Abstract
We analyze data on political violence in the United States and India using various statistical and machine learning methods, paying attention to both geographic location and time of an incident of political violence. We briefly examine an epidemiological model of public outrage leading to public violence developed by Nizamani, Memon & Galam (2014) and critically observe that under the assumption that one of the main agents in the model approaches a stable equilibrium value, the violent actors necessarily decrease to zero. This is in contradiction to years of data from the United State and India. We sketch a field theory of upsetness that acts as a background to acts of political violence. The field theory of upsetness stems directly from actual outlier incidents in both the United States and India.