Social Complexity: can it be analyzed and modelled?

Over the past decade network theory has turned out to be a powerful methodology to investigate complex systems of various sorts. Through data analysis, modeling, and simulation quite an unparalleled insight into their structure, function, and response can be obtained. In human societies individuals are linked through social interactions, which today are increasingly mediated electronically by modern Information Communication Technology thus leaving "footprints" of human behaviour as digital records. For these datasets the network theory approach is a natural one as we have demonstrated by analysing the dataset of multi-million user mobile phone communication-logs. This social network turned out to be modular in structure showing communities where individuals are connected with stronger ties and between communities with weaker ties. Also the network topology and the weighted links for pairs of individuals turned out to be related.These empirical findings inspired us to take the next step in network theory, by developing a simple network model based on basic network sociology mechanisms to get friends in order to catch some salient features of mesoscopic community and macroscopic topology formation. Our model turned out to produce many empirically observed features of large-scale social networks. Thus we believe that the network theory approach combining data analysis with modeling and simulation could open a new perspective for studying and even predicting various collective social phenomena such as information spreading, formation of societal structures, and evolutionary processes in them.