In the forest near Wamba, a village in the Democratic Republic of the Congo, some of the last remaining bonobos breed, feed and lounge in the trees. Like other great apes, these animals have a rich social life, communicating with their fellows using some 80 types of gestures. Primatologist Kirsty Graham of the University of St. Andrews in Scotland has spent hundreds of hours among this screeching, scratching endangered troupe to decode its members’ nonverbal interactions. This work has confirmed, for example, that when one of the animals repeatedly swipes the black fuzz on its chest, it is begging to be groomed. And when it cups its hand under another’s chin, it is asking for food.
Graham’s St. Andrews colleague Catherine Hobaiter built a similar body-language dictionary by observing the East African chimpanzees at the Budongo Central Forest Reserve in Uganda. The gestures of both species, which are humans’ two closest relatives, are more complex and varied than their vocalizations, which mainly reflect urgent needs such as finding food or spotting predators.
By contrast, the apes’ gestures serve as a deliberate way of conveying specific everyday goals, leading some scientists to believe that these signals are the precursors to human language. “They are using gestures in a way that is more languagelike, and so there’s this theory that human language might have evolved from this gestural basis,” Graham says.
In a paper published today in PLOS Biology, Graham and Hobaiter provide startling evidence that this ancestral ability may persist in modern humans. They show that our species can make a pretty good guess of the meanings of chimp and bonobo gestures, another hint that language may have evolved from an elaborate system of hand and body signals.
In the study, when thousands of people watched online videos of wild apes raising an arm, scratching and striking various poses, they got the gist of the animals’ lingo far more often than would be expected by chance. “Humans without any training and without seeing any of the outcomes or surrounding behaviors can understand what chimpanzee and bonobo gestures mean,” Graham says.
The finding suggests that humans still have some grasp of this ancestral vocabulary. “Maybe this is something that was shared with our last common ancestor and that we, in fact, retain, this ability to understand and use the great ape gestures,” Graham says.
The work fills a hole in the case for a shared linguistic lineage. Scientists have learned that great ape vocabularies overlap extensively: about 95 percent of bonobos’ gestures are the same as those chimpanzees use. What’s more, in a 2019 pilot study, Hobaiter and her colleagues found that the vast majority of gestures used by toddlers younger than two years old, who do not yet primarily use spoken or sign language to communicate, are the same as those of chimpanzees.
But it has been difficult to find any such gestural overlap between wild apes and humans once the latter reach adulthood. Adult human communication is dominated by spoken or sign language, along with a huge variety of gestures, many of which are culturally specific. “Unpicking whether we still have access to that great ape gesture communication is really difficult just by observing people,” Graham says.
A strategy to address this question came to Graham and Hobaiter several years ago, as they were collecting and analyzing ape videos. The researchers deciphered the meaning of each ape gesture by looking at what happened after. For instance, if an ape bent its back knee and raised its foot, and then its child hopped on its back, they might conclude that the foot lift meant “I’ll give you a ride.” The decoding process took years and involved sifting through thousands of examples of such behaviors. But all along, the scientists had a sense that they just got what the animals were saying. “We are putting all of this time into understanding them,” Graham says, “but always in back of our minds, it was like, ‘We know what the gestures mean.’”
What the researchers didn’t know was whether ordinary people who did not spend large parts of the year hanging out with apes had the same intuition. In 2017 Graham and Hobaiter decided to find out. They designed what they expected to be a small pilot study in which lay people would try to identify chimp gestures from videos in an online quiz. Each test-taker was shown 20 short clips of chimps or bonobos making gestures and asked to decide which of four possible answers described each gesture’s meaning. In one of the clips, a bonobo gave its comrade a shove. In this case, the animal was saying, “Climb on my back.” But when a chimp did the same, it was telling its companion to move to a new spot.
After the media picked up the team’s work, more than 17,000 people logged on to watch the videos. The researchers excluded participants who didn’t watch all the clips, who saw the clips more than once or who said they had experience with primates, leaving 5,656 scores. “It’s a really impressive piece of citizen science,” says Erica Cartmill, an anthropologist at the University of California, Los Angeles, who was not involved in the study. “The number of people that they have looking at ape gestures is fantastic.”
If people had no idea what the apes were doing, they would be right 25 percent of the time by chance, correctly choosing one of the four answers. But the average score was slightly above 50 percent, a statistically strong result, given the study’s size. “What is interesting is that people seem to be able to do [this task], and somehow their guesses, while not being perfect, are definitely above chance,” says Federico Rossano, a comparative psychologist at the University of California, San Diego, who was not involved in the study. Humans’ ability to make sense of the behaviors of pets, such as cats and dogs, is more limited, he says. “Some people are very good at interpreting pets, but many others are not,” Rossano says. “Indeed, you might get bit.”
The study participants even understood ambiguous ape gestures, those with more than one possible meaning in ape society, with one exception: When a chimp shakes an object such as a branch, it often means “Let’s have sex,” but it could mean “Move away from me,” depending on the context. People picked the correct meaning for this gesture, which the researchers called “object shake,” no better than chance. And in general, accuracy rates in the study varied quite a bit by the type of gesture. “That tells you that things that are very specific to the language of the chimps, we don’t [understand],” says Thibaud Gruber, a primatologist at the University of Geneva, who was not involved in the study.
But the achievement of the test takers shouldn’t be underestimated. Apes get important context for interpreting ambiguous gestures by living with one another. “Given that people don’t get any of that in these videos, it is really striking that they can still understand the gestures,” Graham says.
Why humans may understand the ape gestures remains to be discovered. The human and chimp lineages are separated by up to six million years. One possibility is that all great apes, including humans, inherit a common set of gestures. Another is that humans and other great apes share the ability to use body movements as communication tools, what Graham calls “embodied communication.”
A third explanation is that the similarity in body shape among humans and other great apes, combined with humans’ cognitive abilities, makes it easy for people to infer meaning from ape movements. The study leaves these possibilities unresolved, Cartmill says. She calls the work “a beautiful first step” showing that humans can recognize the intent and meaning of the communications of another species. But she wonders, “Is this because we are very good at making inferences, or is it because there is some common underlying gestural system?”
“I believe gesture had a strong role to play in the evolution of language,” Cartmill continues. “I think this paper contributes to that story and helps open up new possibilities of there being either a deep-rooted set of gestures or a deep-rooted sensibility to seeing, perceiving and understanding gesture.”
But Rossano is not convinced that humans and other great apes share an innate repertoire of gestures. “Humans can recognize the meaning of a bark from a dog or the roar of a lion as a threat, yet this does not mean we share barking or roaring with them as a communicative tool,” he said.
Even if Rossano is right, the study itself still has its own value as a teaching tool. Engaging so many people with this research has benefits beyond the science, Gruber says. “What is really cool is the methodology and how people engaged with it,” he says. “That makes them understand ‘Oh, we’re so really close with our closest relatives.’ That, for me, is a win. It’s a win for conservation. It’s a win for showing how important it is to keep these guys in our mind, to protect them, to save them.”