Violence: Clarified

See allHide authors and affiliations

Science  19 Oct 2012:
Vol. 338, Issue 6105, pp. 327-328
DOI: 10.1126/science.338.6105.327-b

Two recent news stories discuss the high average rates of violence in nonstate societies summarized in my book The Better Angels of Our Nature: Why Violence Has Declined (1). Both require clarification.

“The battle over violence” (2) quotes an archaeologist who claims that I selected “a few populations…biased toward supporting [my] argument.” This is untrue. Better Angels reports all the published estimates of per capita rates of violent death in the archaeological and anthropological literature I could find, including those from societies that have been claimed to be nonviolent. Figure 2-3, for example, reports war death estimates from 27 hunter-gatherer and hunter-horticulturalist societies (most data collected between 1825 and 1982), which range from 0 to 1450 war deaths per 100,000 people per year (per capita rates based on the local tribe or village population), with an average of 524 (1). By comparison, the highest estimate of worldwide deaths from war and genocide in the 20th century is 60, and the rate today is far less than 1 (1).

The Enga people of Papua New Guinea.


“Turning from war to peace in Papua New Guinea” (E. Culotta, News & Analysis, 28 September, p. 1593) notes that I praise Wiessner and Pupu's recent study on peacemaking among the Enga (3), “even though it contradicts one of [my] key conclusions.” The study does not contradict my findings. Modern guns did not bring high rates of violence to New Guinea, nor have the Enga's impressive peacemaking efforts brought the rate down to those of modern Western societies. Even before the gun-fueled rise in Enga violence starting around 1990, the average annual war death rate in published estimates from 11 New Guinean societies is 498 per 100,000 (46). During the worst period among the Enga, the authors' estimate is around 100, and after the Enga's peacemaking efforts, it was 32. Compare that to the worldwide average rate of violent deaths today of around 6 to 8 (mostly homicides); the figure for the worst year for American crime, 10; and the range for most Western democracies, 1 to 3 (1). The Enga at their most peaceable still have a rate of violent death that is 3 to 30 times as high as that of much of the modern West.

Although Better Angels documents the pacifying effects of successful states, it did not, of course, claim that any state, no matter how weak or inept, magically reduces violence, nor that state control is the only means by which violence may be reduced. Indeed, the book devoted several pages to Wiessner's studies of peacemaking among the Enga, an excellent example of a “civilizing offensive” that has reduced violence at several points in history.

Facts about violence in nonstate peoples have often been politicized because people worried that high estimates would imply that efforts at reducing violence are futile [e.g., (710)]. The value of Wiessner and Pupu's meticulous study is to show why this fear is misguided.

Regarding Benjamin et al.'s Letter, it is encouraging that “anthropologists of peace” now see their discipline as empirical rather than ideological, a welcome change from the days when many anthropologists signed manifestos on which their position on violence was “correct,” and censured, shut down, or spread libelous rumors about colleagues who disagreed (710).


Navigate This Article