Ravens Can Plan for the Future

They join an elite group of animals that includes great apes, but not monkeys or 3-year-old human children.

Nevemore. (Mario Anzuoni / Reuters)

When you’re a scientist who studies birds as inquisitive and intelligent as ravens, doing experiments is easy. “If they find the task interesting enough, they all line up, and almost fight over who gets to take part in the study,” says Mathias Osvath from Lund University, who keeps 16 of the birds on his farm. As we talk on the phone, I can hear them cawing in the background.

For years, Osvath has been trying to work out if animals have foresight—if species other than humans can plan for the future. And his latest experiments suggest that ravens can. Based on their previous experiences, the birds will select tools that can help them solve a puzzle in the future, or pick up tokens that they can later use to barter for food with human experimenters. And that, Osvath says, puts them “on par with apes.”

As was the case with many mental abilities, scientists used to think that only humans could take themselves out of the present and anticipate their future needs. They assumed that other animals are entirely motivated by their current state, thinking only about whether they’re hungry right now. If animals looked like they were preparing for tomorrow, as a squirrel does when it buries nuts, that’s just because they’re carrying out rote, instinctive behaviors.

Around a decade ago, notions of human exceptionalism started to crumble. Osvath and others found evidence that other great apes, like chimpanzees, bonobos, and orangutans, could choose and save tools for later use. One particular chimp named Santino, who lived in a Swedish zoo, would stockpile arsenals of stones that he would later throw at visitors.

It might seem intuitive that our closest evolutionary relatives should share our capacity for premeditation, but in 2007, Nicola Clayton from the University of Cambridge discovered the same ability in the corvids—a group of brainy birds that includes crows, ravens, jackdaws, jays, and magpies. In a clever experiment, she housed jays in cages with various rooms, one that always contained food in the morning and another that never did. When given access to nuts, the jays stored three times as many in the room that would otherwise remain empty, suggesting that they were anticipating their future needs.

Not everyone bought that interpretation. Skeptics argued that the jays’ success in Clayton’s experiment may just reflect a highly specific aptitude for burying food—something that these birds spend an inordinate amount of time doing. Even if they were planning for the future, they were only doing it in a narrow way, displaying what some call “laser-beam intelligence.” We humans, by contrast, can plan for the future in all kinds of ways, from stocking freezers to buying insurance to saving money—and there was no evidence that corvids share the same flexibility.

That’s where Osvath’s ravens come in. His student Can Kabadayi trained the birds to open a baited puzzle box by putting a stone of particular weight and shape into a tube. He then moved the box out of sight. After an hour, he offered the birds a tray containing various objects, including the magic box-opening stone. Fifteen minutes later, after the birds had chosen an item from the tray, Kabadayi brought back the puzzle box. The stone has no value on its own, and the birds had no idea that the box would return. And yet, they chose the stone from the tray around 86 percent of the time.

In a similar test, Kabadayi trained the birds to exchange a blue bottle cap for a bit of food. An hour after he left, the birds were moved to a different location, where another experimenter offered them a tray containing the token and some other bric-a-brac. Again, the cap had no intrinsic value and the birds didn’t know of Kabadayi would return with food. Nonetheless, when he did after 15 minutes, the ravens almost always had a blue bottle cap waiting for him.

These experiments are essentially the same as the ones that Osvath and others performed with great apes, and the ravens passed in a similar way. Based on their past experience, they picked the object that would be most useful to them in the future. They did so even when Kabadayi extended the waiting period from 15 minutes to 17 hours. And they did so even when the trays contained a piece of food, showing that they can forgo an immediate reward in anticipation of a better one later.

This is all the more impressive because ravens are not frequent tool-users, and they’ve never been known to barter for food in the wild. Unlike Clayton’s jay studies, Osvath’s experiments forced the birds to do things that they don’t naturally do. Their success suggests that they can plan for the future in the same flexible way that humans and other great apes can.

Again, not all researchers are convinced. If the ravens don’t actually know when they’ll encounter the puzzle box, wonders Jennifer Vonk from Oakland University, who also studies corvids, how do they know which item to select? Since they were trained to use the stone to open the box, and since the stone was the only object that ever did so, “it makes sense that the birds would develop a preference for that tool,” Vonk adds. “It isn’t clear that this preferential selection reflects future planning.”

Osvath argues that the birds were only given minimal training, in a different location than where they were tested. “These animals had only used a tool five times in their entire life, and they aren’t predisposed to tool use,” he says. And when tested, all the birds picked the right object on their very first go, suggesting that they didn’t just learn the correct answer over the course of the experiment.

Even if you don’t buy any of that, he says, you would still need to explain why monkeys consistently fail at these kinds of tasks, or why human children only succeed after they turn 4. “They’re really good at learning and have predispositions for tool use,” says Osvath, and yet they don’t pick the right tool or token when tested in the same way as the ravens.

“This study contributes to lots of mounting evidence for the amazing [parallel evolution] between apes and corvids,” says Laurie Santos from Yale University. They are separated by 320 million years of evolution, but “in terms of their tool-use, elaborate social cognition, and now even forward planning, these two groups are surprisingly similar, which raises some cool questions about why the two groups wound up so similar.”

To plan for the future in a flexible way, animals must be able to learn, represent a future event in their minds, understand that their current actions can contribute to an abstract and unobservable goal, and restrain the pull of their present senses to focus on that goal. Do corvids have all of these skills? Santos isn’t sure. “The current results show that birds value objects that they know might be valuable later. It’s nice evidence for forward planning, but I don’t think it shows that the birds are thinking of themselves in the future in any rich way,” she says.

To the extent that corvids and apes share similar skills, did they evolve their abilities independently? Or did they build upon abilities that were present in their common ancestors? “Those are questions I’ll work on for the rest of my life,” says Osvath.

It should be a fun ride, especially given how surprising ravens can be. For example, Osvath and Kabadayi had to exclude one of the females from their study because she invented a new way of opening the puzzle box. The duo thought that they had eliminated all such possibilities by combing the raven’s cage for any object that was similar to the special stone.

But alas they were mistaken. There was one more way to break in:

Fill the tube with bits of bark that had been scattered on the floor.

So one raven, thoughts unspoken, just ignored the offered tokens,

Used the bark instead to open up the puzzle box’s door

Thus her part within that study, which the duo had in store,

Would continue nevermore.

Related Video