Chimpanzees weigh evidence before making decisions, just like humans

Chimpanzees are not just grabbing at food on instinct. In a new set of experiments, they changed their choices when stronger clues appeared and stayed the course when only weaker hints came in.

Working with dozens of chimpanzees at a forested sanctuary in Uganda, researchers staged simple games with boxes and hidden apple slices. 

By watching how the apes reacted as new information arrived, the team showed that they could use evidence to reach decisions that lined up with formal models of rational choice.

Searching for missing clues

Humans are famous for weighing evidence, pausing, and sometimes changing our minds. That skill depends on metacognition, the capacity to think about one’s own thinking.

The work was led by Hanna Schleihauf, a psychologist at Utrecht University (UU). Her research focuses on how humans and other primates reflect on what they know and adjust their decisions accordingly.

In 2025, Schleihauf’s team published a large study showing that chimpanzees revised their beliefs about where food was hidden only when new clues were clearly stronger than the ones they already had. 

This pattern matched a mathematical model that predicts when a perfectly rational decision maker should stick or switch.

Earlier experiments with language trained chimpanzees found that they checked a container before naming food when they lacked key information in front of them, according to one experiment on information seeking. 

Those results suggested that great apes sometimes notice gaps in their own knowledge and actively look for the missing piece.

A chimpanzee test of rational choice

The new research took place at Ngamba Island Chimpanzee Sanctuary (NICS), which cares for 53 apes rescued from the illegal wildlife trade.

The animals sleep in secure enclosures at night, roam a patch of tropical forest by day, and can choose whether to take part in research sessions.

For each trial, a chimp sat behind a barrier facing two opaque boxes, only one of which contained a slice of apple. 

The experimenter provided a clue about which box held the reward, sometimes by letting the chimp see the apple through a glass panel and sometimes by shaking a box so something inside rattled.

When chimps first saw the apple through a clear window, they almost always chose that box and continued to back it even after they later heard the rival box rattle. 

When they first heard a rattle from one box, however, then later saw an apple in the other, they abandoned the earlier choice and moved to the box backed by stronger visual evidence.

“Chimpanzees were able to revise their beliefs when better evidence became available. This kind of flexible reasoning is something we often associate with 4-year-old children. It was exciting to show that chimps can do this too,” said Emily Sanford, a researcher in the UC Berkeley Social Origins Lab (UCB). 

Weighing strong clues against weak ones

The researchers also tried a setup where the visual clue was barely helpful. Chimps saw only a faint hint of food in one box, while the other made a clear rattle when shaken, and they chose the box with the stronger sound.

Across conditions, the apes treated a clear sight of food as the most reliable cue, a loud food-like rattle as the next best, and faint visual traces as the weakest. 

Their choices showed that they were not just reacting to one sense but ranking different sources of information by reliability.

In another condition, chimps saw a piece of apple in one box, then heard a single apple drop into the second box, and still chose the one they could see. 

After the experimenter dropped a second piece into the hidden box, so that it now likely held more reward, they switched to that box, as if tallying up total evidence on each side.

When the researchers replaced the real glass-panel view with a picture of an apple, the chimps realized the clue was fake and switched to the other box. 

Metacognition in animals

Psychologists call this kind of updating belief revision, changing what you think when new evidence arrives and conflicts with earlier reasons.

The chimpanzees did not merely chase the newest clue but seemed to compare how much support each side had overall.

The team fitted their data with a probabilistic model that predicts the best supported choice after each new hint arrives. 

For years, critics argued that great apes might solve earlier metacognition tasks with simple habits, such as always checking containers they have not seen inside, without truly monitoring their own uncertainty. 

Another research project had already complicated that view by showing that apes flexibly adjusted how often they peeked into tubes depending on how costly checking was and whether they had alternative cues like sound. 

“All of this suggests they’re not just driven by simple, emotional responses. They have rather complex awareness,” said animal cognition researcher Christopher Krupenye. He noted that the patterns are hard to explain with simple stimulus-response rules alone.

An ancient evolutionary story

These findings do not erase the differences between ape and human reasoning, but they do narrow the psychological gap.

The chimps behaved as if they could represent an option, consider why they believed it, and then change course when a better supported alternative emerged.

“This is really what makes humans so special. We give and ask for reasons,” said Schleihauf. According to the research team, the crucial human twist may involve social reasoning, where people ask each other for explanations and trade arguments. 

The evolutionary story behind this ability likely stretches back millions of years. If chimps can evaluate conflicting evidence in this way, then some ingredients of human-style rationality were probably present in the last common ancestor we share with them and with bonobos.

Studying intelligent animals

The researchers point out that sanctuaries play a key role in studies like this. The chimps at Ngamba Island move freely in their groups, eat normally, and choose whether to join the sessions. 

Their behavior reflects animals living in good conditions rather than controlled lab settings. This approach offers a model for studying intelligent animals without sacrificing welfare standards.

Future work will adapt the same tasks to bonobos and to young children, testing how far this kind of reflective reasoning extends across species and development. 

The study is published in the journal Science.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–