John Horgan - http://blogs.scientificamerican.com/cross-check/2013/07/18/new-study-of-… .
Of the 21 [hunter-gatherer] societies examined by Fry and Soderberg, three had no observed killings of any kind, and 10 had no killings carried out by more than one perpetrator. In only six societies did ethnographers record killings that involved two or more perpetrators and two or more victims. However, a single society, the Tiwi of Australia, accounted for almost all of these group killings.
Some other points of interest: 96 percent of the killers were male. No surprise there. But some readers may be surprised that only two out of 148 killings stemmed from a fight over “resources,” such as a hunting ground, water hole or fruit tree. Nine episodes of lethal aggression involved husbands killing wives; three involved “execution” of an individual in a group by other members of the group; seven involved execution of “outsiders,” such as colonizers or missionaries.
Most of the killings stemmed from what Fry and Soderberg categorize as “miscellaneous personal disputes,” involving jealousy, theft, insults and so on. The most common specific cause of deadly violence—involving either single or multiple perpetrators–was revenge for a previous attack.
The anthropologist , R. Brian Ferguson, considered to be the foremost expert on the early history of war, conducted an extensive global survey of archaeological records and has found no substantive evidence of systematic violence in prehistoric human societies.