Political Science, Department of


Date of this Version

October 2006


The willingness of people to risk their lives fighting on behalf of their nation (which we call heroism) is a background assumption in the study of war, thus of international relations, but also an evolutionary puzzle. We use two computer simulations to explore the possibility that heroism could have evolved as a domain specific form of altruism, selected through humans’ ancient past as a consequence of warfare. In the first, “altruism” is modeled as a generalized disposition that promotes both heroism and other, non-military, forms of group-benefiting behaviors—which we call communitarianism. In the second, heroism and communitarianism are modeled as domain specific dispositions free to evolve independently. Warfare promotes weak selection on generalized altruism, somewhat stronger selection on communitarianism, and substantial selection on heroism. Heroism evolves more readily when groups are small and relatively equal in size. However, the level of evolved heroism is unaffected by whether war is rare or common. An analytic model indicates that heroism should evolve to higher levels when the rate of casualties in defeated groups is high. Our results suggest why special purpose modes of altruism might evolve more readily than a generalized propensity for altruistic behavior.