Azara Blog: Why Apes and Humans Kill

Blog home page | Blog archive

Google   Bookmark and Share

Date published: 2005/02/04

The third lecture of the Darwin Lecture Series 2005 was by Richard Wrangham, about why apes and humans kill. It's possible that there might be reasons due to evolution by natural selection. A traditional scientific explanation would be that killing was the result of an escalating conflict that went too far.

He mentioned two views of human nature. The first was the "Rousseau" view, whereby killing was:

and so most humans were innately reasonable and "nice" and killing was a freak occurrence.

The second view of human nature was the "Hobbes" view, whereby killing was:

and so killing was just part of human nature.

As a small support for the "Rousseau" view he quoted a claim that "1% of people in the USA commit 50% of the murders".

To investigate the question further he looked at other species. Apparently there is a species of spider monkey, the Muriqui, in Brazil, which has been observed for many years and there has been no evidence of violence and even the traditional male dominance hierarchy has not been detected (to some laughter he said that even with mating the males would hang around and practically say to each other "after you").

On the other hand chimpanzees, one of the closest relatives of humans, were observed to have regular violence, including fatal fighting. So the question became, why do chimpanzees kill?

He then went through some of the evidence of chimpanzee killing. In particular he was interested in killing between chimpanzee communities (not inside one community). Jane Goodall was the first to observe chimpanzees killing chimpanzees in Gombe starting in the 1960s/1970s. The killings resulted from lone individuals being caught by a group of male chimpanzees from another community in a surprise "raid", who then beat it to death in a gang frenzy. Apparently one of the chimp communities was eliminated largely due to such raids (at least six males were killed this way).

Males did the killings and males were usually the victims.

He showed part of a film showing such a raid in another reserve (Kanyawara in Uganda). The killing was all over in ten minutes. One chimp each held down the victim's arms and legs and then the others took turns biting, beating and ripping the victim.

Wrangham said there was long-term research about five sites where the chimpanzees had neighbouring communities (otherwise there was no likelihood of such killings), with eleven communities and more than 180 years of data. In this time there was direct evidence of 33 killings and another 16 deaths were suspected of being such killings.

There were also intragroup killings, but these were mainly of infants because of male dominance attacks. (He did not say, but presumably similarly to what happens, for example, with lions when a new male takes over a group.)

He said that for a release of chimps by humans into the wild in the Congo, 40%-50% of the males were attacked by wild chimps (and would have died were it not for human intervention). A similar thing happened with a release of chimps in Senegal.

So the bottom line was that such killings were "natural" with chimps, supporting the "Hobbes" view.

Wrangham then turned to the causes of the killing. Apparently chimps like to hang out in large groups, only resorting to small parties when there was not much food, and in that case they would avoid territorial boundaries. (Presumably this is due to wanting to avoid being the subject of attack.)

The killing attacks happened only when there was a clear imbalance of power (lots of males on one side, not so many on the other) and in this case the attackers were never seriously hurt. The attacks happened because there was an opportunity to kill. There was usually an element of surprise.

He showed a map of the territory of one chimpanzee community. There was an inner core, where they felt comfortable spending the night, and there was an outer section, where they would sometimes hunt for food during the day, but usually only in groups with a reasonable number of males. There were many observations of conflicts between communities, even in the inner core, but these were different in that usually one side backed down without any serious injury on either side. The "home" side often "lost" these conflicts (i.e. had to back down) and the "winner" was usually the side with more males in the particular conflict. There was no element of surprise in these conflicts.

He quoted a study when the recording of an alien chimp was played and the probability of a group calling back was highly correlated with the number of males in the group. With only one male in the group there was a low (almost zero) probability of reply, but by the time there were four or more males there was a high (almost one) probability of reply.

Of course chimp communities benefit in the long term from eliminating their neighbours, because they get a bigger territory with more food and this creates an opportunity for having faster reproduction (observed). But Wrangham said the killing attacks offered no obvious short term reward. (Well, it presumably offers some social cohesion.)

With intergroup killing the violence was used as a strategy:

Wrangham wanted to assure the audience it was not just primates who behaved this way. Wolves have also been observed to have similar behaviour, with both defensive but particularly offensive territoriality killing. He quoted one study in Alaska where 39%-65% of deaths were killings by other wolves. Wolf groups were unstable, wolves were often alone, and when there were meetings between groups there was frequently a power imbalance.

And some primates were not like chimps. He mentioned bonobos (apparently another close relative of humans). Here it seems there have been no observed kills or attempts at kills. The groups were stable, bonobos were rarely alone, and when there were meetings between groups there was no large-power imbalance.

What was the relevance to war? He looked at nomadic hunter-gatherer groups, in particular hunters who lived in the world of other hunters. Apparently nomadic hunter-gatherers lived in communities of typical size 25-40 but were part of tribes of typical size circa 500, and the territories were defended not by communities but by tribe.

He quoted one recent study which said that "all the evidence indicates very high killing rates among all known simple hunter-gatherer societies" (which he then gave some caveats to). The killings between groups were:

He quoted some small-scale societies where there were also ritualised battles, which were rarely lethal (so no element of surprise).

He then moved onto states. Apparently with some state wars there are no battles, just lots of killing (decolonialism, drug wars, guerilla wars). And some state battles involved "safe" killing, safe because of a temporary imbalance of power in the killing zone (e.g. three against one) and with helpless victims (e.g. apparently the largest death of soldiers in battles is when one side is running away).

So the following kill while safe: chimps, social carnivores (e.g. wolves), hunter-gatherers, small-scale societies and states. But states also kill while at risk. There are real battles with "unsafe" killing. Apparently most soldiers avoid killing and find killing traumatic. Soldiers fight in fear (of being punished by their own side or killed by the opposition) and alchohol has been used throughout human history to promote fighting. States unlike chimps conduct lethal battles (when both sides risk being killed).

Wrangham quoted some statistics about the percentage of a population killed in intergroup aggression per year. The numbers quoted were 0%-1.4% for chimps and foragers, 0%-1% for farmers and 0%-0.2% for industrial states (and the latter was for 1900-1990 when there were two world wars, so perhaps surprisingly low). (One would think one has to be a bit suspicious of such studies since the data must be patchy and states are so big that there is a lot of intragroup killing.)

He said it was clear that chimps were much worse behaved than humans, with almost daily intragroup aggression in chimp communities. The fighting rate within communities for chimps was estimated to be 150 to 1000 times that for humans.

So the claim is that humans behave in the "Rousseau" view at home and in the "Hobbes" view abroad. (Obviously simplistic but a nice strap line.)

Why do humans get into lethal battles? Perhaps this is partly due to the safety of leaders (e.g. Bush will never himself fight in Iraq; he won't even appear in London without 10000 policemen protecting his every move). But Wrangham thinks it is also due to the positive illusions leaders give themselves about their ability to win. (Which begs the question, is there an evolutionary reason for this?) He quoted the Bay of Pigs, but managed to avoid mentioning Iraq, which of course is the relevant example today. (Wrangham is at Harvard so perhaps there is some necessary self-censorship here to avoid labels of traitor by the Republican scum. Or perhaps it was all just too obvious.)

All material not included from other sources is copyright For further information or questions email: info [at] cambridge2000 [dot] com (replace "[at]" with "@" and "[dot]" with ".").