How can we detect narratives that legitimize the use of force? Using the ‘hero-protector narrative’ as a theoretical model to understand how political leaders manufacture collective emotions when advocating the use of force as the only effective and morally acceptable option, we show how such an argument can be tested using computer-assisted narrative analysis. Based on a case study of political speeches and statements legitimizing-and others rejecting-the use of force in the context of the 2003 invasion of Iraq, we suggest that our methodology offers an innovative approach to uncover similarities (and differences) across narratives used to mobilize collective emotions of moral anger and compassion.