Scientific Method —
Ravens ignore a treat in favor of a useful tool for the future
Planning ahead means they're even smarter than we thought.
Cathleen O'Grady - 7/14/2017, 11:21 AM
If humans, chimps, gorillas, and orangutans can all do something, but monkeys can’t, that tells a certain evolutionary story: it suggests that the ability emerged sometime after the apes split off from the monkeys on our evolutionary tree. But if a bird comes along with that ability, it throws the whole story off course. Corvids—a family of birds that includes ravens, jays, and crows—seem to delight in doing just that.
Humans pretty obviously plan for the future, from packing a brown bag lunch to saving for retirement. Other apes also seem to be able to plan for the short-term future, at least up to one night. Monkeys don’t. But a paper in Science this week reports a small group of corvids succeeding at future-planning tasks. That points to a complex evolutionary story.
Two cognitive scientists at Lund University in Sweden, Can Kabadayi and Mathias Osvath, conducted a series of experiments with five captive, hand-raised ravens. Obviously, that’s not a lot of ravens, and hand-raised ravens do not behave like wild ravens. But when it comes to figuring out the outer bounds of cognitive abilities for a species, those aren’t the most important problems to worry about. Testing more ravens, and wild ravens, comes later.
Jelly bean now, or burger tomorrow?
First, the ravens had to be given an experience they could plan for. They learned how to operate a puzzle box that opened to yield a reward, but the box could only be opened with a specific tool. The next day, they were shown the puzzle, loaded with food, but no tool. Only an hour later they were shown a tray of objects, including the tool, and given the opportunity to choose just one thing. Fifteen minutes later, the puzzle came back, and if the raven had chosen the right tool, they could open it.
Repeatedly, through 14 trials, the birds chose and used the correct tool 79 percent of the time—much higher than chance. The average would have been higher if one of the ravens hadn’t thwarted the researchers by figuring out a way to open the puzzle without using a tool—she didn’t pick the tool, because she didn’t need to. Without her, the average was 86 percent.
The ravens could also solve a similar future-planning problem that involved bartering. They learned that a researcher would give food treats in exchange for a particular token. Then, the researchers would start asking the raven for the token, even when it was nowhere to be found. Once the ravens were shown a tray of objects, including the token, they chose and used the correct token at a rate higher than chance—78 percent of the time.
Perhaps most importantly, four out of five ravens got these tasks right on the first trial, before they had any chance of learning the particular task by experience or building habits. They also aced the tasks when the waiting period was extended to overnight.
These ravens weren’t done yet: they passed a task that required them to choose the tool for opening the puzzle—which contained a superior treat—over an immediate but inferior treat. Doing that requires not just planning, it also requires self-control.
These studies were carefully set up to show that the ravens could plan under flexible conditions—different time delays, and solving either a mechanical (puzzle box) or social (bartering) problem. They didn’t just match apes in their performance; they beat them. The ravens even performed better than 4-year-old children.
Ape brains, bird brains
The simplest assumption in evolution is that all species that share a trait share an evolutionary history, and that the trait emerged at the beginning of that evolutionary history. Take the spine: all vertebrate creatures share an evolutionary history right back to the point at which the first nerve cord appeared. We share the spine because we share the history.
But sometimes it’s more complicated. Birds, butterflies, and bats all have wings—but not because all their ancestors right back to their last common ancestor had wings. Those three groups all evolved wings independently in response to similar evolutionary pressures.
The same ideas apply to cognitive abilities. Can corvids and apes plan for the future and think about other minds because those abilities are shared way back on the family tree, where mammals shared a last common ancestor with birds? That was around 320 million years ago, suggesting that everything from snakes to rabbits should have the same abilities. It seems incredibly unlikely. So instead, the best explanation is convergent evolution: we developed the same cognitive traits in response to similar evolutionary pressures.
That means, write Kabadayi and Osvath, that these “avian dinosaurs” don’t just show what bird brains are capable of—they “open up avenues for investigation into the evolutionary principles of cognition.” Looking for the evolutionary pressures that lead to this kind of flexible cognition can help us to understand why humans, our close relatives, and a handful of other species ended up the way we are.
Science, 2016. DOI: 10.1126/science.aam8138 (About DOIs).
