| In this dissertation, I develop a sociological conception of violence. I use the concept of 'social agency' to represent violence as a social phenomenon with both embedded relations and an autonomous role in society. I argue that political violence projects use this agency to develop relations within a social context, and also to intervene in it. In Chapter 2, I analyze the regional incidence of political violence in India between 1981 and 2013 to show that there are significant shifts in the form of political violence and its causal profile. I hypothesize that the form shifts from mass-based, localized, intergroup to specialized, hybrid, and multi-level. The causal profile shows a significant shift from identity factors to economic factors and also an increased salience of political factors. In Chapter 3, I argue that there is a common pattern of relations between different types of political violence and their social contexts. I compare the regional incidence of left extremist and ethnic identity-based violence in India between 1981 and 2013 to explain how the social agency of violence accounts for both their commonalities and differences. I hypothesize that left and ethnic violence orient differently to processes of socio-cultural and economic differentiation but have similar relational orientations to political contention and regular social conflict. In Chapter 4, I argue that violence has a dual social agency of both mobilizing protests and reproducing inequality in society. I analyze the district-level incidence of left extremist violence in India to show that the use of violence in left political projects conforms to the interests of the upper castes. I hypothesize that the incidence of left violence involves the conversion of relations of inequality into organizational resources, re-activation of tribal histories of conflict, and ideological subversion of caste-based inequality. |
