Pointing at something is one of those human universals that we do without thinking. We start showing infants objects by pointing very early, and when they get it—when they look at what we’re indicating or point at something themselves—it’s a significant developmental benchmark. That’s because pointing indicates some rather complex cognition—an understanding not just that you have thoughts in your own head but that other people have different thoughts and that they can share them. From that so-called theory of mind comes learning, sharing, empathy, even humility, all the stuff of the human collective. And the simple business of pointing a finger is its first indicator.
Over the last few decades, that has led cognition researchers to an inevitable question: Can other creatures understand human pointing, and if so, what does it imply about their cognitive abilities? The findings have been mixed, and suggest that a gesture that seems so self-evident to us can be exceedingly hard to understand. Dogs are naturals at understanding pointing—no surprise, perhaps, given what we like to think about their emotional complexity. Other animals come to an understanding of it much less naturally and more laboriously—if they get it at all. What’s fascinating is that chimps, our closest animal kin, seem to have a lot of difficulty with pointing, implying that perhaps humans’ use of the gesture arose after our species parted ways.
Now, researchers publishing in Current Biology have found that a group of African elephants living in captivity in Zimbabwe seem to be able to understand pointing naturally, showing an impressive ability to choose a bucket with food in it as opposed to an empty one when a human is pointing toward the chow. Moreover, the animals did not have to be trained by the researchers to perform the task. They picked the right bucket with the same frequency regardless of how many chances they’d had to practice the test.
The pointer in this study was Anna Smet, a psychology graduate student at University of St. Andrews. She set up the experiments with 11 elephants belonging to a tourism company called Wild Horizons, which provides elephant rides in the area of Victoria Falls. First, while a subject elephant was held some way off by a handler, Smelt put up a screen and pantomimed putting food into two buckets, but actually filled only one. Then she removed the screen and pointed at the correct bucket in one of a series of postures—with her full arm, say, or with her arm crossed over her body. On every trial, she looked back and forth between the elephant and the bucket, in the same way that people do when trying to catch each others’ attention.
Crunching the numbers on the elephants’ responses, Smelt and her adviser, St. Andrews psychology professor Richard Byrne, found that the animals chose the bucket containing the food a greater proportion of the time than by chance alone when she took any of the pointing poses but one (a contortionist’s gesture involving pointing at something with the opposite arm and poking out the elbow distractingly. Human babies don’t recognize that as pointing either.) It didn’t matter if Smelt was nearer the food bucket or farther away, suggesting that the elephants were not just choosing the bucket physically closest to her. Looking at the bucket or staring at the elephant until it chose did not result in it choosing the food bucket more either. Pointing had to be a part of things.
“What really amazed us about the elephants was that we didn’t find any sign that they were learning it,” says Byrne. “I think it implies that they knew how to do it in the first place.”
It’s possible of course that Byrne is wrong about that. The elephants might have learned to interpret pointing from their time around humans, but if that were true, the ones that had been in captivity longest would do better than the comparative newcomers. Yet neither that factor nor the elephants’ age seemed to make a difference. The trainers at Wild Horizons specifically use verbal cues, rather than gestures, in their communication with the elephants, and Smet did not see any pointing in the months she spent observing them. And even if the animals did learn pointing from humans, that still puts them a step ahead of every other species the researchers have tested, which never make the connection between the gesture and the intended target.
In some ways, it’s a wonder that scientists didn’t look for pointing awareness among African elephants earlier, since the animals have some of the most complex social groups outside of our own, and they are extremely intelligent. Various studies have observed elephants showing signs of what appears to be grief over dead companions and of cooperating with each other. Their social abilities clearly evolved separately from humans’, which means that studying them can help shed light on what situations result in the evolution of complex societies.
Brian Hare, a professor of psychology at Duke University who performs pointing studies with dogs, says he welcomes the new study’s insights and looks forward to further experiments that can help make clearer just what elephants understand human pointing to mean. “The question is are elephants slow[ly] learning some simple cues or are they thinking about [what others are] thinking and flexibly adjusting their behavior to others based on that understanding? One is a very simple cognitive mechanisms seen in a range of species; the other is very complex and only seen in humans,” he wrote in an email.
Do elephants point among themselves? Byrne says that so far there doesn’t appear to be any record of it, but he wants to study the question further. “The special thing about elephants is just how very social and empathic they are. Something about their social lives and their social understanding is perhaps what has selected for this ability,” he says. “I think it also implies that they probably ought to show some pointing ability themselves. As you can imagine, Anna and I are quite keen to find out.”
(MORE: Your Cat is 95% Tiger)