UMR 5169, Université Paul Sabatier, 118 route de Narbonne, 31062 Toulouse Cedex 4, France Jacques Gautrais Centre de Recherches sur la Cognition Animale, CNRS UMR 5169, Université Paul Sabatier, 118 route de Narbonne, 31062 Toulouse Cedex 4, France Vincent Fourcassié Centre de Recherches sur la Cognition Animale, CNRS UMR 5169, Université Paul Sabatier, 118 route de Narbonne, 31062 Toulouse Cedex 4, France Ricard V. Solé (1) ICREA-Complex Systems Lab, Pompeu Fabra University, Dr. Aiguader 80, 08003 Barcelona, Spain (2) Santa Fe Institute, Hyde Park Road 1399, NM, 87501, USA Abstract Understanding the complex dynamics of communities of software developers requires a view of such organizations as a network of interacting agents involving both goals and constraints. Beyond their special features, these systems display some overall patterns of organization not far from the ones seen in other types of organizations, including both natural and artificial entities. By looking at both software developers and social insects as agents interacting in a complex network, we found common statistical patterns of organization. Here, simple self-organizing processes leading to the formation of hierarchies in wasp colonies and open source communities are studied. Our analysis further validates simple models of formation of wasp hierarchies based on individual learning. In the open source community, a few members are clearly distinguished from the rest of the community with different reinforcement mechanisms. Keywords Social Networks, Self-organization, Open-source software development. Introduction In both nature and engineering, complex designs can emerge as a result of distributed collective processes. In such cases, the agents involved (such as social insects or humans) only have limited access to the global pattern being developed. Of course, an important difference emerges when dealing with the knowledge and goals of