As researchers in the field of complex systems, we are used to applying tools inherited from statistical physics to the study of social behavior. In practice, however, significant barriers still exist between researchers with backgrounds in physics and mathematics and those trained in sociology, anthropology, or social psychology. The former often develop highly simplified models and focus on their theoretical properties, such as thermodynamic limits and phase transitions, while paying little attention to empirical data. The latter, in contrast, tend to produce rich experimental studies without aiming to extract general or universal behavioral principles. In this talk, we discuss recent empirical results that connect fundamental concepts from statistical mechanics with the dynamics of networks of personal relationships, linking microscopic interaction patterns to macroscopic network structure. We show how these results can be used to construct realistic models of social network evolution that retain the simplicity and explanatory power of statistical physics. We also explore the limitations of such models, highlighting the risks of both overly simplistic and overly complex approaches. Finally, we examine how these dynamics can be studied at the mesoscopic scale and discuss the limitations of community detection methods in capturing socially meaningful structure.
Contact details:
Juan Antonio García Contact form