By now you probably know about Walter Mischel's famous 1972 marshmallow task, which used the tasty snacks to measure impulse control in children. What you might not know is why marshmallows can help us better understand what it means to be human. A new study, in which crows and ravens were subjected to a version of the task, might explain why.
The original experiment was fairly straightforward. A child is brought into a room where an experimenter produces a marshmallow. The child is told she can eat it now, or they can wait a few minutes, and eat two. But if the kid doesn't wait, they won't get the second marshmallow. Then the child is left alone with the marshmallow. If she waits for the larger reward, then she's got decent impulse control, at least according to the theory. It's called "delayed gratification."
The researchers followed up on those kids, and they found that the ones who had the willpower to avoid gobbling up the marshmallow right away had, on average, higher SAT scores than the ones who couldn't wait. They were also less likely to use drugs. And so on. As Slate reported last year, it probably isn't actually that straightforward, but while the conclusions are subject to reinterpretation, the findings themselves are robust. Some kids are more likely to delay their gratification than others, depending on the circumstances.
Now, researchers have given a version of the marshmallow task to a brainy group of birds called corvids - such as ravens and crows - and their findings are now online in the journal Animal Behaviour. They tested seven crows and five ravens, all of whom were raised in zoos or were injured, rescued, and rehabilitated, but for some reason could not be released.
Researcher Friederike Hillemann and a female crow named Resa.
The researchers began by figuring out each bird's favorite food: beans, bread, cheese, corn, grapes, sausage, or fried pork fat. (To nobody's surprise, fried pork fat was usually the most prized item.) They used that to pit the quantity of a food reward against the quality. For example, if one crow's favorite food was sausage but he was given bread, his least favorite, it might make sense to wait for a second snack. If given a bean, a low-value food, is it really worth trading it up for more beans? Maybe, maybe not. It might be worth waiting to get more sausages, but could the birds resist the impulse to snack when faced with their preferred treat? By running these sorts of tests, the researchers could determine in what circumstances the birds would delay their gratification, and when their impulses were too strong to ignore.
Each bird was given one bit of food and after a delay that lasted between two seconds and ten minutes, they could trade up: either for a snack they liked better or for more of the original snack. In a second type of test, four food items were visible to the birds but out of reach. The experimenter provided the birds with each bit of food in order, with a fixed delay between each item. If the birds were willing to wait, they could wind up with all four bits of food, but if they ate anything before it was all dispensed, they'd miss out on what was left over.
In this video, a crow named Resa participates in two trials in the 40s delay condition. First, she eats the initial item, so the trial is ended. In the second trial she waits the 40 seconds to exchange bread for corn. Watching her struggle with the impulse to just gobble up the bread is absolutely charming.
Impressively, the birds were willing to avoid eating the original snack, but only if they could trade for something they liked better. If given sausage, they wouldn't trade for beans, but if given beans, they'd happily resist temptation to wait out the sausage. But that didn't hold for the quantity trade. They were perfectly happy eating one bit of food and not waiting for more of the same, no matter if the food they were given was low-value or highly preferred. The same was true for the second task: if they saw a high-value treat coming, they could wait. But once they had the sausage in their grubby little beaks, they didn't bother waiting for more.
Together, the experiments show that crows and ravens can delay their gratification, up to an impressive ten minutes, but only when it leads to a reward of higher quality. That reflects both their incredible sophistication, but also their cognitive limitations. They know to wait for a more tasty reward, but once one piece of sausage is available to them, they find it incredibly hard to wait very long for more. "The birds exchanged more often when the potential reward was highly preferred, the initial item was of low quality, and when the relative value of the reward was clearly distinct from the initial item." That means that they weigh the loss of the initial food item against the gain of the future reward.
Lead researcher Friederike Hillemann said she hopes to figure out how to test wild corvids, but they are extremely shy and don't usually approach experimental apparatuses. Still, she's got her eyes on the rook colonies and cooperatively breeding crows of Northern Spain.
Hillemann and a female crow, Resa, inspect eachother.
Animal behavior researchers are interested in delayed gratification since the cognitive machinery that underlies the ability to reason about various rewards and to incorporate the passage of time into that reasoning is important for foraging decisions. Avian researcher Nathan Emery pointed out that many wild corvids cache their food, and the time between hiding and eating can be up to several months. "They are also frequently hungry when caching, so it's not as if the caching bird is hiding food because it is satiated," he says. In addition, he points out that delayed gratification has important social ramifications as well. Some corvids engage in what's called "courtship feeding," where a male feeds a female, even when he hasn't yet eaten himself. The same is true for feeding of offspring, where either a male or female finds food for their chicks. At minimum, this behavior requires some self-control to ensure that the food remains uneaten until fed to the hatchlings.
Hillemann points out that postponing an action in favor for a future, better option may pay off in more abstract social situations as well, and not just for corvids. When it comes to mating, you might prefer spending your time and energy seeking out an ideal partner, rather than focusing on the first available. "[And] while arguing with your supervisor, you should hold back in the heat of the moment in order to keep your job," she says.
Indeed, those mechanisms probably underlie the social interactions that characterize the complex societies of elephants, cetaceans, at least some birds, such as corvids and parrots, and primates, including humans. Reciprocal altruism, for example, requires the mental accounting of favors given and received over time, and for dominance hierarchies to remain stable, subordinate individuals must be able to inhibit their desires in the presence of dominant ones.
How do the corvids stack up against us? "Crows and ravens performed comparably to primates and children tested in [similar] tasks," Hillemann said. But one difference between the avian brain and the primate brain may be the way that they consider quality versus quantity. Birds usually aren't willing to wait in order to increase the amount of food, only to optimize the quality of their reward. "This may be unique to birds," she adds, "because carrying a large amount of food can be disadvantageous in flight." On the other hand, a human child can fit two marshmallows in his mouth just as easily as one. While more research is needed, this lends support to the notion that this sort of impulse control is a case of convergent evolution, having evolved independently in the avian lineage and in our primate lineage.
In a way, the marshmallow task might tap into one of the most fundamental tradeoffs that defines human society: waiting your turn versus taking what's yours.
Hillemann F., Bugnyar T., Kotrschal K. & Wascher C.A.F. (2014). Waiting for better, not for more: corvids respond to quality in two delay maintenance tasks, Animal Behaviour, 90 1-10. Link to paywalled article.
Crow photos copyright Claudia Wascher, used with permission. Video via Friederike Hillemann, used with permission.