Do Chimps Have Culture?
An entourage of subordinate chimps is gathered eagerly around Steward, the big alpha-male chimpanzee, at the Yerkes National Primate Research Center outside of Atlanta. They're watching Steward brandish a Plexiglas stick at an oblong polycarbonate box sitting behind a wide mesh fence - a rudimentary candy vending machine. As the subordinates look on, he liberates M&M after M&M from the contraption - called Pan-pipes after the chimpanzee genus, Pan - designed to make chimps work for their treats.
He's doing so by raising an internal trap door with the Plexiglas stick. In the upper pipe on the Pan-pipes is a piece of candy, trapped by a blockage (see first image below). Once the blockage is removed, the candy...
Victoria Horner, a postdoc at Yerkes' Living Links Center, keeps the Pan-pipes stocked with fresh M&Ms. A steady rhythm develops: the click of Horner introducing an M&M into the back of the Pan-pipes, the clunk of Steward lifting the internal gate, and the tinkle of the treat rolling down the shoot into his hand. Steward scores at least a dozen M&Ms. His entourage looks on, rapt but patient.
Tool use is well documented in both wild and captive chimpanzees. So, the fact that Steward can operate the Pan-pipes using a simple tool isn't groundbreaking. What is remarkable, however, is how Steward obtained this skill. His human caretakers didn't teach him this relatively complex behavior, nor did he figure out the lift on his own. Steward learned the secret of the Pan-pipes from one of the very chimps huddled around him: alpha-female Ericka.
Most of the subordinate chimps surrounding him are expert lifters as well. They learned the lift after Horner taught Ericka the behavior, in collaboration with Andrew Whiten, a psychologist at Scotland's University of St. Andrews, and Frans de Waal, a primatologist at Yerkes. Horner also taught a high-ranking female in another group of Yerkes chimps to use the poke technique to push treats from the Pan-pipes. It has been six months since Steward has seen the Pan-pipes, and he's sticking with lifting.
"Imagine what the poke group looks like, and then you've got your two cultures," says Horner, watching Steward. "I think these guys are very representative of chimps in the wild." She adds, "They're doing things that nobody thought that they could, and they're learning with this high fidelity."
The Yerkes chimps may be doing an unprecedented, or at least undocumented, thing by sharing and maintaining specific behaviors within social groups. Some researchers say that's evidence of culture among apes. But is it?
Jane Goodall, the renowned primatologist, was one of the first scientists to rigorously draw a link between human and chimp behavior when she reported, in a 1964 Nature paper,1 the manufacture of twig tools used for termite foraging in Tanzanian chimpanzees. Then in the 1970s, William McGrew and Caroline Tutin noted a curious hand-clasp (see Primate customs ) performed during grooming bouts in one Tanzanian chimpanzee community. That behavior hadn't been seen in Goodall's Gombe Stream Reserve chimps living only 170 km to the north and separated by recently deforested tracts. Because both the hand-claspers and the Gombe chimps were of the same subspecies, Pan troglodytes schweinfurthii, and because they inhabited similar ecologic niches, McGrew and Tutin wrote in a 1978 Man paper,2 that the grooming hand-clasp was a "social custom."
-Frans de Waal
It was the most robust evidence yet for an established chimpanzee behavior that was not obviously mediated by genetic or ecologic differences between chimpanzee groups. It was also a seminal moment for the field of "cultural primatology."
In the decades of field observation that followed, researchers described more behaviors that seemed to be transmitted socially through chimpanzee populations. Descriptions of regionally varying behaviors, such as nut cracking and using bunches of leaves as sponges to sop up water from tree hollows or as napkins to clean muddy feet, built the case for a cultural dimension to wild chimpanzee life.
This accumulation of observations culminated in a 1999 Nature paper3 in which Whiten and a collection of other leading field primatologists compiled decades of field data to identify 39 behaviors, or "traditions," that vary across African chimpanzee communities in tell-tale patterns. They wrote that those patterns "resemble those in human societies."
The authors of the paper didn't go as far as to propose the exact mechanism by which these traditions were maintained and spread through chimp communities, but they did suggest that some form of cultural or social transmission was likely at play. And, they equated the resulting patterns in chimpanzee behavioral traditions to what humans call culture. "It was the first time that the claim was made that one could distinguish different [chimpanzee] communities on the basis of a whole rich and complex profile of multiple traditions," says Whiten, "which I think then holds up a mirror to what we think of as culture in the human case."
Not everyone agrees, however, that chimpanzee behavior can be framed in the same context as human culture. The paper's premise sparked a heated debate that continues today. "If you want to call what chimpanzees have culture, then we'll have to find a new word for what it is that humans have, because they're just totally different from each other," says Bennett Galef, former president of the Animal Behavior Society. "It's reflected in the fact that we're building cathedrals and walking on the moon, and they still sit naked in the rain."
Galef, a psychologist at McMaster University in Canada who has studied social learning's role in the behaviors of animals such as Norway rats and Japanese quail, says that field primatologists prematurely affixed the term culture to the behavioral patterns they observed in wild chimpanzees. But de Waal, who was not an author on the 1999 Nature paper, defends the claim. "In science we push limits, you know? We say, 'Let's call it culture because that's what we think it is,' and then others can try to debunk it," de Waal says. "You can call it provocative."
And provoke it does. The claim of chimpanzee culture fuels arguments from those who consider human culture a distinct and unparalleled evolutionary advance. The basis for this contention is twofold: One is a disagreement over definitions, and the other is a perceived lack of empirical data supporting chimpanzee culture.
Defining the word culture becomes problematic when scientists from anthropologists and sociologists to psychologists and primatologists throw their hats into the ring. Primatologists tend to focus on a pared-down definition of culture, emphasizing a fundamentally biological approach, while social scientists tend to define the term using specific mechanisms of cultural acquisition and transmission, such as language, teaching, or symbolic mediation.
University of Chicago anthropologist Russell Tuttle, for example, stresses symbolism as a defining aspect of culture. He suggests that chimpanzee behaviors, such as nut cracking, should not be called cultural until their inherent symbolism can be demonstrated. "If they were doing it because that rock and that nut and anvil represented something," says Tuttle, "then it would be culture."
Horner, however, defines culture simply as the "transmission of behavior by nongenetic means." This would make nut cracking, if it is indeed learned socially in wild chimps, a cultural behavior. de Waal shares Horner's more inclusive definition, saying that it places chimpanzees on the same cultural continuum as human beings. "You can emphasize the unique parts, but I still think that essentially human culture would fall under what we define as culture, which is behavior that is socially transmitted," he says. "I think it's based on the same learning mechanisms (that you learn from others in social learning), and that it has the same results - that groups do different things. In humans we definitely call that cultural variation."
Tuttle and Galef both say that they avoid using the word "culture" to describe chimpanzee behavior. "Why in the world do you have to insist that whatever it is that chimpanzees do is called culture?" asks Galef. "I'm perfectly happy to call it tradition in those cases where it seems likely that it's been learned socially." Tuttle shares Galef's sentiment. "You want to call it culture? Fine," he says. "I'll call it learned behavior patterns."
Semantics aside, critics of Whiten et al.'s 1999 Nature paper also argue that observational evidence was too scant at that time to support the claim that chimpanzees were sharing and maintaining behaviors as part of a culture. "I personally believe that one should be fairly conservative in attributing human-like properties to animals," says Galef. "I don't mean to be overly critical, but the primatologists have a way of claiming that they've shown something which they haven't yet shown, and then spend years collecting data that suggested they were right in the first place."
Due to the inherent limits of chimpanzee field observation, primatologists could provide little more than circumstantial evidence that behavioral patterns among wild chimp communities resulted from social transmission. They could not easily observe this transmission at work, pinpoint the mechanisms mediating the transmission, or conduct active experiments that might elucidate these facets of chimpanzee behavior.
For example, transplant experiments, in which a chimp from one community is moved to another community with a different set of traditions, are virtually impossible to conduct due to ethical and logistical considerations and the nature of chimpanzee intergroup aggression. This left a gap between documenting sets of wild chimpanzee traditions that resembled cultures and providing evidence that these traditions were passed along socially, and not the result of undetected genetic or ecologic differences.
Given the limitations of field-based experimentation, some focus shifted to captive chimpanzee groups, like Steward's, to elucidate fundamental facets of chimpanzee behavior and provide convincing experimental evidence for the assertion that chimps do indeed have the capacity to develop distinct cultures. Enter the Pan-pipes study, published in Nature in 2005.4
Horner, Whiten, and de Waal, the authors of the study, wanted to use a scenario that tracked socially transmitted behavioral variation and approximated natural foraging. They came up with the Pan-pipes because it's close to tasks such as using a stick to fish for ants, which wild chimps perform with slight variations in technique between groups.
In the Pan-pipes study, while Steward and Ericka's group was learning to lift the Pan-pipes, Yerkes' other chimp group was learning to poke treats from the apparatus (Kix cereal, not M&Ms, was used in the actual study). Horner trained Georgia, a high-ranking female in the poke group, who then demonstrated the behavior for the chimps in her group. It spread among them just as the lift behavior had through Steward's group. Though in both the lift and poke groups some individuals experimented with the alternate technique, the researchers found a high fidelity for the specific behavior each group learned from their respective models. Individuals who did poke among lifters or lift among pokers tended to revert back to the dominant group behavior when the chimps were retested with the Pan-pipes two months after their initial exposure.
A similar experiment,5 published earlier this year, also tracked an introduced behavior as it diffused through chimp groups. In this study, Yerkes' two chimp groups learned that depositing tokens in receptacles resulted in food rewards. As in the Pan-pipes study, the groups learned different, but equally effective, reward-garnering tactics. The chimps received a treat by depositing tokens in either blue or white garbage cans. While Georgia showed her group how to get treats by dropping tokens in the blue garbage can, Ericka was demonstrating how putting tokens in the white garbage can could yield the same reward. Though all chimps were rewarded for placing a token in either receptacle, the two groups again showed an overwhelming tendency to stick to the specific behavior that was originally seeded in their community.
According to de Waal, the fidelity for socially entrained behaviors shown in these two experiments represents a previously undemonstrated propensity for conformity in chimps, which may be related to a major driver of human culture: the urge to fit in. Such results expand the developing picture of chimpanzee social capabilities and bolster the claims of field primatologists studying chimpanzee culture. "Ten years ago, when we started, the prevalent opinion was that chimps have no imitation and have very limited capacities for social learning. I think what we have shown is that you can actually artificially introduce cultures and you get reliable transmission. So this whole story about chimpanzees being limited in that regard, I think we can say, has been disproved," he says. "I think that's a very strong finding."
Owing to its experimental design, the token study provided an additional insight into the chimpanzee's capacity for culture. Unlike the Pan-pipes experiment, which tracked a behavior that directly resulted in a food reward, the token study examined a behavior that was not, in and of itself, directly tied to getting a treat. While the Pan-pipes can be viewed as an artificial termite mound, the behavior explored in the token study is closer to arbitrary human behaviors such as working for a paycheck that one uses to buy groceries.
Kristin Bonnie, a PhD student in de Waal's group and lead author of the token study, says that depositing tokens into a garbage can for a food reward parallels seemingly arbitrary wild chimpanzee behaviors. Some wild chimps, for example, noisily tear leaves between their teeth as a way to initiate sex or play with a companion. "We've seen those observations in the wild. We assume that they're learned through social learning in some way, but it's hard to observe social learning in the wild," Bonnie says. "This study really built off of that to show that they can learn to do things with objects that are seemingly arbitrary and have no meaning until somebody does something with it to give it a meaning."
Bonnie says that chimps participating in her token study displayed the kind of learning that many feel is essential to the development and spread of culture in chimps or in humans. "It's complex learning because they have to generalize," she says. "They have to do this multistep process that involves going to look for a token, picking it up, bringing it back, putting it in the right place, and then looking up at me standing in the tower and waiting for that reward to come."
Results from both these studies have supported the claim that wild chimpanzees share and maintain behavioral patterns that approximate human culture. "What we have established here is that you can get reliable transmission of behavior in chimpanzees," says de Waal. "That supports the claims of the field workers, who say that the variation we see is probably due to social learning." Insights garnered from field work must be complemented and substantiated by controlled experiments in a captive setting, de Waal says. "Each study has its limitations, but if you put them all together, you get a pretty strong picture."
It's now Socrates' turn to perform. The alpha-male of Yerkes' other captive chimp community sits at his enclosure's fence, reaching through the chain link to grab the top of a dusty-blue plastic garbage can and bang it against the ground. A white garbage can with a plastic tube protruding from its top sits next to the blue one, but Socrates pays it no mind. He continues to bang the blue garbage can with flicks of his hairy wrist and seems to be urging Bonnie to get the demonstration started.
Bonnie, with two buckets in hand, ascends the wooden observation tower that rises above Socrates' enclosure. One bucket contains halved bananas and the other contains tokens, which are actually short segments of PVC pipe, painted bright orange. Bonnie flings tokens into the chimps' enclosure, one by one, scattering the orange tubes among the wooden climbing structure, concrete culverts, and truck tires on which the chimps normally cavort.
Socko (as Bonnie calls him) ambles over and picks up two tokens, stashes one in his mouth, and hurries back to his position in front of the blue garbage can. Taking a seat, he drops one of the tokens into a hole in the garbage can, and it rattles to the bottom with a thump. Socko tilts his head up towards the tower, fixing his gaze on Bonnie and the bucket in her hand. Bonnie lobs him a banana, Socko eats it, and then pops his second token into the garbage can. Again a banana sails down to him from the tower. As Socko relishes his second tossed banana, more chimps venture out from their indoor enclosure or rouse from shaded naps to brave the mounting Georgia humidity and collect their own tokens.
They appear to share and maintain clever behaviors within their communities, but are the Yerkes chimps really uncovering some of the fundamental aspects of culture? Moreover, are these attributes of chimpanzee behavior related to human culture? After all, they're using socially shared foraging techniques that differ only slightly from techniques employed by another conspecific group.
"If you eat with chopsticks and I eat with knife and fork, we call that a cultural difference, because we know it's not in your genes to eat with chopsticks," says de Waal, director of the Living Links Center. "So we call that cultural variation in humans, and for the same reason, we use that term [for chimpanzees]."
Though no suite of ape behaviors can match the rich tapestry of human cultural traditions, chimps like Steward and Socko may be revealing a primitive form of cultural behavior that originated somewhere in our shared evolutionary history.
"I think we all assume, in primatology at least, that our ancestors started out with the kind of culture that chimpanzees have," de Waal says. "We started adding on to it and expanding it, so it is all a continuum."