Book Manuscript: Military Disobedience in China: Networks, Loyalties, and Strategy


Obedience to authority is often thought to be paramount in military organizations, but in fact disobedience in various forms — including desertion, foot-dragging, rebellion, compromise, negotiation, surrender, mutiny, and outright insubordination — is common. This dissertation explores the dynamics that drive such behaviors by investigating how military commanders respond to their orders in war, with a particular focus on explaining when well-trained, ostensibly loyal commanders disobey their superiors. 

My theory of commanders' responses to orders proceeds in two stages. In the first, commanders judge whether their orders are appropriate or inappropriate — appropriate orders will be obeyed, but there are multiple potential responses to inappropriate orders. Judgments of inappropriateness result when commanders think their orders will increase their risks or create tensions between competing social groups to which they belong. In the second, the interaction between social network brokerage and what I term "command centrism" — the strength of identification with the command authority that issued one's orders — determines how commanders respond to orders they deem to be inappropriate. Brokerage gives commanders power to resist orders and also makes them less willing to change their social positions, while command centrism makes them want to support their superiors. 

I argue that the interaction between brokerage and command centrism leads commanders to pursue four types of responses to orders they deem inappropriate: they can refine their orders, defy their orders, obey their orders, or exit their military roles. All brokers remain in their military roles in hopes of maintaining their brokerage-derived social power – command-centric brokers work with their superiors to refine and improve their orders; brokers who are not command centric defy their orders outright. Non-brokers who are not command centric feel little obligation to support their superiors and have little social power to lose by changing their social positions, so they respond to inappropriate orders by exiting their roles in the military. Non-brokers who are command centric, however, lack the power that might make them risk disobedience and, because they identify strongly with the command authority that issued their orders, they do not see exit as legitimate. They therefore simply obey. 

To substantiate my argument, I identified one case of each type of response to inappropriate orders among Chinese commanders during the Sino-French War (1883-1885): Xu Yanxu (徐延旭) defied his orders, Liu Ming-ch'uan (劉銘傳) refined his orders, Bao Chao (鮑超) grudgingly obeyed his orders, and Liu Yongfu (劉永福) considered exit. I then applied a combination of case study and network analysis techniques to primary sources I gathered in the Beijing and Taipei archives from 2012-2015, in order to investigate each commander's brokerage status, command centrism, and judgment of his orders' appropriateness. In addition to leveraging personal texts — including private correspondence, internal government documents, military communiqués, media reports, and official pronouncement — to evaluate commanders' social positions and judgments of appropriateness, I constructed novel egocentric databases that allow me to model changes in commanders' social networks over time while directly measuring brokerage. 

Article Manuscript: "Rogues, Degenerates, and Heroes: Insubordination and Identity in Military Organizations" (with Sarah Parkinson).


Disobedience in military organizations affects critical outcomes such as civil-military relations, casualty rates, civilian abuse, battlefield effectiveness, and negotiations. However, existing work on military disobedience focuses on the role played by group-level pressures, especially those that operate on low-ranking soldiers. This paper instead investigates the circumstances under which high-level, disciplined, ostensibly loyal members of a military organization disobey. Drawing on social network theory, we note that military officers maintain social ties both to the military and to other social groups. We argue that because officers interpret orders based on frames of reference distilled from both military and non-military identities, and because they often serve as brokers between these networks, the potential for disobedience presents itself. Specifically, certain military orders “activate” overlapping constitutive networks. This process creates tensions that must be adjudicated when the officer decides whether to obey or disobey. 

We abductively develop this theory by deploying a paired comparison of elite disobedience in the Chinese military and the Palestine Liberation Organization. Relying on primary sources and personal texts, we demonstrate how overlapping identity claims led two ostensibly dissimilar individuals to disobey in similar ways during the Sino-French War (1883-85) and the Lebanese Civil War (1975-1989). Our theory thus shows how overlapping social networks create conditions of possibility for even previously loyal, well-trained commanders to disobey their superiors. This paper thus highlights one of the ways relational conceptions of identity influence intra-military dynamics, while also pointing towards new research agendas on the interactions between war and everyday social systems.