Read our COVID-19 research and news.

Children as young as 3 showed a grasp of justice and punishment by taking treats from a "thief" puppet sitting at the table and returning them to a "victim" puppet who originally had the object.

Children as young as 3 showed a grasp of justice and punishment by taking treats from a "thief" puppet sitting at the table and returning them to a "victim" puppet who originally had the object.


Even 3-year-olds have a sense of justice

More than one parent has likened their 3-year-old to a chimpanzee—the noise, the unbridled energy, even the occasional biting. But kids have one big leg up on their ape cousins: They’re much more likely to help victims of injustice, according to a new study. The findings indicate that even young children may have advanced moral reasoning.

Scientists know that chimpanzees have a sense of justice—but it has limits. In lab experiments, for example, the primates will punish a chimp that takes food from them by triggering a trapdoor that makes the food disappear. But they don’t step in and inflict similar punishment when they see a food theft between two other chimpanzees.

Such “third-party” punishment is considered vital for maintaining cooperation in human societies. People are more likely to follow the rules if they are enforced by everyone, not just the victims of a misdeed. So if chimps don’t do it, what about young humans?

To find out, scientists at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, gave 168 3- and 5-year-old German children the power to punish after witnessing an injustice between two other parties—in this case, two childlike puppets. In the experiment, a child was seated at a round table divided into four quadrants by 25-centimeter-tall clear plastic walls. One puppet sat in front of the section to their left and the other sat in front of the section directly across the table. On the child’s right was a covered space dubbed “the cave.” The child could pull on a rope to make the table rotate, and the puppet opposite them had a control rope as well.

The children witnessed a variety of scenarios in which the puppet opposite them wound up with a treat, such as a cookie or a marble game. Sometimes that puppet “took” the treat from the child by making the table rotate to move the item from the child to the puppet. Other times this puppet took the treat from the “victim” puppet to the child’s left.

The children had the option to intercede by rotating the table. In some variations, they could only rotate the table so that the treat went into the cave, where nobody could have it. In other cases, they could rotate the table more freely, shifting the treat to whomever they chose.

To the scientists’ surprise, the children were almost as likely to intercede when the theft didn’t directly involve them. The 3-year-olds turned the table roughly half the time when the treat was taken from them and approximately 40% of the time when they witnessed stealing between the two puppets. The 5-year-olds took action nearly 80% of the time for themselves, and roughly 70% of the time on behalf of the victim puppet in response to a theft, the team reports online today in Current Biology.

But the research suggests kid justice is a bit different from adult justice. The 3-year-olds intervened just as much in situations modified so that the thief puppet had less of a role in the moving of the treat. In addition to direct theft, scientists wanted to see what children did when an additional party stepped in. So in some cases they added a third “stranger” puppet not seated at the table. This stranger puppet would move the treat to the “thief” puppet, or to a vacant spot at the table where no one sat. The 3-year-olds interceded in equal measure in all three scenarios, returning the treat to where it started.

Adults might see these on a sliding scale from worst (the thief keeps the treat) to least bad (the victim loses the treat, but no one gains), says Keith Jensen, an expert in primate and child psychology at the University of Manchester in the United Kingdom and an author of the study. For instance, a person in a restaurant would probably be more incensed if a neighboring diner grabbed a plate of food from another nearby table than if a waiter moved the plate to a vacant table. The child seemed more focused on the common fate of the victim puppet, which lost the treat in each scenario.

This was part of a broader theme, in which the children—particularly the 3-year-olds—were more focused on helping the victim than punishing the perpetrator. Given the option, the children preferred to rotate the table to return the treat to whoever first had it. In situations where they could only put the treat in the cave or leave it with the perpetrator, some were frustrated to the point of tears, Jensen says.

“When all you can do is take it away from the puppet who took it, this is a very tense situation for the children,” Jensen says. “In the restoration condition it seemed much more calm for them.”

The study is the first to find this grasp of third-party punishment in children so young. And it offers intriguing differences from earlier work. A 2014 study, for example, found that 5-year-olds didn’t show an urge to intervene when another child split candy unevenly with their partner. Six-year-olds, however, did.

Katherine McAuliffe, a developmental psychologist at Yale University and an author of the earlier study, praises the Leipzig experiment as elegant and simple enough for children to grasp. In her research, children were presented with an explanation of how the candy had been divided, but they didn’t witness it firsthand. That could explain the difference in the findings. In the new study, “they’re watching it get stolen,” she says. “So that might be much more salient to the children.”

But people shouldn’t be so quick to draw a bright line between chimps and kids, says Frans de Waal, a primate behavior expert at Emory University in Atlanta and author of the book The Bonobo and the Atheist: In Search of Humanism Among the Primates. He says proving the absence of something—like a grasp of third-party punishment—is difficult. And artificial studies like the chimp one in Leipzig can obscure normal primate behavior. “If one chimp attacks another, third parties may intervene and take sides. This is common behavior. So, chimps do pay attention to how others are being treated,” he writes in an e-mail.

What’s not clear is whether the behavior among the German children is universal for kids this age, or is influenced by the culture in which they are raised. But developmental psychologists are beginning to look more closely at such questions. “More and more people are starting to take these games cross-culturally. But we don’t have much data from these,” McAuliffe says. “So watch this space.”