Grey parrots may sometimes be impulsive—think about how often you may have had to give your bird multiple timeouts for the same behavior (like chewing on your sunglasses) in a very short time period. However, my students and I have shown that our parrot, Griffin, can actually exhibit quite a bit of self-control. We found that out by giving him a classic task used to test children.
The task is informally called the “marshmallow test,” and was designed by the psychologist Walter Mischel in the 1970s. He examined about 60 children, all approximately 4 years old, in a laboratory setting. He used several versions of the task, but in the most common version, he sat each child down behind a table, on which he placed a plate with one marshmallow.
He told the child that he was going to run an errand, and that he would be back in about 15 minutes. He told the child, “If you can keep from eating the marshmallow on the plate in my absence, I’ll give you a second marshmallow when I return.” He also told the children that they could eat the first marshmallow at any time, but that if they didn’t wait for his return, they could not have the second one. And then he left the room (Mischel, 1974).
Now, if you know anything about 4-year-olds, for them 15 minutes is a lifetime. Many children gave up and ate the first marshmallow. But a number of children were able to wait, and almost all of them managed the task by figuring out how to distract themselves. They got up from the table and danced around; they sat at the table and sang to themselves. Sometimes they licked the marshmallow but didn’t eat it. They played with their hair or their clothes. Some even tried to nap.
Mischel argued that those who waited had more self-control, a form of executive function. He surmised that the children with more self-control would be more successful in certain situations, such as being able to finish their homework before going out to play. It wasn’t until he went back and interviewed these children some 30 years later that he found out that his hypothesis was correct: The children who had waited the longest had, for the most part, done better in school, had gone to higher-ranked colleges, had better jobs and fewer divorces! [One note: All of the children he tested had come from middle-class families.
When experimenters subsequently tested children from impoverished communities, these children often didn’t wait, and gave reasons such as not trusting the experimenter to return, or said that they feared that someone else would eat the second marshmallow or even that someone would come and steal the first one…a sad commentary, for sure.]
Nonhumans Put To The Test
My students and I had read a few papers in which researchers used this delayed gratification task on nonhuman subjects. Many papers involved experiments on nonhuman primates, but none of those studies had actually tested the animals in exactly the same way as Mischel had tested the children. In fact, we felt that some of the tests on the nonhuman primates probably had some design flaws. However, on tests very similar to those used by Mischel, cockatoos and some corvids seemed to be able to wait for a better reward, although not for more of a reward (see Auersperg et al., 2014; Hilleman et al., 2014).
The cockatoos waited for many seconds; the corvids for up to about 10 minutes. Interestingly, the corvids, who often cache food in nature, sometimes cached the first reward while waiting—possibly a case of “out of sight, out of mind?” These studies made us wonder whether an African grey parrot might do as well as the children, or maybe even better than the cockatoos and corvids. We knew that our grey parrot, Griffin, understood the word “wait”—although he had no choice in the matter, he heard the word every day when we told him to wait for his cooked luncheon grains to cool, and to wait while people entering into the lab used hand sanitizer and took off their outdoor shoes before going over to greet him. So we proceeded to test him, using Mischel’s exact task (Koepke et al., 2015).
We noted that in almost all the other tests with nonhumans, researchers tweaked Mischel’s procedure. For example, they slowly increased the waiting time—so if their subjects could refrain from eating a peanut for 10 seconds, on the next trial they would try making the subject wait 20 seconds, and so forth, until the subjects failed. Such a procedure could actually be training the subjects to wait, rather than testing their basic behavior.
To control for this possibility, we mixed up all the waiting times. Other researchers sometimes used only one pair of treats; we wanted to ensure that one treat wasn’t some kind of signal for another treat, so we used several different pairs of treats. And we also wanted to ensure that Griffin didn’t interpret the word “wait” as a trained command (like that given a dog who has a biscuit placed on his nose and has to wait for a command to flip it and eat it). So on some trials, we presented the more-favored reward and asked Griffin to wait for the less-favored one.
African Grey Griffin Plays The Waiting Game
Griffin succeeded on the task—he waited in 108 out of 120 trials. He was as successful on the 15-minute delays as on the 10-second delays. He didn’t learn to wait…he made as many errors at the end of testing as at the beginning, and there were as many long delays at the beginning as at the end of the experiment. The different types of treat pairing mostly didn’t matter, although on two of his failures, the rewards were very close in desirability (a cashew and a candy), and he decided to eat the slightly less-favored reward almost immediately.
On the trials in which we asked him to wait for something less favorable, he barely waited at all—he spent a second or two looking up at us as though we were demented and then immediately ate the reward, showing that he was very aware of when he should and shouldn’t wait. And he wasn’t simply losing interest in the reward—on a few of his failures, he waited until almost the end of the delay period—in one case, over 14 minutes—before succumbing to the lure of the reward that was present.
What was particularly interesting, however, were the behavior patterns he exhibited while waiting: He did almost the same things as the young children! He talked to himself, he tried to nap, he preened, he turned his head away. Like the children, he sometimes licked the reward, but didn’t eat it. We actually made a split-screen video of the children (from YouTube) and Griffin to use for a presentation at a conference, although we couldn’t use it in our published paper.
Clearly, Grey parrots can sometimes act quite rashly. However, when it is worthwhile, they are definitely willing to wait, particularly for a better reward. In the wild that makes a lot of sense: If foraging birds come across some fruit, but know that more calorically dense nuts are a bit further along, it’s far better for them to wait for the nuts than to fill up on the fruit. The next step, of course, is to see if Griffin will wait for more of a reward…and that might not be so easy, as it makes less ecological sense: There is no reason not to stop at a small patch of nuts and eat them while en route to a larger patch. Keep tuned to see what Griffin will do!
Auersperg, A. M. I., Laumer, I. B., & Bugnyar, T. (2013). Goffin cockatoos wait for qualitative and quantitative gains but prefer ‘better’ to ‘more’. The Royal Society: Biology Letters, 9, Article 20121092.
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.
Mischel, W. (1974). Processes in delay of gratification. In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 7 (pp. 249–292). New York, NY: Academic Press.