Modeling Insurgent Dynamics Including Heterogeneity: A Statistical Physics Approach

摘要

Despite the myriad complexities inherent in human conflict, a common pattern has been identified across a wide range of modern insurgencies and terrorist campaigns involving the severity of individual events-namely an approximate power-law x~(-α) with exponent α≈2. 5. We recently proposed a simple toy model to explain this finding, built around the reported loose and transient nature of operational cells of insurgents or terrorists. Although it reproduces the 2. 5 power-law, this toy model assumes every actor is identical. Here we generalize this toy model to incorporate individual heterogeneity while retaining the model's analytic solvability. In the case of kinship or team rules guiding the cell dynamics, we find that this 2. 5 analytic result persists-however an interesting new phase transition emerges whereby this cell distribution undergoes a transition to a phase in which the individuals become isolated and hence all the cells have spontaneously disintegrated. Apart from extending our understanding of the empirical 2. 5 result for insurgencies and terrorism, this work illustrates how other statistical physics models of human grouping might usefully be generalized in order to explore the effect of diverse human social, cultural or behavioral traits.