This data review presents quantitative information on violent deaths based on archaeological and ethnographic studies.

Homicide rates for modern times are routinely published by statistical offices or other state agencies, and research institutes publish reliable data on war deaths.

For the study of lethal violence in non-state societies we have generally two different sources of information: for the more recent past (since the late 19th century) ethnographic evidence is abundant.

For the more distant past, we have information on the prevalence of violence from archeologists who have studied violence in past societies through skeletal remains at archeological sites.

We present and describe data from both archaeological and ethnographic studies below.

Importantly, this is not all the information we have on violence in different societies. Non-quantitative information on violence is abundant, such as in art.

And both types of studies often cannot distinguish between whether a violent death was inflicted by a member of the same group or by another, and therefore is more similar to what we think of today as a homicide or a death from war.

We see that the levels of violence in prehistoric times and in non-state societies differ widely between societies. But on average they tend to be significantly higher than in modern state societies and the world today.

Rate of violent deaths in state and non-state societies

Click to open interactive version

Click to open interactive version

Share of violent deaths in prehistoric, state and non-state societies

Click to open interactive version

Click to open interactive version

Click to open interactive version

Data sources

This section collects all of the ethnographic and archaeological data sources on violent deaths presented above.

Procedures in collecting this data

The starting point for my research was the sample presented by Pinker (2011).¹ Pinker does not present the numerical data itself but instead includes two bar charts (Figures 2-2 and 2-3) from which the exact data cannot be discerned. Therefore I looked up all the sources for the data he presents.

The topic of Pinker's book is the decline of violence in humanity's history. That would give him an incentive to cherry-pick data that suits his narrative. For this reason I had contact with several of his critics and asked them for their evidence. After this assessment of Pinker's data, I did not find any reason to assume that he cherry-picked. He included all the data that his sources included.

Through my contact with critics of Pinker's book I was referred to a critique of Pinker's data by Douglas Fry.² Fry criticizes some of Pinker's data. I checked the original sources for the criticized information, referred to the criticisms and the original sources in my comments and corrected the data when the criticism was justified.

An additional source that I added to my dataset and which was not yet available to Pinker is a study published by Robert Walker and Drew Bailey (published in 2013).³

Main sources

I checked the following sources and collected the data these sources presented. For every information I also included the original source:

• Bowles (2009) - Did Warfare Among Ancestral Hunter-Gatherers Affect the Evolution of Human Social Behaviors?. In Science, 324, 5932, 1293--1298.
• Gat (2006) - War in Human Civilization. Oxford University Press, USA.
• Knauft, Bruce M. et al (1987) - Reconsidering Violence in Simple Human Societies: Homicide among the Gebusi of New Guinea. In Current Anthropology, 28, 4, 457-500.
• Keeley (1997) - War Before Civilization: The Myth of the Peaceful Savage. Oxford University Press, USA.
• Pinker (2011) - The Better Angels of Our Nature: Why Violence Has Declined. Viking.
• Walker and Bailey (2013) - Body counts in lowland South American violence. In Evolution and Human Behavior, 34, 1, 29--34.

Sources for rates of violent deaths in non-state societies

Society: Kato (Cahto) 1840s (California)

Violent deaths per 100,000 people per year: 1450

• My source for the data: Keeley (1996)
• Source that my source quoted: Kroeber 1965. "A Kato War." In The California Indians: A Source Book, ed R Heizer and M. Whipple, pp. 397-403. Berkeley: University of California Press

Society: Grand Valley Dani (New Guinea) (New Guinea)

Violent deaths per 100,000 people per year: 1000

• My source for the data: Keeley (1996)
• Source that my source quoted: Heider, K. 1970. The Dugum Dani. Chicago: Aldine.

Society: Piegan (North American Plains)

Violent deaths per 100,000 people per year: 1000

• My source for the data: Keeley (1996)
• Source that my source quoted: Livingscone, F. 1968. "The Effect of Warfare on the Biology of the Human Species." In War, ed. M. Fried, M. Harris, and R. Murphy, pp. 3-16. Garden City, N.Y.: Natural History Press.

Society: Dinka 1928 (N.E. Africa)

Violent deaths per 100,000 people per year: 970

• My source for the data: Keeley (1996)
• Source that my source quoted: Kelly, R. (1985). The Nuer Conquest. Ann Arbor: University of Michigan Press

Society: Fiji 1860s (Melanesia)

Violent deaths per 100,000 people per year: 870

• My source for the data: Keeley (1996)
• Source that my source quoted: Carniero R. (1990). "Chiefdom-Level Warfare as Exemplified in Fiji and the Cauca Valley." In The Anthropology of War, ed. J. Haas, pp. 190-211. New York: Cambridge University Press.

Society: Hewa (New Guinea)

Violent deaths per 100,000 people per year: 731

• My source for the data: Knauft et al (1987)
• Sources that my source quoted: For a first time period (1940-62) Knauft (1985) - Good company and violence: Sorcery and social action in a lowland New Guinea society. Berkeley:University of California Press. cites 778 violent deaths per 100,000 people per year. For a second time period (1959-68): Steadman (1971) - Neighbors and killers: Residence and dominance among the Hewa of New Guinea. Ph.D. diss., Australian National University, Canberra, Australia. cites 683 violent deaths per 100,000 people per year. I calculated the mean which is 731.

Society: Chippewa 1825-1832 (Minnesota)

Violent deaths per 100,000 people per year: 750

• My source for the data: Keeley (1996)
• Source that my source quoted: Hickerson, H. 1962. "The Souchwestern Chippewa: An Echnohistorical Study." American Anthropologist 64: Memoir 92.

Society: Telefolmin 1939-1950 (New Guinea)

Violent deaths per 100,000 people per year: 740

• My source for the data: Keeley (1996)
• Source that my source quoted: Morren, G. 1984. "Warfare on the Highland Fringe of New Guinea: The Case of the Mountain Ok." In Warfare, Culture, and Environment, ed. R. Ferguson, pp. 169— 208. Orlando, Fla.: Academic Press.

Society: Buin (Salomon Is.)

Violent deaths per 100,000 people per year: 710

• My source for the data: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol. 1 (abridged ed. published 1964). Chicago: University of Chicago Press

Society: Mtetwa 1806-1814 (S. Africa)

Violent deaths per 100,000 people per year: 590

• My source for the data: Keeley (1996)
• Source that my source quoted: Otterbein, K. (1967) - "The Evolution of Zulu Warfare." In Law and Warfare, ed. P. Bohannon, pp. 351-57. Garden City, N.Y.; Natural History Press.

Society: Goilala (from Papua New Guinea)

Violent deaths per 100,000 people per year: 550

• My source for the data: Gat (2006)
• Source that my source quoted: Hallpike, Bloodshed and Vengeance in the Papuan Mountains, pp. 54, 202.
• Comment: Gat (2006) writes 'among the Goilala, whose total population was barely over 150, there were 29 (predominantly men) killed during a period of 35 years'. I calculated the rate of violent deaths given assuming that the population was 150 (the lower bound of his estimate).

Society: Dugum Dani 1961 (New Guinea)

Violent deaths per 100,000 people per year: 480

• My source for the data: Keeley (1996)
• Source that my source quoted: Heider, K. (1970) - The Dugum Dani. Chicago: Aldine.

Society: Manga 1949-1956 (New Guinea)

Violent deaths per 100,000 people per year: 460

• My source for the data: Keeley (1996)
• Source that my source quoted: Pflanz-Cook, S., and E. Cook. 1983. "Manga Pacification." In The Pacification of Melanesia ed. M. Rodman and M. Cooper, pp. 179-98. Lanham, Md.: University Press of America. Vayda (1967) War in Ecological Perspective. Mew York: Plenum

Society: Modoc (California)

Violent deaths per 100,000 people per year: 450

• My source for the data: Keeley (1996)
• Source that my source quoted: Ray, V. (1963) - Primitive Pragmatists: The Modoc Indians of Northern California. Seattle: University of Washington Press.

Society: Auyana 1924-1949 (New Guinea)

Violent deaths per 100,000 people per year: 420

• My source for the data: Keeley (1996)
• Source that my source quoted: Robbins, S. 1982. Auyana: Those Who Held onto Home. Seattle: University of Washington Press.

Society: (Hunter-Gatherer) Murngin 20 years (Australia)

Violent deaths per 100,000 people per year: 330

• My source for the data: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol 1 (abridged ed. published 1964). Chicago: University of Chicago Press
• Comment: Knauft et al (1987) give the same number but cites not Wright (1942) but Warner (1969:146-47) - A black civilization: A social study of an Australian tribe. New York: Peter Smith.

Society: Tauade 1900-1946 (New Guinea)

Violent deaths per 100,000 people per year: 320

• My source for the data: Keeley (1996)
• Source that my source quoted: Hallpike, C. (1977) - Bloodshed and Vengeance in the Papuan Mountains. Oxford: Oxford University Press.

Society: Mae Enga 1900-1950 (New Guinea)

Violent deaths per 100,000 people per year: 320

• My source for the data: Keeley (1996)
• Source that my source quoted: Meggitt (1977) - Blood Is Their Argument. Palo Alto, Calif.: Mayfield.

Society: Yanomama 1938-1958 (Brazil)

Violent deaths per 100,000 people per year: 290

• My source for the data: Keeley (1996)
• Source that my source quoted: Early, J., and J. Peters. 1990. The Population Dynamics of the Mucajai Yanomamo. San Diego, Calif.: Academic Press.

Society: Yurok (California)

Violent deaths per 100,000 people per year: 240

• My source for the data: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol. 1 (abridged ed. published 1964). Chicago: University of Chicago Press

Society: Mohave 1840s (Calif.-Ariz.)

Violent deaths per 100,000 people per year: 230

• My source for the data: Keeley (1996)
• Source that my source quoted: Stewart, K. 1965. "Mohave Warfare." In 77jt? California Indians, ed. R. Heizer and M. Whipple, pp. 369-82. Berkeley: University of California Press.

Society: Gebusi 1940-1982 (New Guinea)

Violent deaths per 100,000 people per year: 200

• My source for the data: Keeley (1996)
• Comment and source that my source quoted: Knauft B (1985) Good Company and Violence. Berkeley. University of California Press. Knauft et al (1987) cite a violent death rate of 419 per 100,000 people per year.

Society: Yanomamo 1970-74

Violent deaths per 100,000 people per year: 165.9

• My source for the data: Knauft et al (1987)
• Source that my source quoted: Melancon (1982:33,42) - Marriage and reproduction among the Yanomamo Indians of Venezuela. Ann Arbor: University Microfilms.

Society: (Hunter-Gatherer) Tiwi 1893-1903 (Australia)

Violent deaths per 100,000 people per year: 160

• My source for the data: Keeley (1996)
• Source that my source quoted: Pilling, A. 1968. "Discussion: Predation and Warfare." ln Man the Hunter, ed. R. Lee and 1. Devore, p. 158. Chicago: Aldine Atherton.

Society: Boko Dani 1937-1962 (New Guinea)

Violent deaths per 100,000 people per year: 140

• My source for the data: Keeley (1996)
• Source that my source quoted: Ploeg, A. 1983. "The Establishment of the Pax Neerlandica in the Bokondini Area." In The Pacification of Melanesia, ed. M. Rodman and M Cooper, pp. 161-78. Lanham, Md.: University Press of America.

Society: Eskimos (central Canadian Arctic)

Violent deaths per 100,000 people per year: 100

• My source for the data: Gat (2006)
• Source that my source quoted: Gat (2006) is referring to: Donald Symons, The Evolution of Human Sexuality. New York: Oxford University Press, 1979, p. 145; Bruce Knauft, ‘Reconsidering violence in simple societies: Homicide among the Gebusi of New Guinea’, Current Anthropology, 1987; 28: 458; Jean Briggs, ‘ “Why don’t you kill your baby brother?” The dynamics of peace in Canadian Inuit camps’, in L. E. Sponsel and T. Gregor (eds), The Anthropology of Peace and Nonviolence. London: Lynne Rienner, 1994, p.156.
• Comment: Gat (2006) quotes Jean Briggs with the following: ‘Readers of Canadian Inuit ethnography, my own Never in Anger (1970) in particular, have sometimes concluded that Inuit are always and everywhere pacific. Nothing could be farther from the truth.’

Society: !Kung Bushmen of the Kalahari

Violent deaths per 100,000 people per year: (before the establishment of a state authority): 42 // (after the establishment of a state authority): 29

• My source for the data: Gat (2006) (Knauft et al (1987) cite the same source that my source quoted.)
• Source that my source quoted: Richard Lee, The !Kung San. New York: Cambridge University Press, 1979, p. 398.
• Comment: Quotation of Gat (2006): 'The rate for the !Kung Bushmen of the Kalahari, the famous ‘harmless people’, was 0.29 per 1,000 per year, and had been 0.42 before the coming of firm state authority'.

Society: Andamanese 30 years (Indian Ocean)

Violent deaths per 100,000 people per year: 20

• My source for the data: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol. 1 (abridged ed. published 1964). Chicago: University of Chicago Press

Sources for rates of violent deaths in state societies

Society: C. Mexico, 1419-1519 (Mesoamerica)

Violent deaths per 100,000 people per year: 250

• My source for the data: Keeley (1996)
• Source that my source quoted: Thieme, F. 1968. "The Biological Consequences of War." In War, ed. M. Fried, M. Harris, and R. Murphy, pp. 16—21. Garden City, N.Y.: Natural History Press.

Society: Mexican mestizo village (1961-65)

Violent deaths per 100,000 people per year: 251.2

• My source for the data: Knauft et al (1987)
• Source that my source quoted: Nash (1967) - Death as a way of life: The increasing resort to homicide in a Maya Indian community. American Anthropologist 66:1-19.

Society: Tepoztlán, Mexico (1922-55)

Violent deaths per 100,000 people per year: ca. 59

• My source for the data: Knauft et al (1987)
• Source that my source quoted: Lewis (1951 26-28, 227) - Life in a Mexican village: Tepoztlán restudied. Urbana: University of Illinois Press.

Society: Germany 1900-1990

Violent deaths per 100,000 people per year: 160

• This data calculated by: Keeley (1996)

Society: Russia 1900-1990

Violent deaths per 100,000 people per year: 150

• This data calculated by: Keeley (1996)

Society: France 1800-1899

Violent deaths per 100,000 people per year: 70

• This data calculated by: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol 1 (abridged ed. published 1964). Chicago: University of Chicago Press

Society: Japan 1900-1990

Violent deaths per 100,000 people per year: 30

• This data calculated by: Keeley (1996)

Society: World, 20th Century (wars & genocides)

Violent deaths per 100,000 people per year: 60

• Own calculation: I am following Pinker calculating this data: White (2011)⁴ estimates a total number of deaths of all violent deaths in the 20th-century of 180 million deaths. Pinker assumes an average annual world population for the 20th century of 3 billion. (see his footnote 65 in chapter 2).180.000.000 / 100 = 1.800.000 violent deaths per year. Divided by an annual population of 3.000.000.000 and multiplied by 100,000 is 60.

Society: US, 20th century (war deaths)

Violent deaths per 100,000 people per year: 3.79

• Own calculation: I am following Pinker calculating this data: From the US Census I know that the average population of the US over the 20th century was: 163,275,000 (The source for this data is U.S. Census.The number of 'Total Deaths of U.S. Military Personnel Serving and Casualties' for the 20th century: 619,538 This is taken from Leland & Oboroceanu (2010)⁵. According to this source these are the number of 'Total Deaths of U.S. Military Personnel Serving and Casualties':

Society: World, 2007 (battle and one-sided violence deaths)

Violent deaths per 100,000 people per year: 0.33

• Own calculation: I am following Pinker calculating this data: The Human Security Report Project. (accessed April 2013). Miniatlas of human security. (World Bank) has data on the global number of war deaths and is available online.The last year that this source has data for is 2007. These are the numbers by category:

Share of violent deaths in prehistoric societies

Site: Crow Creek, 1325 CE (South Dakota)

Share of Violent Deaths: >60%

• My source for the data: Keeley (1996)
• Comment and Source that my source quoted: Keeley (1996) writes: 'Contrary to Brian Ferguson's claim that such slaughters were a consequence of contact with modem European or other civilizations, archaeology yields evidence of prehistoric massacres more severe than any recounted in ethnography.(Footnote 23- 23. (Middle Missouri) Zimmerman 1980; Wilky 1990; Bamforth 1994; (Southwest) Haas 1990: 187, and personal communication.) For example, at Crow Creek in South Dakota, archaeologists found a mass grave containing the remains of more than 500 men, women, and children who had been slaughtered, scalped, and mutilaced during an attack on their village a century and a hdf before Columbus's arrival (ca. A.D. 1325). The attack seems to have occurred just when the village's fortifications were being rebuilt. AU the houses were burned, and most of the inhabitants were murdered. This death toll represented more than 60 percent of the village's population, estimated from the number of houses to have been about 800. The survivors appear to have been primarily young women, as their skeletons are underrepresented among the bones; if so, they were probably taken away as captives. Certainly, the site was deserted for some time after the attack because the bodies evidently remained exposed to scavenging animals for a few weeks before burial. In other words, this whole village was annihilared in a single attack and never reoccupied.'

Site: Nubia (site 117), 14-12000 years before present

Share of Violent Deaths: 46%

• My source for the data: Bowles (2009)
• Comment and sources: Bowles (2009) quotes F. Wendorf, in The Prehistory of Nubia F. Wendorf, Ed. (Southern Methodist University Press, Dallas, 1968), vol. 2, pp. 954-998. Keeley (1996) cites Wendorf (1968) - 'Site 117: A Nubian Final Palaeolithic Graveyard near Jebel Sahaba, Sudan.' In The Prehistory of Nubia, vol. 2, ed. F. Wendorf, pp. 954-95. Dallas: Southern Methodist University Press. Keeley refers to a time 10,000 B.C and quotes a share of violent deaths of 40.7%.

Site: N. British Columbia (1500 B.C.-A.D. 500)

Share of Violent Deaths: 32.4%

• My source for the data: Keeley (1996)
• Source that my source quoted: Cybulski, J. (1994) "Culture Change, Demographic History, and Health and Disease on Che Northwest Coast" In In the Wake of Contact, ed. G. Milner and C. Larsen. New York: Wiley-Liss.

Site: Sarai Nahar Rai, N.India, 3140-2854 years before present

Share of Violent Deaths: 30%

• My source for the data: Bowles (2009)
• Source that my source quoted: G. R. Sharma, Proceedings of the Prehistoric Society 39, 129 (1973).
• Comment: Here Pinker appears to have made a mistake by converting the dates from years before present to BCE/CE. Bowles gives the time period as '3140-2854 years before present' which Pinker erroneously converted to 2140-850 BCE.

Site: British Columbia (A.D. 500-1774)

Share of Violent Deaths: 27.6%

• My source for the data: Keeley (1996)
• Source that my source quoted: Cybulski, J. (1994) "Culture Change, Demographic History, and Health and Disease on Che Northwest Coast" In In the Wake of Contact, ed. G. Milner and C. Larsen. New York: Wiley-Liss.

Site: British Columbia (30 sites, 5500–334 years before present)

Share of Violent Deaths: 23%

• My source for the data: Bowles (2009)
• Source that my source quoted: Cybulski (1994) - J. Cybulski, in In the wake of contact: Biological responses to conquest C. S. Larsen, G. Milner, Eds. (Wiley-Liss, 1994) pp. 75-85.

Site: Late prehistoric Indian site of Madisonville, Ohio

Share of Violent Deaths: 22% of adult male skulls had wounds; 8 % of them were fractured

• My source for the data: Gat (2006)
• Source that my source quoted: Livingstone, ‘The effects of warfare on the biology of the human species’, p. 9.

Site: Volos’ke, Ukraine ("Epipalaeolithic”)

Share of Violent Deaths: 22%

• My source for the data: Bowles (2009)
• Source that my source quoted: V. N. Danilenko, Sovietskaya Etnografia 3, 54 (1955).
• *Comment: Pinker (2011) dates this to '±7500BCE'.

Site: Vasiliv’ka III, Ukraine; 11000 years before present

Share of Violent Deaths: 21%

• My source for the data: Bowles (2009)
• Source that my source quoted: Telegin, D. (1961) Arkheologiya XIII, 3.
• Comment Keeley (1996) has data on 'Ukraine: Vasylivka (Mesolithic)' and quotes Vencl 1991. "Interprétation des Blessures Causée par les Armes au Mésolithique" L'Anthropologie 95: 219-28. as his source. He gives a share of violent deaths of 15.9%.

Site: Nubia: Qadan burials 10,000 B.C.

Share of Violent Deaths: 21.4%

• My source for the data: Keeley (1996)
• Source that my source quoted: Wendorf, F. 1968. "Site 117: A Nubian Final Palaeolithic Graveyard near Jebel Sahaba, Sudan." In The Prehistory ofNubia, vol. 2, ed. F. Wendorf, pp. 954-95. Dallas: Southern Methodist University Press.

Site: Illinois A.D. 1300

Share of Violent Deaths: 16.3%

• My source for the data: Keeley (1996) and Gat (2006)
• Source that my source quoted: Milner, G, E. Anderson, and V. Smith. 1991. "Warfare in Late Prehistoric West-Central Illinois." American Antiquity 56: 581-603. Gat (2006) quotes a share of violent deaths of 16%.

Site: Northeast Plains 1325-1650

Share of Violent Deaths: 15%

• My source for the data: Keeley (1996)
• Source that my source quoted: Willey, P . 1990. Prehistoric Warfare on the Great Plains. New York: Garland.

Site: Denmark: Vedbaek 4100 B.C.

Share of Violent Deaths:13.6%

• My source for the data: Keeley (1996)
• Source that my source quoted: Price, T. D. 1985. "Affluent Foragers of Mesolithic Southern Scandinavia." In Prehistoric Hunter-Gatherers, ed. T. Price and J. Brown, pp. 361-64. Orlando, Fla.: Academic Press.

Site: Ile Teviec, France, 6600 years before present

Share of Violent Deaths: 12%

• My source for the data: Bowles (2009)
• Source that my source quoted: R. R. Newell, T. S. Constandse-Westerman, C. Meikeljohn, Journal of Human Evolution 8, 29 (1979).

Site: Bogebakken, Denmark, 6300-5800 years before present

Share of Violent Deaths: 12%

• My source for the data: Bowles (2009)
• Source that my source quoted: R. R. Newell, T. S. Constandse-Westerman, C. Meikeljohn, Journal of Human Evolution 8, 29 (1979).

Site: S. California: Ven-110 A.D. 100-1100

Share of Violent Deaths: 10%

• My source for the data: Keeley (1996)
• Source that my source quoted: Walker, P., and P, Lambert 1989. "Skeletal Evidence for Stress During a Period of Cultural Change in Prehistoric California." In Advances in Paleopathology, Journal of Paleopathology: Monographic Publ. no. 1, ed. L. Cappasso, pp. 207-12. Chieti, Italy: Marino Solfanelli.

Site: Brittany 6000 B.C.

Share of Violent Deaths: 8%

• My source for the data: Keeley (1996)
• Source that my source quoted: Vencl 1991. "Interprétation des Blessures Causée par les Armes au Mésolithique" L'Anthropologie 95: 219-28.

Site: Central California, 2415-1773 years before present

Share of Violent Deaths: 8%

• My source for the data: Bowles (2009)
• Source that my source quoted: V. Andrushko, K. Latham, D. Grady, A. Pastron, P. Walker, American Journal of Physical Anthropology 127, 375 (2005).

Site: Sweden (Skateholm I), 6100 years before present

Share of Violent Deaths: 7%

• My source for the data: Bowles (2009) and Keeley (1996)
• Source that my source quoted: T. D. Price, in Prehistoric Hunter-Gatherers: The emergence of cultural complexity T. D. Price, J. A. Brown, Eds. (Academic, Orlando, 1985) pp. 341- 363.
• Comment: Keeley (1996) cites Price, T . D., and J. Brown, eds. 1985. Prehistoric Hunter-Gatherers. Orlando, Fla.: Academic Press. and gives a share of violent deaths of 3.8%.

Site: S. California (28 sites), 5500-628 years before present

Share of Violent Deaths: 6%

• My source for the data: Bowles (2009)
• Source that my source quoted: P. Lambert, in Troubled Times: Violence and Warfare in the Past D. L. Martin, D. W. Frayer, Eds. (Gordon and Breach, Amsterdam, 1997) pp. 77-109.

Site: Kentucky: 2500-3000 B.C.

Share of Violent Deaths: 5.6%

• My source for the data: Keeley (1996)
• Source that my source quoted: Webb, W. 1974. Indian Knoll Knoxville: University of Tennessee.

Site: Central California 1500 B.C.-A.D. 500

Share of Violent Deaths: > 5%

• My source for the data: Keeley (1996) and Bowles (2009)
• Source that my source quoted: Moratto M. (1984) - California Archaeology. Orlando, Fla.: Academic Press.
• Comment: Bowles cites 5%.

Site: Central California (2 sites), 2240-238 years before present

Share of Violent Deaths: 4%

• My source for the data: Bowles (2009)
• Source that my source quoted: R. D. Jurmain, American Journal of Physical Anthropology 115, 13 (2001).
• Comment: Here Pinker appears to have made a mistake by converting the dates from years before present to BCE/CE. Bowles gives the time period as '2240-238 years before present' which Pinker erroneously converted to 240 - 1770 CE.

Site: Calumnata, Algeria, 8300-7300 years before present

Share of Violent Deaths: 4%

• My source for the data: Bowles (2009)
• Source that my source quoted: M.-C. Chamla, Memoires du Centre de Recherches Anthropologiques, Prehistoriques, et Ethnographiques XV, 1 (1970).

Site: Nubia (near site 117), 14-12000 years before present

Share of Violent Deaths: 3%

• My source for the data: Bowles (2009)
• Source that my source quoted: F. Wendorf, in The Prehistory of Nubia F. Wendorf, Ed. (Southern Methodist University Press, Dallas, 1968), vol. 2, pp. 954-998.

Site: Algeria: Columnata ca. 6000 B.C.

Share of Violent Deaths:1.7%

• My source for the data: Keeley (1996)
• Source that my source quoted: Vencl 1991. "Interprétation des Blessures Causée par les Armes au Mésolithique" L'Anthropologie 95: 219-28.

Site: Gobero, Niger, 16,000-8,000 years before present

Share of Violent Deaths: 0%

• My source for the data: Bowles (2009)
• Source that my source quoted: P. Sereno et al., PLoS ONE 3, 1 (2008).
• Comment: Based on 35 of the approximately 200 burials at the site.

Share of violent deaths in non-state societies

The following statistics refer to the share of violent deaths out of all deaths in a given society.

Society: The Waorani (Auca) of the Ecuadorian Amazon

Share of Violent Deaths: 60 % and 56 %

• My source for the data: Gat (2006) and Walker and Bailey (2013)
• Source that Gat (2006) quoted: J.A.Yost,‘Twenty years of contact: The mechanism of change in Wao (Auca) culture’, in N. A. Whitten (ed.), Cultural Transformation and Ethnicity in Modern Ecuador. Urbana, IL: University of Illinois, 1981, pp. 677–704; C. A. Robarchek and C. J. Robarchek, ‘Cultures of war and peace: A comparative study of Waorani and Semai’, in J. Silverberg and J. P. Gray (eds), Aggression and Peacefulness in Humans and Other Primates. New York: Oxford University Press, pp. 189–213.
• Comment: Quote from Gat (2006): The tribe 'resembles the Yanomamo in their subsistence patterns and in the causes and style of fighting'. The data is collected over five generations '60 % of adult deaths over five generations were caused by feuding and warfare'.
• Source that Walker and Bailey (2013) quoted: Larrick et al., 1979
• Comment: 60% of victims were male. Total numbers: 272 killings in 484 deaths

Society: Achuar

Share of Violent Deaths: 42%

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: J.B. Ross -- A balance of deaths: Revenge feuding among the Achuarä Jívaro of the northwest Peruvian AmazonColumbia University, New York (1988)
• Comment: 69% of victims were male. Total numbers: 106 killings in 250 deaths

Society: Arawete

Share of Violent Deaths: 35%

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: E. Viveiros de Castro -- From the enemy's point of view. Humanity and divinity in an Amazonian Society -- University of Chicago Press, Chicago (1992)

Society: Kayapo

Share of Violent Deaths: 35%

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: Werner, D. W. (1980). The making of a Mekranoti chief: The psychological and social determinants of leadership in a native South American Society. PhD dissertation. City University of New York, New York.
• Comment: 75% of victims were male, Total numbers: 83 killings in 237 deaths

Society: Enga (Papua New Guinea)

Share of Violent Deaths: 34.8% (male population)

• My source for the data: Gat (2006)
• Source that my source quoted: Meggitt, Blood is their Argument, pp. 13–14, 110.
• Comment: Meggitt had records of 34 wars among them in 50 years

Society: Gebusi (Papua New Guinea)

Share of Violent Deaths: 35.2 % (male population) & 29.3 % (female population)

• My source for the data: Knauft et al (1987)
• Source that my source quoted: Knauft (1985) - Ritual form and permutation in New Guinea: Implications of symbolic process for socio- political evolution. American Ethnologist 12:321-40.
• Comment: It is noted that the extraordinary high homicide rate for women might be explained by 'the fact that killing was mainly related to failure to reciprocate in sister exchange marriage'.

Society: Jivaro Share of Deaths due to Warfare: 32.7%

• My source for the data: Keeley (1996)
• Source that my source quoted: Ross,J. 1984. "Effects of Contact on Revenge Hostilities Among the Achuarajivaro." In Warfare, Culture, and Environment, ed. R. Ferguson, pp. 83—110. Orlando, Fla.: Academic Press.
• Comment: Male Deaths: 59% - Female Deaths: 27%

Society: (Hunter-Gatherer) The Murngin of Arnhem Land

Share of Violent Deaths: 28.6% (male population) & 21% (adult population)

• My source for the data: Keeley (1996), Gat (2006) and Bowles (2009)
• Source that my source quoted: Warner, A Black Civilization, pp. 157–8
• Comment: Warner estimated 200 violent deaths of a population of 700 men. total population was 3000 but he mentions no violent mortality of women and children. Bowles (2009) calculated the number for adult mortality due to violence/warfare. Keeley (1996) also quotes a share of violent deaths of 28% for the male Murngin population (his source is Harris 1975. Culture, People, Nature. 2d ed. New York: Crowed.)

Society: Dani (Highland Papua New Guinea)

Share of Violent Deaths: 28.5 % (male population) & 2.4 % (women population)

• My source for the data: Gat (2006)
• Source that my source quoted: Heider, The Dugum Dani, p. 128.

Society: Wari

Share of Violent Deaths: 28%

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: Conklin, 1989 -- Conklin, B. A. (1989). Images of health, illness and death among the Wari' (Pakaas Novos) of Rondonia, Brazil. PhD dissertation. University of California, San Francisco.
• Comment: 110 violent deaths of total 400 deaths

Society: Tribal Montenegro (beginning twentieth century)

Share of Violent Deaths: ca. 25 %

• My source for the data: Gat (2006)
• Source that my source quoted: Boehm, Blood Revenge, p. 177.

Society: Yanomamo-Shamatari

Share of Violent Deaths: 20.9%

• My source for the data: Keeley (1996)
• Source that my source quoted: Chagnon. 1974. Studying the Yanomamo. New York: Holt, Rinehart and Winston.
• Comment: Male Deaths: 37.4% - Female Deaths: 4.4%

Society: (Hunter-Gatherer) Ayoreo (1920-1979)

Share of Violent Deaths: 20% (all ages)

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: Bugos, P. E. Jr. (1985). An evolutionary ecological analysis of the social organization of the Ayoreo of the Northern Gran Chaco. PhD dissertation, Northwestern University, Evanston, IL.
• Comment: Based on 276 killings.

Society: Mae Enga

Share of Violent Deaths: 18.6%

• My source for the data: Keeley (1996)
• Source that my source quoted: Meggitt - 1977. Blood Is Their Argument. Palo Alto, Calif.: Mayfield.
• Comment: Male Deaths: 34.8% - Female Deaths: 2.3%.

Society: Xilixana

Share of Violent Deaths: 16%

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: J.D. Early, J.F. Peters -- The Xilixana Yanomami of the Amazon: History, social structure, and population dynamics -- University Press of Florida, Gainesville (2000)
• Comment: 80% of the victims were male, total killings 10 out of 64 deaths.

Society: Dugum Dani

Share of Violent Deaths: 15.5%

• My source for the data: Keeley (1996)
• Source that my source quoted: Heider, K. 1970. The Dugum Dani. Chicago: Aldine.
• Comment: Male Deaths: 28.5% - Female Deaths: 2.4%

Society: Yanomamo

Share of Violent Deaths: 15% (entire population) / 24% (male population) / 7% (female population)

• My source for the data: Gat (2006)
• Source that my source quoted: Mildred Dickemann, ‘Female infanticide, reproductive strategies, and social stratification’, in N. Chagnon and W. Irons (eds), Evolutionary Biology and Human Social Behavior. North Scitnate, MA: Buxbury Press, 1979, p. 364; slightly different (and inconsistent) figures in: Marvin Harris (ed.), ‘Primitive war’, in Cows, Pigs, Wars and Witches. New York: Random House, 1974, p. 69.Keeley (1996) and Walker & Bailey quote N. Chagnon 'Studying the Yanomamö' (Holt, Rinehart and Winston, New York) (1974). For the male and female population Keeley (1996) quotes a share of violent deaths of 15.3%.

Society: Modoc, Northern California

Share of Violent Deaths: 13%

• My source for the data: Bowles (2009)
• Source that my source quoted: V. F. Ray, Primitive Pragmatists: The Modoc Indians of Northern California (University of Washington Press, Seattle, 1963), pp.
• Comment: Refers to “Aboriginal times”.

Society: Huli

Share of Violent Deaths: 13.2%

• My source for the data: Keeley (1996)
• Source that my source quoted: Glasse, R. 1968. Hull of Papua: A Cognathic Descent System. The Hague: Mouton.
• Comment: Male Deaths: 19.6% - Female Deaths: 6.1%

Society: (Hunter-Gatherer) Casiguran Agta (Philippines)

Share of Violent Deaths:12% were attributed to homicide (In the less violent time period (1962-1977); adult male population). Between 1977 and 1984 the homicide rate is 326 homicides per 100,000 people per year. In this time period, that was described as unusually violent (and which is therefore not quoted in my statistic) 21% of all adult male deaths were due to homicide (see comment).

• Original source:Headland, T. (1989) - Population decline in a Philippine Negrito hunter-gatherer Society. American Journal of Human Biology, 1, 59–72.
• Comment: The source of Pinker on the percentage of deaths in warfare is Bowles (2009) and he made a mistake that Fry (2013) corrects: Bowles (2009) based his calculation on ten killings when in fact 9 killings are reported. In the time between 1936 and 1950 not 10 but 9 Casiguran Agta were killed - 1 in 1950, 5 in 1947; 1 in1938 raid and 2 “shortly after World War II”. Bowles therefore miscalculated his data and Pinker carries this wrong number over.As Headlands study of the Agta focusses on a later time (1962-84) I take the data from this better researched time. As Headland states that the latter half of his research stay (post-1977) was a time of 'unusual high violence' I quote the death rate due to violence from the earlier time (1962-77). In this more 'peaceful' time the share of violent deaths was 12% (and this also is not counting the number of killings of non-Agta by Agta - which is also high (see below)).Homicides, as opposed to deaths in warfare are very common in the Agta Population as Headland points out: 'In the sample of adult male deaths before 1977, 12% were attributed to homicide; in 1977-1984, 21% of all adult male deaths were from homicide. The overall homicide rate for the Casiguran Agta, based on the data for 1977-1984, is 326 homicides per 100,000 people per year. The 14 homicides on which this calculation is based include only homicides of members of the de jure population.' (page 68). He also notes that 'of the 33 Agta deaths by homicide since my arrival in 1962 up to June 1984, only eight were killed by non-Agta. Of the 14 Agta homicides in the de jure population in 1977-1984, only five were killed by non-Agta. Of the 53 homicides, only 30% of the victims were killed by non-Agta (16153)And he says: 'It should be noted here that Agta kill outsiders, too. I know of 24 cases of Agta killing non-Agta since 1930 and 15 cases since I began fieldwork in the area in 1962.' (page 69)

Society: Anggor

Share of Violent Deaths: 11.9%

• My source for the data: Keeley (1996)
• Source that my source quoted: Huber, P. 1973. "Defending the Cosmos: Violence and Social Order Among the Anggor of New Guinea." In War, Its Causes and Correlates, ed. M. Nettieship, R. Givens, and A. Nettieship, pp. 619-61. The Hague: Mouton.

Society: Gebusi

Share of Violent Deaths: 8.3%

• My source for the data: Keeley (1996)
• Source that my source quoted: Knauft B 1985 Good Company and Violence. Berkeley. University of California Press
• Comment: Male Deaths: 8.3% - Female Deaths: 8.2%

Society: (Hunter-Gatherer) Hiwi

Share of Violent Deaths: 7,2 % (Hiwi killed by Hiwi in Precontact times); in total 28% of Hiwi died a violent death in Precontact times

3,8% (Hiwi killed by Hiwi in Postcontact times)

• My source for the data:Hill,K.,Hurtado,A.M.,&Walker,R.S.(2007). High adult mortality among Hiwi hunter-gatherers: Implications for human evolution. Journal of Human Evolution, 52, 443–454.
• Comment: The paper distinguishes clearly between deaths inflicted by Venezuelans on Hiwis and death inflicted by Hiwis on Hiwis. Table 4 (page 450) shows the statistics on both groups and makes an additional distinction between precontact and postcontact times. In the precontact times 153 deaths are tabulated. There were 11 killings of Hiwi by Hiwi in the precontact times. Out of the 160 killings in the postcontact times 6 were killings of Hiwi by Hiwi. The Source that my source quoted states: "The Hiwi mortality profile is characterized by notably high rates of violence and accidental trauma" (page 449)The numbers I quoted and the share of violent deaths is not counting infanticides that were very common among the Hiwi (and are reported by Hill, Hurtado (2007)).

Society: Tsimane

Share of Violent Deaths: 6% (adults)

• My source for the data: Walker and Bailey (2013)
• Source that my source quoted: M. Gurven, H. Kaplan, A. Zelada Supa Mortality experience of Tsimane Amerindians: Regional variation and temporal trends - American Journal of Human Biology, 19 (2007), pp. 376–398
• Comment: 73% of the victims were male, total killings 30 out of 525 deaths

Society: (Hunter-Gatherer) Tiwi

Share of Violent Deaths: 5.75% (entire population; average of arid and non-arid areas) -- 10% (men only) - 5% (entire population - arid areas) - 6,5% (entire population - non-arid areas)

• My source for the data: Gat (2006)
• Comment and Source that my source quotes: Gat (2006) quotes death rates from Pilling and Kimber and refers to three different sources: Warner, A Black Civilization, pp. 157–8; Pilling, in Man the Hunter, p. 158; R. G. Kimber, ‘Hunter-gatherer demography: The recent past in Central Australia’, in B. Meehan and N. White (eds), Hunter-Gatherer Demography. Sydney: University of Sydney, 1990, p. 163.Gat (2006) refers to Pilling saying: 'Pilling’s estimate of at least 10 per cent killed among the Tiwi men in one decade.'Gat (2006) refers to Kimber saying: 'Kimber’s estimate, for a generation, of 5 per cent mortality in fighting in arid areas and about 6.5 per cent in well-watered ones refers to violent mortality in relation to the entire population’s overall mortality rates.' Pinker quotes the 10%, which according to Gat (2006) only refers to the male population. I decided to quote the average of the two shares specified for the entire population. That is 5.75%.

Society: (Hunter-Gatherer) Anbara, N. Australia

Share of Violent Deaths: 4%

• My source for the data: Bowles (2009)
• Source that my source quoted: L. R. Hiatt, Kinship and Conflict: A Study of an Aboriginal community in Northern Arnhem Land (Australian National University, Canbera, 1965), pp.

Share of violent deaths in state societies

Society: Ancient Mexico

Share of Violent Deaths: 5%

• My source for the data: Keeley (1996)
• Source that my source quoted: Thieme, F. 1968. "The Biological Consequences of War." In War, ed. M. Fried, M. Harris, and R. Murphy, pp. 16—21. Garden City, N.Y.: Natural History Press.

Society: France 19th Century

Share of Violent Deaths: 3%

• My source for the data: Keeley (1996)
• Source that my source quoted: Thieme, F. 1968. "The Biological Consequences of War." In War, ed. M. Fried, M. Harris, and R. Murphy, pp. 16—21. Garden City, N.Y.: Natural History Press.

Society: Western Europe 17th Century

Share of Violent Deaths: 2%

• My source for the data: Keeley (1996)
• Source that my source quoted: Wright Q. 1942. A Study of War. Vol. 1 (abridged ed. published 1964). Chicago: University of Chicago Press

Society: U.S. and Europe 1900-1960

Share of Violent Deaths: less than 1% (Male Deaths only)

• My source for the data: Keeley (1996)
• Source that my source quoted: Harris 1975. Culture, People, Nature. 2d ed. New York: Crowed.

Society: World, 2007 (battle and one-sided violence deaths)

Share of Violent Deaths: 0.04%

• My own calculation:* The Human Security Report Project - Miniatlas of human security (accessed April 2013) has data on the global number of war deaths. It is available onlineThe last year that we have data for is 2007. These are the numbers by category:Number of State-Based Battle Deaths: 16773 Number of Non-State Battle Deaths: 1865Number of One Sided Violence Deaths: 3501The total sum of the above is: 22139 deathsThe world population in 2007 was 6.658.000.000 (World Bank) - The death rate of the world is approximately 8 per 1000 per year according to the CIA Factbook. This means around 53264000 people died in 2007. Therefore the ratio of people dying due to warfare in 2007 is 22139/53264000. This means 0.04% of the global deaths were due to warfare in 2007.

Max Roser (2013) - “Data review: ethnographic and archaeological evidence on violent deaths” Published online at OurWorldinData.org. Retrieved from: 'https://ourworldindata.org/ethnographic-and-archaeological-evidence-on-violent-deaths' [Online Resource]

@article{owid-ethnographic-and-archaeological-evidence-on-violent-deaths,
    author = {Max Roser},
    title = {Data review: ethnographic and archaeological evidence on violent deaths},
    journal = {Our World in Data},
    year = {2013},
    note = {https://ourworldindata.org/ethnographic-and-archaeological-evidence-on-violent-deaths}
}

Reuse this work freely

All visualizations, data, and code produced by Our World in Data are completely open access under the Creative Commons BY license. You have the permission to use, distribute, and reproduce these in any medium, provided the source and authors are credited.

The data produced by third parties and made available by Our World in Data is subject to the license terms from the original third-party authors. We will always indicate the original source of the data in our documentation, so you should always check the license of any such third-party data before use and redistribution.

All of our charts can be embedded in any site.