Bonobo Kanzi Tracks Imaginary Juice in Tea Party Tests

Researchers have found that a bonobo named Kanzi can follow imaginary objects during make-believe tea parties. The 43-year-old ape lives at a research center in Iowa and took part in tests with scientists from Johns Hopkins University and other places. The work came out in a science journal this month and shows apes may think in ways once seen as human-only. The tests happened over several sessions where Kanzi sat at a table with empty cups and jars.

Background

Bonobos are close relatives to humans, sharing about 99 percent of our DNA. They live in forests in central Africa, but Kanzi has spent his life in human care. Born in 1983, he learned to use a board with symbols called a lexigram to communicate. He knows over 300 symbols and responds to spoken words too. Stories from his keepers said he played pretend games before, like chasing make-believe friends or eating fake blueberries from photos. But no one had tested this in a strict way until now.

Scientists long thought pretend play was special to humans. Kids start it around one year old, pouring air into cups or talking to stuffed animals. By age three, they build whole pretend worlds. This skill helps with big thinking tasks, like planning or solving problems by picturing what is not there. Experts called these 'secondary representations' – mental pictures of things that are not real. Past signs of pretend play in apes were just stories, like a chimp using a stick as a baby. No hard proof existed that apes really knew the difference between real and fake.

Kanzi met one researcher, Amalia Bastos, for the first time. He used his lexigram to ask her and a helper to chase each other in pretend play. He watched with joy, even though no one ran for real. This sparked the idea for tests. Bastos worked with Christopher Krupenye, a psychologist at Johns Hopkins. They set up games to check if Kanzi could track pretend items.

Key Details

The main tests used a table with two empty glasses or jars. In the first setup, a researcher pretended to pour juice from an empty clear jug into both glasses. Kanzi likes juice, so they picked that. The researcher then acted like she poured one glass back into the jug. She asked Kanzi, 'Where is the juice?' He pointed to the glass that still had the pretend juice. He got it right 68 percent of the time over many trials. Chance would be 50 percent, so this stood out.

Checking Real vs. Pretend

To make sure Kanzi knew pretend from real, they did another test. One glass had real orange juice, the other was empty but got pretend juice poured in. When asked which he wanted, Kanzi picked the real one 80 percent of the time. This showed he saw the difference and still played along with the fake stuff.

They repeated the idea with pretend grapes. The researcher acted like she put a grape in one of two jars, then moved it around in pretend ways. Kanzi pointed to the right jar 68.9 percent of the time, correct even on the first try. Over 45 no-reward trials, he stayed steady. In mix-in training trials with real rewards, he scored near perfect, like 99 percent. No signs of upset, like shaking fences or scratching, when he pointed to empty spots.

The team ruled out other ideas. Kanzi did not learn by reward because pretend trials gave nothing. He did not copy the researchers, who poured but he just pointed. Looks or smells did not trick him. He stayed motivated the whole time.

"I was amazed and thrilled to see him perform so well across all of the experiments!" – Christopher Krupenye, Johns Hopkins University

Kanzi is special because he understands spoken words and points or uses symbols to answer. The tests built on his skills, with researchers talking to him during the play.

What This Means

This work gives clear proof that at least one ape can handle pretend objects in shared games. It means the skill to picture unreal things may go back millions of years to when humans and bonobos split from common ancestors, around six to nine million years ago. Kanzi grew up around humans, so his skills might come from that setting. Still, it opens doors to check wild apes or other animals.

Pretend play links to bigger mind powers, like thinking ahead or guessing what others think. If apes do this, it changes how we see animal smarts. Kids use pretend to learn social rules and solve problems. Apes might do the same in their groups. The study pushes against old views that humans alone imagine beyond the moment.

Researchers now plan more tests. They want to try other apes, maybe without human training. Questions remain about future thinking or full pretend stories. Kanzi's success came from simple setups with his language skills. Broader tests could show if this is common in apes.

The findings highlight bonobos in the wild face threats from logging and hunting. Understanding their minds may help people care more about saving them. Kanzi's tests ran smoothly because he enjoyed the games. He chased pretend friends and tracked invisible food without mix-ups. This points to a shared spark in primate brains for play that blurs real and make-believe.