War arose recently, anthropologists contend

A battle has broken out among scientists trying to untangle the origins of war.

The fighting is over whether hunter-gatherer communities in recent centuries have tended more toward war — defined as banding together in groups to kill people in other populations — than toward one-on-one attacks within their own communities. A second front has broken out over how to extrapolate from modern behavior to the Stone Age. Some  anthropologists regard  the nomadic groups as helpful if imperfect models of Stone Age human behavior. Others suspect that too much evolutionary change and irregular contact with outsiders make hunter-gatherers unreliable signposts of the past.

Lethal attacks on one community by another rarely occurred during the 19th and 20th centuries, according to a new analysis of data previously gathered from nomadic hunter-gatherer populations. Murders of one person by another in the same group accounted for a majority of intentionally caused deaths, anthropologists Douglas Fry and Patrik Söderberg of Åbo Akademi University in Vasa, Finland, report in the July 19 Science.

Ten of the hunter-gatherer groups  had no recorded killings involving more than one attacker, effectively making those societies no-war zones, Fry and Söderberg say.

The new evidence suggests that humans have evolved a tendency to avoid killing in general, the researchers contend. War originated only within the past 10,000 years, in their view, with armed conflicts intensifying as the first states expanded between 6,000 and 4,000 years ago.

“Fry and Söderberg go against the popular tide in science … and win hands down,” says anthropologist R. Brian Ferguson of Rutgers University in Newark, N.J.

Archaeological evidence from Europe, the Middle East and western Asia contains relatively few signs of murder and war until after 10,000 years ago, he says.

But the new study has attracted fire from other investigators. “Fry and Söderberg use the hunter-gatherer record inappropriately to push the idea that because many modern hunter-gatherers were not seen to have war, ancestral hunter-gatherers also did not often have war,” says Harvard anthropologist Richard Wrangham.

Wrangham and others say that the new paper ignores relatively high homicide rates previously documented in hunter-gatherer groups, including some in the study. Critics also point to reports of regular fighting among neighboring hunter-gatherer communities; the groups that Fry and Söderberg studied were largely isolated. From critics’ perspective, war probably goes back tens of thousands of years and stoked the evolution of intense cooperation within, but not between, human groups.

Murders cause more deaths than war in both traditional and modern societies, with exceptions coming during the 20th century’s two world wars, says Harvard University psychologist Steven Pinker. Given war’s rarity, researchers are unlikely to observe raids and other attacks on rival groups when studying small hunter-gatherer samples such as those in the new study, he says. Rates of violent death are higher among hunter-gatherers and in other non-state societies than in state societies, he adds.

Fry and Söderberg’s finding that mobile hunter-gatherer bands infrequently organize warlike attacks does not surprise anthropologist Polly Wiessner of the University of Utah. But raiding and war does take place in a few such groups, as well as among sedentary hunter-gatherers that live year-round in bountiful settings near coasts or rivers. The great unanswered question, Wiessner says, concerns “how different societies harnessed and tamed aggression to build larger societies throughout human evolution.”

Fry and Söderberg identified data on 148 killings in 21 mobile hunter-gatherer groups. Just over half of those killings were committed by lone perpetrators. Almost two-thirds resulted from disputes within families, executions of group members, competition among men over women and other conflicts within groups.

About one-third of killings involved attacks by one group on another. Reasons included disputes over resources, thefts of women and revenge attacks for past stealing or other offenses.

Australia’s Tiwi had an exceptionally large number of killings, 69, and accounted for most of the lethal attacks across groups.

Economist Samuel Bowles of New Mexico’s Santa Fe Institute criticizes Fry and Söderberg for choosing relatively peaceful groups, including the Tiwi, that mostly live in places where state-run armies discourage intergroup conflict. In his 2009 analysis of eight hunter-gatherer societies, the Tiwi ranked near the bottom in estimated rates of war-related deaths. None of the other seven groups he studied were part of Fry and Söderberg’s work.

A handful of reports have likewise found fairly regular, usually low-level warfare among neighboring hunter-gatherer societies, Wrangham says. In those cases, hunter-gatherers had little or no contact with more powerful farming communities that could have discouraged fighting. Warring groups had different customs and spoke different languages or dialects. Fry and Söderberg mainly addressed conflicts between bands of hunter-gatherers with common customs and languages, which reveal little about the evolution of war, Wrangham contends.

Wrangham considers periodically warring hunter-gatherer groups to be the best available models of Stone Age practices. In a 2012 paper, he and a colleague proposed that ancient people in groups such as these evolved a tendency for males to band together and opportunistically kill members of rival groups — much as chimpanzees do.