Unlike humans, chimpanzees only punish when they’ve been personally wronged

By Ed Yong | August 27, 2012 3:00 pm

When Delta Airlines refused to let Arijit Guha board a plane because his T-shirt made passengers uncomfortable, others made Delta aware of their outrage. When Samsung infringed Apple’s copyright, a jury of independent peers awarded Apple more than $1 billion in damages. When Republican Todd Akin claimed that women could stop themselves from becoming pregnant if raped, people called for his head.

These recent events all illustrate a broad human trait: we seek to punish people who do wrong and violate our social rules, even when their actions don’t harm us directly. We call for retribution, even if we have nothing specific to gain from it and even if it costs us time, effort, status or money to do so. This “third-party punishment” is thought to cement human societies together, and prevents cheats and free-riders from running riot. If you wrong someone, and they’re the only ones who want to sanction you, the price of vice is low. If an entire society condemns you, the cost skyrockets.

Do other animals do the same thing? It’s not clear, but one group of scientists believes that our closest relative – the chimpanzee – does not. Katrin Riedl from the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany found that chimpanzees will punish individuals who steal food from them, but not those who steal food from others. Even if the victim was a close relative, the third party never sought to punish the thief. These were the first direct tests of third-party punishment in a non-human animal, and the chimps got an F.

Riedl’s conclusion is stark: “In contrast to humans, chimpanzees do not engage in third-party punishment.” This behaviour, so crucial to our stable societies, seems to have evolved after our ancestors split away from the other apes. “This might explain, in part, how we have been uniquely able to form large-scale societies of unrelated individuals,” says Keith Jensen, who was involved in the study. “It can allow for cooperation to move beyond simple tit-for-tat and allow for norms of cooperation, and institutions to enforce and sanction them.”

Frans de Waal, a primate researcher from Emory University, is more circumspect. “I agree that [the experiments] set up a situation in which chimpanzees could potentially perform third-party punishment,” he says. They didn’t, but why? “Negative evidence has its place in science, but is to be treated with circumspection as there are many reasons why certain behaviours may not occur. Only one of those reasons is that the capacity is lacking.” Absence of evidence is not evidence of absence. “That’s always a worry,” says Jensen, “but then science works on disconfirming, rather than proving, hypotheses.”

There is some other evidence for third- party punishment in the wild. In social insects like ants and bees, some workers police their nestmates and clamp down on reproduction attempts outside the royal line, even though those workers can’t breed either. In cleaner fish, which pick parasites off larger animals, males will attack their female partners if they bite the client’s flesh instead. Closer to us, chimps and pig-tailed macaques both have dominant individuals who will intervene in fights between subordinates.

But in all these cases, there could be selfish rather than cooperative motives. The cleaner fish could just be preventing their partners from annoying their clients. The policing chimps could be asserting their dominance, protecting females, or trying to make allies. “Perhaps,” says Jensen, “it is just because fighting is annoying.” As ever, reading into the motivation of animals is not something to be lightly done at first glance.

Riedl wanted a more direct test. She worked with 13 chimps who took turns playing an actor, a thief and a victim. They sat in three cages surrounding a central space, so they could see each other but not interact directly. In the central space was a Plexiglas puzzle box with food inside. The “victim” chimp was closest, and could move five sliders to get the food to fall down to the bottom of the box.

Then, Riedl pulled on a rope to open a flap in the victim’s cage, allowing it to finally reach the food. She also opened another flap that gave the thief access to a rope. If it pulled, it could tug the food away from the victim. And it could get away with it too, if not for the pesky actor. The actor also had a rope, which opened a trapdoor in front of the thief’s cage and sent the stolen food plummeting away. The actor could never reach the food itself but he could punish the thief for their transgression against the victim.

But they didn’t. Some of them did collapse the trapdoor, but they were just as unlikely to do so in response to third-party theft as they were when the experimenter visibly took the food away. Riedl expected that dominant chimps should punish third parties more than subordinates, since they are the ones who police wild interactions. Not so. She also wondered if close kin would show third-party punishment, since animals are expected to behave more altruistically towards others who share large proportions of their genes. Nope. Not that either.

The only consistent use of punishment was by dominant chimps, castigating subordinates who steal their food directly. As Riedl writes: “Chimpanzee punishment is of the “might makes right” variety.” It’s all about personal harm, and positions of power.

That contrasts sharply to what humans do. Forget the hand-waving complaining about broken modern societies: For the most part, humans adhere to social norms and frown on others who break them. Even three-year-old children show such sensitivities. What does it mean that the chimps in Riedl’s experiments did not?

It could mean that chimps generally don’t. After all, food is a prized commodity in chimp life, and the theft of food would be a punishable offence if ever there was one. Riedl concludes that third-party punishment “would not seem to be an ancestral feature of the last common ancestor to humans and chimpanzees.” That’s not so radical a conclusion. There are many things that unite us with our closest relatives, but it’s not controversial to say that there are differences.

For example, chimps, cooperative though they are, are less cooperative than us. Primatologist Sarah Hrdy imagines what would happen if she was on a plane, and all the people quietly sitting in each other’s company suddenly turned into chimps. She says, “Any one of us would be lucky to disembark with all their fingers and toes still attached… Bloody earlobes and other appendages would litter the aisles.”

The question is where the differences between human and chimp behaviour lie. Michael Tomasello, who led Riedl’s study, has long emphasised those differences. His studies have found: that children share when they work together but chimps do not; that they do not care about fairness (or rather, are “insensitive to inequity”); and that they prefer to work alone to get food rather than working together (unlike children). (This is not to cast Tomasello as some professional downplayer of chimp cooperation: He has also found that chimps help each other to get food and other items (although less readily than humans), and that they’re smart enough to deduce the inferences of their peers.)

On the flipside, Frans de Waal has been more likely to emphasise the similarities we share with chimps. He has found that chimps spontaneously help each other, and that they cooperate as readily as humans if they’re put to the same tasks. About Riedl’s new study, he says, “It may not relate well to natural chimpanzee social life. How often are these apes in a situation to remove another’s food without obtaining it for themselves? That they don’t punish, doesn’t mean they do not have a capacity to do so.”

“From a long line of research we have learned to be cautious when chimpanzees do not show a certain outcome that we humans expect,” De Waal adds. Negative experiments initially argued that chimps don’t spontaneously help each other out, and don’t imitate each other’s behaviour, but better studies later provided evidence for both.

De Waal now wants to see what humans would do in the same situation, using the equivalent tests that I wrote about last year. “Until we test the same set up on uninformed humans, we don’t know,” he says. Jensen says that this particular version was “custom-designed for captive chimpanzees” but they have adapted it for human children. The results are awaiting publication.

Meanwhile, he says, “If chimpanzees (and other animals) have third party punishment, why do we see so little evidence of it and no evidence of large-scale cooperation. Perhaps that is a more interesting question.”

Reference: Riedl, Jensen, Call, Tomasello. 2012. No third-party punishment in chimpanzees. PNAS http://dx.doi.org/10.1073/pnas.1203179109

Images via Lionsgate (sort of) and Possumgirl2

More on chimp/human behaviour:



Comments (12)

  1. Vervets have been described as having vendettas (work of Cheney and Seyfarth).

  2. Mike from Ottawa

    “Absence of evidence is not evidence of absence.”

    The use of that saying is one of my pet peeves. I strongly expect you mean ‘Absence of evidence is not necessarily conclusive evidence of absence.’ but that doesn’t exactly roll trippingly off the tongue. I may be nitpicking but the easy repetition of false sayings doesn’t help in thinking about the world.

    Folk who’ve had the experience of hearing that saying from a Biblical literalist using it to justify the story of Noah’s flood may share my peeve.

    Not like you invented it, Ed, and not like it diminishes the general excellence of you science writing.

  3. Tk

    It seems to me there’s a lot of problematic interpretation going on with the labels “victim” and “thief.” If I were the “actor” chimp, all I would see is the experimenters playing favorites. The so-called “victim” is actually a lucky chimp who gets to play with a box that has food in it. Meanwhile, the hungry “thief” is just sitting there, wondering how come she doesn’t get any. So, when given the chance, why shouldn’t the she claim her fair share? Why shouldn’t she deserve that food just as much as the “victim”? Even if the “victim” had the first chance to get the food, I don’t have to think that it’s rightfully hers. When the “thief” takes the snack, I say, “Way to go; stick it to the researchers’ pet!”

    The point is, how do we know this experiment is really testing for third-party punishment at all? Do the researchers offer any justification for their interpretation of the situation?

  4. Kyle

    What about Bonobos? They cooperate more than chimpanzees.

  5. Charles Sullivan

    Don’t chimps sometimes engage in ‘punishment’ when third-party violence is involved, especially if the victim is kin or close cohort?

    Food distribution is only one area where third-party ‘punishment’ could take place.

  6. SP

    Ed, it is unreasonable of me to ask more of your wonderful site, but I found myself with unanswered questions regarding this research. In particular, I wondered if it had been established that the “actor” was aware that they had the opportunity to punish and would have taken it if they had been the direct victim of the theft. The following aside in your write up made me think that the research may have addressed this point “The only consistent use of punishment was by dominant chimps, castigating subordinates who steal their food directly” but I could not be sure. I know I should look for myself, but I thought you might find the feedback useful. Thanks for everything.


  7. Arno

    The question of third-party punishing seems to be quite closely related to the rules of the society. Maybe chimps do punish transgressions against social norms, too, but simply see this experiment not as an instance of theft, but just of clever acquisition of food? Different human societies have quite different rules…

  8. Lilly

    Like Arno said, maybe what we consider as stealing food is really not a punishable offence for chimps. They certainly seem to be operating on a basis of “winners, keepers” for other things. Maybe the chimps who fight the perpetrators when food is directly stolen from them, are not punishing them at all: maybe they’re just fighting for food.

  9. Todd S.

    “This behaviour, so crucial to our stable societies, seems to have evolved after our ancestors split away from the other apes”

    Hmm, I would say it is crucial to establishment of states, not societies. In other words, it’s more important to political theory than social.

  10. Susan Durham

    @Kyle, I was thinking the same thing, without wanting ANY bonobos to be subjected to becoming lab-animals. I’m not a Peta, but reading this makes me think these chimps may be socially retarded because of their living circumstances.

    Remember when we thought chimps didn’t hunt or eat meat? I think the days of saying chimps do or don’t do this or that may be coming to an end, along with keeping them in relative isolation to use as rats for comparison with humans.

    And by the way, which humans are we comparing them to? The normal ones? Are they next to the unicorns?

    And furthermore, chimps DO preemptive warfare, as when they will kill lion cubs that are born too close to their territory, which is something that most experts on war advise AGAINST! And that’s all I have to say about THAT!

  11. jake

    Interesting point about third party punishment and might makes right.
    People often defend the concept of anarchy as being a natural state of being- but does anyone really consider what might makes right means? Sure if you are a single unattached healthy male then let the good times roll, but for the rest of society- no one wants or likes living in a free-for-all.
    Pro-anarchists claim that societies will naturally form and create laws and protection for people. Sure does, but then guess what? you are back to having a government.

    Sorry if this seems like a tangent- but the topic at hand was parallels between human and chimp society right?

  12. Kim Harrell

    I’d be interested to know if it would make any difference at all if the “actor’s” punishment would also return the food to the “victim”; that is, would apparently righting the wrong appeal to the chimp any more than third-party punishment alone? Maybe they would like to see someone get food as opposed to no one? Just curious.

    Oh, and “Vervet Vendettas” would be a great name for a band, to paraphrase Dave Barry. : )


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar