Human violence
Traditional societies exhibit much variation in their frequency and forms of death by human violence, which usually ranks as either the leading or (after illness) the second-leading cause of death. A significant factor underlying this variation is the degree of state or outside interference in suppressing or discouraging violence. Types of violence can be somewhat arbitrarily dichotomized into either war (discussed in Chapters 3 and 4) or homicide, where war in its clearest form is defined as collective fighting between different groups, while homicide is defined as killings of individuals within a group. However, this dichotomy becomes blurred when one has to decide whether killings between neighboring groups usually on friendly terms should be counted as in-group homicide or out-group war. Further ambiguities involve which types of killings to count: for instance, published tabulations of Ache violence include infanticide and senilicide, but published !Kung tabulations don’t, and different authors hold different opinions about the frequency of infanticide among the !Kung. The choice of victims, and the relation between the victim and the killer, also vary greatly among societies. For example, Ache victims of violence were mainly infants and children, while !Kung victims were mainly adult men.
!Kung studies of violence are instructive for several reasons. Initial accounts of the !Kung by anthropologists described them as peaceful and non-violent, so much so that a popular book published in 1959, early in the history of modern !Kung studies, was entitled The Harmless People. During three years of residence among the !Kung in the 1960s, Richard Lee observed 34 fights leading to blows but no killings, and informants told him that there actually were no killings during those years. Only after Lee had been in the area for 14 months and come to know his informants better were they willing to talk to him about past killings. When they did start talking, by cross-checking accounts of different informants Lee was able to assemble a reliable list of names and sex and age of killers and victims, the relation between killer and victim, and the circumstances, motive, season, time of day, and weapons used for 22 killings between 1920 and 1969. That list did not count cases of infanticide and senilicide, which Lee believed to be rare, but Nancy Howell’s interviews of !Kung women suggest that infanticide did occur. Lee concluded that those 22 cases represented the total number of deaths by violence in his study area between 1920 and 1969.
All 22 of those !Kung killings should surely be considered homicides rather than wars. In some cases the victim and the killer were within the same camp, while in other cases they were in different camps, but no killing involved a group of people from one camp seeking to kill a group of people from another camp (i.e., “war”). In fact, there was no reported event at all suggestive of a war among the !Kung in Lee’s area during the period 1920–1969. But the !Kung did say that they used to have raiding expeditions, apparently similar to the witnessed “wars” of other traditional peoples, during the generation of the grandparents of the oldest living !Kung—i.e., before Tswana herders began making annual visits to the !Kung and trading with them in the 19th century. We saw in Chapter 4 that visits of traders to the Inuit also had the effect of suppressing Inuit war, even though neither the traders with the Inuit nor those with the !Kung purposely suppressed war. Instead, the Inuit themselves abandoned war in their own self-interest in order to have more opportunities to profit from trade, and the !Kung may have done the same.
As for the rate of !Kung homicide, 22 killings over the course of 49 years works out to less than 1 homicide every 2 years. That sounds utterly trivial to readers of urban American newspapers, who can open the newspaper on any randomly chosen day and read about all the murders committed in their city within the last 24 hours. The main explanation for this difference is of course that the base population within which murders can occur is millions of people for an American city, but only about 1,500 people for the !Kung population surveyed by Lee. Referred to that base population, the homicide rate for the !Kung works out to 29 homicides per 100,000 person-years, which is triple the homicide rate for the United States and 10 to 30 times the rates for Canada, Britain, France, and Germany. One might object that the United States calculation excludes violent deaths in war, which would yield a higher rate for the United States. However, the !Kung rate also doesn’t include deaths in !Kung “wars” (i.e., their raiding expeditions that ended over a century ago), whose number is completely unknown for the !Kung but is known to be high for many other traditional peoples.
The figure of 22 !Kung homicides in 49 years is instructive for another reason as well. One homicide every 27 months means that, for an anthropologist carrying out a field study of a people lasting one year, the odds are against any homicide occurring during that period, and the anthropologist would consider the people as peaceful. Even if the anthropologist were resident there for five years, a period long enough for a killing to be likely to occur given !Kung homicide rates, the killing would be very unlikely to take place under the eyes of the anthropologist, whose assessment of the frequency of violence would depend on whether his informants chose to tell him about it. Similarly, although the United States ranks as the most homicidal society in the First World, I have never personally witnessed a homicide, and I have heard only a few first-hand accounts of homicides within my circle of acquaintances. Nancy Howell’s calculations suggest that violence was the second-leading cause of !Kung death, behind infectious and parasitic diseases, but ahead of degenerative diseases and accidents.
It is also instructive to consider why violent deaths ended recently among the !Kung. The last homicide reported to Lee occurred in the spring of 1955, when two !Kung men killed a third !Kung man. The two killers were arrested by the police, put on trial, and jailed, and did not return to their home area. This event occurred only three years after the first instance in which the police intervened to jail a !Kung killer. From 1955 until Lee published his analysis in 1979, there was no further homicide in his study area. This course of events illustrates the role of control by a strong state government in reducing violence. That same role also becomes obvious from central facts of the colonial and post-colonial history of New Guinea in the last 50 years: namely, the steep decrease in violence following establishment of Australian and Indonesian control of remote areas of eastern and western New Guinea respectively, previously without state government; the continued low level of violence in Indonesian New Guinea under maintained rigorous government control there; and the eventual resurgence of violence in Papua New Guinea after Australian colonial government gradually yielded to less rigorous independent government. That tendency for violence to decrease under state government control does not deny the fact that traditional societies have non-violent means of resolving most of their disputes successfully before the disputes become violent (Chapter 2).