At the very least, the new experiment reported in Science is going to make people think differently about what it means to be a “rat.” Eventually, though, it may tell us interesting things about what it means to be a human being.

In a simple experiment, researchers at the University of Chicago sought to find out whether a rat would release a fellow rat from an unpleasantly restrictive cage if it could. The answer was yes.

The free rat, occasionally hearing distress calls from its compatriot, learned to open the cage and did so with greater efficiency over time. It would release the other animal even if there wasn’t the payoff of a reunion with it. Astonishingly, if given access to a small hoard of chocolate chips, the free rat would usually save at least one treat for the captive — which is a lot to expect of a rat.

The researchers came to the unavoidable conclusion that what they were seeing was empathy — and apparently selfless behavior driven by that mental state.

“There is nothing in it for them except for whatever feeling they get from helping another individual,” said Peggy Mason, the neurobiologist who conducted the experiment along with graduate student Inbal Ben-Ami Bartal and fellow researcher Jean Decety.

The presence of a rat trapped in a restrainer elicits focused activity from his cagemate, leading eventually to door-opening and consequent liberation of the trapped rat. (Science/AAAS)

“There is a common misconception that sharing and helping is a cultural occurrence. But this is not a cultural event. It is part of our biological inheritance,” she added.

The idea that animals have emotional lives and are capable of detecting emotions in others has been gaining ground for decades. Empathic behavior has been observed in apes and monkeys, and described by many pet owners (especially dog owners). Recently, scientists demonstrated “emotional contagion” in mice, a situation in which one animal’s stress worsens another’s.

But empathy that leads to helping activity — what psychologists term “pro-social behavior” — hasn’t been formally shown in non-primates until now.

If this experiment reported Thursday holds up under scrutiny, it will give neuroscientists a method to study empathy and altruism in a rigorous way.

Do age and gender affect empathic behavior? Will a rat free a rat it doesn’t know? Is more help offered to individuals an animal is related to, either directly or as a member of the same genetic tribe? What are the genes, and their variants, that determine whether one animal helps another and how much? Answering those questions becomes possible now that there is an animal “model” for this behavior.

“The study is truly groundbreaking,” said Frans de Waal, a primatologist at Emory University who has written extensively about empathy. What is particularly interesting, he said, is there appears to be no clear cost benefit trade-off going on.

“We are entering a distinctly psychological realm of emotions and reactions to the emotions of others, which is where most human altruism finds its motivation.”

Jeffrey S. Mogil, the McGill University neuroscientist who showed emotional contagion in mice in 2006, said that “what is amazing about this is that it shows empathy in such a robust way. This is not something that rats would otherwise be doing.”

A major question that needs to be answered next is whether the free rat liberates the captive one to relieve its own stress or the stress of the other animal.

“It’s more likely to be the former,” Mogil said. “But even if it is the former, I’m not sure that’s so different from humans.”

In the new experiment, the pairs of rats were put in the experimental condition for an hour a day for 12 days. (They had previously spent two weeks together in a cage and knew each other.) The rat opened the door to the trapped rat’s cage by chance the first time, usually freezing in fright when it fell over noisily. In an average of seven days, however, it had learned to open the door intentionally and was no longer spooked when the door fell over.

In 13 percent of the sessions, the trapped animal gave an alarm call, but vocalized distress was clearly not necessary to put the free rat to work. When the cage was empty or occupied by a rat doll, the free rat sometimes opened it, but over the course of days lost interest in doing so.

After liberation, the rats nuzzled and explored the experimental arena. But when the setup was changed so that the captive exited into a different area, the free rat still opened the door for the captive one.

When a cage with five chocolate chips was added to the arena, the free rat opened it, too. That animal consumed all the treats if the other cage was empty. But if it contained a captive rat, the free rat shared the chocolate about half the time, letting its compatriot have 11 / 2 pieces on average.

“To actually share food — this is a big deal to a rat,” Mason said. “I didn’t think they would do that.”

Mason sees two processes at work. The first is one animal’s ability to identify and share another animal’s stress. But equally important is the ability to control the “acquired” stress and keep from becoming overwhelmed. But that was something not every rat could do. All six female rats in the experiment learned to open the captive’s cage, but seven of 26 males never did.

“I don’t think it’s because they didn’t have empathy. I don’t think they had the ability to down-regulate their own stress and act on the empathy,” she said.

Mason thinks that empathy and altruism evolved with females caring for helpless offspring. Natural selection favored those maternal traits, which then became generalized to both sexes. They helped forge social bonds that aided the survival of individuals and groups. She suspects the behavior is “sub-cortical” — closer to a reflex than a thought, and driven by ancient parts of the brain. De Waal, who in 2009 wrote a book called “The Age of Empathy” whose cover featured a chimpanzee shaking hands with a man, agrees up to a point.

“It is an intelligent response, but the motivation is, as in humans, an empathic process that is fairly automatic,” he said.

Jaak Panksepp, a neuroscientist at Washington State University who wrote an accompanying commentary in the journal, said that many people still doubt that animals have emotional lives that can be studied.

“Some skeptics are bound to say that this interpretation is a bit far-fetched,” he said in an interview. “What this provides is reasonably good evidence for empathy, and a model system to study the underlying processes further.”