Danny Hirschel-Burns had a great post earlier this month on whether people are violent by nature over at The Widening Lens. He reviews a number of seminal works on this issue (Stanley Milgram, Christoper Browning, Dave Grossman and Randall Collins) and concludes that while most people are “inherently adverse” to killing others, violent coercion and social sanctioning can lead people to kill. Notably, he estimates that intergroup coercion is probably more effective in this regard than either authority or ideology.
This fits very well with Scott Straus’ findings about adult men’s participation in the Rwandan genocide. In his 2006 book The Order of Genocide, he interviews convicted genocidaires* about their actions, and found that the two most common reasons for killing others were “intra-Hutu coercion” and “wartime fear and combativeness” (p. 136). In his telling, the invasion of the Tutsi-led RPF rebels from Uganda in 1990 deeply unsettled ethnic relations, made violence less unthinkable in daily life, and made people more receptive to genocidal political propaganda. When the genocide began in 1994, however, ideology and fear alone were not enough to convince most men to kill. Genocidaires were provided with material incentives, like claiming the houses and cattle of their victims, but were also violently threatened by other Hutus if they didn’t participate.
*Straus notes that because his respondents had already been convicted, they should not have felt it necessary to downplay any anti-Tutsi ideology which motivated them, since they presumably couldn’t be punished again. I think this misses the possibility that prisoners who continued to be openly anti-Tutsi might be socially sanctioned in other ways within the prison system. However, in general I don’t think this detracts from the plausibility of his overall argument.