What can the micro-sociology of violence contribute to understanding the mass killings in Aurora, Colorado, and similar incidents? In the immediate shock of public attention, there is an imperative to give policy answers. I could join the chorus advocating a ban on weapons in the USA. This is a hope; it is not a guarantee. Mass shootings are very rare events. There are about 15,000 homicides per year in the USA; the great majority are single-victim killings. Less than 1% are mass killings (4 or more victims in the same incident). Spectacular mass shootings, where many persons are killed or wounded, have been happening at a rate of about 1 or 2 per year, in the 30 years since 1980, for the most common type, school shootings; shootings in other venues, apparently imitating school shootings, are rarer but on the rise. It is their rarity that attracts so much attention, and their out-of-the-blue, seemingly random relationship between killer and victims, that makes them so dramatically alarming.
This rarity means that very distinctive circumstances are needed to explain mass killings, and that widely available conditions cannot be very accurate predictors. There are approximately 190 million firearms in the civilian population in America, in a population of 310 million. The vast majority of these guns are not used to kill people. Even if we focus on the total number of yearly homicides by gun (about 12,000), the percentage of guns that kill someone is about 12,000 / 190,000,000, or 1 in 16,000. Another way to put it: of approximately 44 million gun owners in the US, 99.97% of them do not murder anyone. It is not surprising that their owners resist being accused of abetting murder.
(H/T Geoff Nathan)