Monday, January 08, 2018

A marshmallow test for ravens

You’ll certainly have heard of it, for the research is nearly fifty years old: The marshmallow test, done by Walter Mischel and his colleagues of Stanford University. You can find it on the Internet and here I’ll simply quote one of the descriptions there, in this case one by Stewart Brand: “A researcher whom the child knew and trusted, after playing some fun games together, suggested playing a ‘waiting game.’ The researcher explained that the child could have either one or two of the highly attractive treats the child had chosen and was facing (marshmallows, cookies, pretzels) – depending on how long the child waited for them after the researcher left the room. The game was: at any time the child could ring a bell, and the researcher would come back immediately and the child could have one treat. To practice, the researcher left the room, the child rang the bell and the researcher came right back, saying, ‘You see, you brought me back. Now if you wait for me to come back by myself without ringing the bell or starting to eat a treat you can have both of them!!’ The wait might be as long as 15 or 20 minutes. [The kids varied widely in how long they could stand it before ringing the bell, and about one third waited till the researcher came back by himself.]” (
After the experiment Mischel followed the children for many years and it became clear that it said much about what kinds of persons the children would become later in life. However, that’s not what I want to talk about. Here it’s relevant that we can delay gratification in order to be better off in future. A part of us can do it already at a young age and most people can do it better when they grow older.
I think that you consider this ability to think ahead and to control yourself typically human. Look around: Isn’t it so that animals always immediately take what they can get? Okay, maybe there are some apes and monkeys who can refuse to take now what they like, expecting that later they’ll get something what they like more. And maybe there are other mammals that can do it as well. But birds?
Can Kabadayi and Mathias Osvath of Lund University decided to test the ability of flexible planning in birds and took five captive ravens. They had to do tasks they do not do in the wild. Let me quote how The Guardian describes it:
“The birds were shown a box that had a tube sticking out of the top, plus three stones. They learned that they could use a stone as a tool. If they dropped it down the tube, the box would release a doggie treat. They also learned that other familiar objects, such as a small wooden wheel or a ball, would not work. In one experiment, the ravens were shown the box without any stones available. Then the box was taken away. An hour later, in another location, they were presented with a tray containing a stone plus three objects the birds knew would be useless. They were allowed to choose one thing from the tray. Fifteen minutes later, the box would show up again. In 14 cases of encountering the tray and later seeing the box reappear, the birds usually chose the stone and proceeded to use it correctly. The same thing happened in another experiment, when the box did not show up again until the next day, a delay of 17 hours. Further work showed the ravens would pass up an immediate reward if they could get a better one by waiting.” ( To be exactly, in 80-90 % of the cases the ravens selected the correct tool! The result is the more impressive since “Monkeys have not been able to solve tasks like this,” so Osvath (on the same website) Moreover, the ravens were also better than toddlers in such experiments.
As it happens, experiments like these are always difficult to interpret and alternative interpretations are always possible. It’s likely that the test shows that ravens (and possibly other birds as well) have a planning capacity that is more than stashing food away for later (like squirrels do, for example). However, as Alex Taylor, an animal cognition expert of the University of Auckland, says to National Geographic: “The ravens may not be thinking about the future at all, they may instead just be choosing the object that has been associated the most with food.” What’s true must yet have to be found out. Nevertheless, the result is remarkable. Until now scientists thought that flexible planning for unexpected future events was limited to humans and great apes. In the test, the ravens – so birds – were as good in such pre-planning tasks for novel behaviour. If so, this pre-planning ability must have been evolved more than once.
It even seems that ravens are more patient than humans, since they go somewhat less for immediate rewards than humans! Indeed, it might have happened that ravens would not have shown the same behaviour, if they had been given marshmallows. Simply, because they don’t like them so much as children do, they might postpone picking at them. Be it as it may, we think that nothing is as unique as how humans think, but apparently we are not as unique as we think. There are white ravens in nature but they are not human.

- see text

No comments: