At the level of the individual, the army, and the state, there are very few actions that involve higher stakes than warfare. Individuals can die, armies can be annihilated states can go bankrupt or be conquered and lose sovereignty.

Yet in spite of these high stakes, soldiers, armies, and states often do not try to win wars by any means necessary. They often abide by norms that limit their actions and thus decrease their odds of victory. Norms of war refer to general agreements shared by more than one state that govern the nature of acceptable conduct (both prescriptive and proscriptive) during war. Several aspects of warfare are regulated by norms of war: who should fight, based on class (Samurai) and age (Geneva Conventions against military recruitment of children), weapons (papal injunctions against the use of crossbows in medieval Europe, conventions against chemical weapons, landmines, indiscriminant fire), treatment of non-combatants (medieval norms of ransom, Geneva Conventions on prisoners of war, non-combatants, and mercenaries), combat duration (campaigning seasons, holidays, and regularized truces such as Greek Olympics), who can kill whom (prohibitions against killing unarmed combatants in flight or nobility slaying peasantry), the number of soldiers deployed (such as classical monomachy, single combat between champions), how soldiers should dress (British red coats) and where fighting can take place (Greek hoplites fought only on level ground). From the perspective of rational choice theory, abiding by norms like this appears to be anomalous. Can this sort of normative action be explained with a more nuanced rational choice model? We use rational choice institutionalism to explain (1) the existence and content of norms governing action in warfare and (2) variations in compliance with norms of war. The main goal of this paper is theoretical, to outline and assess various answers to these two questions. In addition to this, we will use a few empirical cases, the Battle of Agincourt 1415 and Aztec Flower Wars in the 14th and 15th centuries, to illustrate the causal mechanisms producing norms of war and determining compliance with them.