Ravens Are More Cunning Than Human Preschoolers | @GrrlScientist
A recent study finds that ravens are capable of planning and bartering, and they also forgo an immediate reward to obtain a better reward in the future
This piece was a Forbes Editor’s Pick.
Anyone who knows birds knows that many corvids cache food. Corvidae is a songbird family that comprises 120 species of ravens, crows, jackdaws, magpies and jays; and this caching behavior presumes that these birds are planning for the future. But are they really?
In a recent paper, cognitive zoologist, Mathias Osvath, at Lund University in Sweden, and his doctoral student, Can Kabadayi, wanted to better characterize the ability of common ravens, Corvus corax, to plan ahead. To do this, the team replicated a series of key experiments originally designed to test great apes’ planning abilities (ref). These experiments relied upon either tool use or bartering to obtain a food reward. An advantage of replicating these primate experiments in ravens was the direct comparisons that could be made between these two unrelated — and very different — groups of animals.
In a series of four experiments, Mr. Kabadayi and Dr. Osvath found that ravens will use tools to obtain food rewards and moreover, the birds saved tokens that they later presented to a human experimenter in exchange for food, demonstrating that ravens are more intelligent than we give them credit for. Additionally, these experiments indicate that birds are just as capable of complex thought as mammals are, and further, these findings add to a growing body of data demonstrating that intelligence evolved independently more than once.
Do ravens make decisions in anticipation of a future event?
To determine how long into the future that a raven can plan, Mr. Kabadayi designed a puzzle box containing a favored food reward. He trained the birds to drop a rock into a tube on this puzzle box so the treat (a piece of dog kibble) would fall out.
The following day, the ravens were first presented with the puzzle box, but not with the stone opening tool. Then the box was removed and one hour later the researchers presented each raven with a tray containing the stone opening tool along with three “distractors”. The ravens had previously discovered that these distractors were useless for obtaining the treat, either being too light or too big to do the job. After each bird had selected an item, the puzzle box was then returned either 15 minutes later (experiment 1) or 17 hours later (experiment 2).
This video captures one of the trials:
I found it intriguing that, despite not knowing whether they’d ever see the puzzle box again, the ravens chose the functional opening tool almost 86% of the time. Even after the time delay was lengthened from 15 minutes to 17 hours — until the next day, basically — the ravens still chose the correct opening tool 89% of the time.
These findings are interesting because the ravens proved they were just as adept as great apes in these tool-using tasks, despite lacking the predisposition for tool use. Additionally, great apes were the only animals that could pass these tests, until Mr. Kabadayi and Dr. Osvath conducted these studies with ravens.
“Monkeys have not been able to solve tasks like this,” Dr. Osvath pointed out, and, he added, even preschoolers can’t pass this test until sometime after their fourth birthday.
Will ravens barter with humans for food?
In the third experiment, Mr. Kabadayi and Dr. Osvath trained their ravens to exchange a specific token (a blue plastic bottle cap) for an immediate food reward (a piece of dog kibble). In addition to tool use, bartering (especially with humans) is another behavior that wild ravens don’t perform naturally.
To impress upon the ravens the value of the tokens, and to motivate them to save any tokens that they might find, the experimenter asked the birds for the token when they did not have it.
One hour later, after the ravens had been moved to a different enclosure, the birds were permitted to choose one item from each of three trays presented to them in rapid succession by an experimenter whom they’d never bartered with before. Each tray included several distractors along with the bartering token.
All of the ravens did extremely well in these tests, selecting the bartering token nearly 74% of the time — demonstrating they were more adept at this task than were great apes.
“In the bartering conditions, the ravens outperformed orangutans, bonobos, and particularly chimpanzees,” Mr. Kabadayi and Dr. Osvath noted in their paper (ref).
Do ravens have self-control?
The fourth experiment built upon studies previously conducted in primates by testing whether the ravens placed a higher value on a food reward that was immediately available (but of poorer quality), or were the birds instead willing to wait for their favorite treat. In this experiment, the ravens were allowed to choose only one item from a tray bearing either the apparatus-opening tool or the bartering token along with distractor items, and an immediate food reward.
In both conditions (tool use and bartering), Mr. Kabadayi and Dr. Osvath found that their ravens ignored the immediate reward in favor of choosing the functional item in 100% of the trials — even when their favorite treat was delayed until the next day.
In these tests, the ravens’ performances were comparable with the self-control seen in great apes.
Thinking outside of the (puzzle) box
One raven, whose name was None, was especially creative. Basically, she hacked the puzzle box test by stuffing it full of tree bark — a nifty solution that produced the desired result, without the bother of having to wait for the pesky opening tool to appear.
“In trial nine, one female invented a way to open the apparatus without the tool (and was therefore excluded from subsequent tool conditions in the rest of the study),” the authors report dryly regarding None’s Eureka Moment (ref).
Is a complex and dynamic social structure prerequisite for evolution of intelligence?
“Very interesting and elegant study, comparing ravens directly with the great apes,” said Frans de Waal, a professor of psychology who studies primate behavior at Emory University, who was not involved in the study.
“Many people think that animals are trapped in the present, but there is increasing evidence that they think ahead and plan for the future. Not just in an instinctive manner, the way squirrels do when they collect nuts for the winter (the squirrels do not know about winter, certainly not the 1st year juveniles, but still show the behavior),” Professor de Waal explained in email, “but in an intelligent manner which means that they imagine a future and collect the tools they need for it.”
“In this study, the ravens even were asked to compare an immediate reward with a tool they could use later on. They needed to control the urge to take the reward,” Professor de Waal continued.
“The ravens performed in every way either the same or better than apes,” Professor de Waal said in email. “Better even than 4-year-old children.”
“The most remarkable [aspect] about the study is that the ravens have been tested on tasks they do not usually perform in the wild. Until now corvids were mostly tested on food-hiding. This is something they do all the time, and so the criticism was that they may have a very specialized intelligence related to caching behavior,” Professor de Waal explained.
“The present study counters this argument by showing that ravens can do all of these things on tasks (with tools, and barter) that are quite different and unusual for them. Their planning is not dependent on the food-hiding paradigm, which means that, as in humans and apes, it is a general cognitive capacity that they can use for almost anything.”
These findings are all the more remarkable because ravens and apes are separated by 320 million years of evolution. Thus, parallel cognitive abilities arose independently in hominids and corvids, and further, these animals use very different brain structures to accomplish these tasks — yet these two very different lineages ended up becoming very similar.
One possible explanation for this cognitive similarity is convergent evolution, where unrelated lineages develop similar traits in response to similar environmental challenges. In this case, a complex and dynamic social environment could be the main driver of these cognitive traits. Young ravens, like humans and great apes, are highly social and live in groups where social alliances can change quickly. Only after they reach sexual maturity, at roughly three years of age, do ravens form a monogamous pair bond and settle down to defend a territory and raise young.
This study reveals that birds’ intellectual capacities have been grossly underestimated for far too long. Further, as revealed by these ravens’ comparable performances to great apes, this study also highlights how the environment contributes to cognition and drives the evolution of intelligence.
Can Kabadayi and Mathias Osvath (2017). Ravens parallel great apes in flexible planning for tool-use and bartering, Science, 357(6347):202–204 | doi:10.1126/science.aam8138
Nicholas J. Mulcahy and Josep Call (2006). Apes Save Tools for Future Use, Science, 312:1038–1040 | doi:10.1126/science.1125456
Enjoy my writing? Please give me a few handclaps to recommend this piece. Follow me on Medium for more like this.
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
Originally published at Forbes on 25 July 2017.