Can bonobos use their imagination?
In a series of tea party-like experiments, a bonobo was able to identify imaginary juice and grapes.
The ability to pretend and imagine scenarios that aren’t real may seem uniquely human. But new research reveals that our closest living relatives, bonobos, can also understand the concept of imaginary objects.
Researchers set up a series of tea party-like experiments with Kanzi, a language-trained bonobo who had provided decades worth of scientific knowledge before his death last year. Their findings, published today in the journal Science, reveal that Kanzi could consistently demonstrate an understanding of pretend objects.
Scientists say the new findings provide compelling evidence that at least some other primates are capable of viewing the world through an imaginative lens.
“It shows us that we're not the only animals out there with rich mental lives,” Says Christopher Krupenye, an evolutionary cognitive scientist at Johns Hopkins University and a coauthor of the new paper. Given that bonobos are endangered in their home of the Democratic Republic of Congo, he added, “My hope is that that kind of insight compels people to care for these animals.”
(How female bonobos team up to gain power over males)
How to test pretending in primates
Humans begin experimenting with imagination through pretend play, when a child conjures or taps into an imagined state, known as a secondary representation, that is different from the reality in front of them. Researchers have long thought that only humans can concoct imagined scenarios.
Anecdotally, scientists have observed captive juvenile chimpanzees pushing imaginary blocks across the floor and wild female chimpanzees cradling sticks and logs like infants. However, these peculiar behaviors had never been explored in an experimental setting, says Amalia Bastos, a comparative psychologist at the University of St. Andrews in Scotland and co-author on the new bonobo study. She and colleagues wanted to find out once and for all: Do these anecdotes suggest apes are merely good at imitation or responding to researchers’ prompts? Or could they actually engage in make-believe?
Researchers turned to the legendary bonobo Kanzi for answers. Kanzi was born at the Yerkes Field Station at Emory University in 1980, then moved to Georgia State University and finally to the Ape Initiative research center in Des Moines, Iowa, in 2005.
Over his 44 years of life, Kanzi provided groundbreaking scientific insights into primate cognition, early tool-making and the evolutionary origins of language. “We were quite star-struck by Kanzi, because he's really special in our field,” Bastos says.
The brilliant bonobo could even respond to verbal prompts and communicated with human researchers through lexigram symbols and pointing. This made him an ideal test subject. “We could test him in a way that's almost identical to how you would test a small child for imagination,” Krupenye says. This particular study was conducted when Kanzi was 43, prior to his death in March 2025.
The researchers ran Kanzi through a series of experiments that resembled a tea party. One of the researchers sat across a table from Kanzi and placed an empty pitcher and a pair of transparent cups between them. The scientist then “poured” imaginary juice into each empty cup before pretending to dump one back into the pitcher. They then asked Kanzi, “Where’s the juice?”
Kanzi showed a knack for finding fake juice. Despite the researchers swapping which cup contained the imaginary juice, Kanzi correctly pointed at the cup containing imaginary liquid in 68 percent of the trials. To confirm that the bonobo knew the cups did not actually contain juice, the researchers ran him through a second series of trials where Kanzi was asked to choose between an empty cup and a glass containing orange juice. The bonobo routinely chose the real juice.
The team also recreated the first experimental set-up with pretend grapes that were placed in a pair of jars. After the researcher “emptied” one of the jars, Kanzi was asked to find the remaining grape. He successfully pointed at the fictitious fruit in 69 percent of the trials. While the ape was never perfect during the trials, his consistency tracking different pretend objects throughout the experiments illustrated his ability to pinpoint things that are not truly there, Bastos says.
(Why these bonobos are 'playing airplane')
Is it only Kanzi?
According to Christine Webb, a researcher at New York University who studies social behavior in nonhuman primates and was not involved in the paper, the findings provide “compelling evidence” that bonobos can utilize secondary representation—in other words, understand what it means to pretend. “This study adds to a substantial and growing body of evidence demonstrating that our closest primate relatives are capable of complex cognitive processes once thought to be uniquely human,” she says.
The new findings suggest that the ability to imagine pretend scenarios and objects may date back between 6 and 9 million years, when the last common ancestor of humans and bonobos lived. It also raises the possibility that apes can utilize imagination for other purposes, like visualizing possible futures or perceiving the thoughts of others. Bonobos have already shown a knack for inferring ignorance in others: In 2025, Krupenye and his team found that Kanzi and other bonobos could tell when humans did not know the location of hidden treats.
Still, Kanzi may not be representative of all—or even any—wild bonobos. After all, the primate prodigy spent his entire life in research facilities interacting with human scientists. That’s why Natalie Awad Schwob, a comparative psychologist at Bucknell University who studies primate cognition, is curious if bonobos besides Kanzi also possess an imaginative streak.
“I would love to see data from more typically reared bonobos and chimpanzees to see if this is something special Kanzi could do, or if other bonobos can [also] track and represent pretend objects,” says Schwob, who has worked with Kanzi in the past but was not involved in the new study.
Krupenye agrees about the need to test the imaginative capabilities of other apes and thinks that it is possible to craft an experimental set-up without verbal cues that cater to less conversant bonobos.
Still, Kanzi’s contributions will be sorely missed. “He offered us direct access to the mind of a bonobo because he was so richly communicative with humans,” Krupenye says. “Throughout his life, we’ve learned so much not only about the minds of apes but also what makes human minds unique.”
