Put two boxes in front of an African Grey parrot. The parrot knows one has a treat and the other doesn’t, but doesn’t know which is which. Shake the empty box so the parrot can hear that it’s empty. She will go over to the other box and open it.
This is called inferential thinking, and (as usual) it was once thought to be unique to humans.
Of course, most of what was thought to be unique to humans turns out not to be. We’ve already learned that great apes are inferential thinkers. And that human children don’t develop it until they’re at least three years old.
African Grey parrots are known to have a very high level of cognitive thinking. (Alex the parrot was a great example.) So German and Austrian researchers developed an experiment to check out their level of inferential thinking.
The team showed the birds two opaque boxes, one of which contained food. They then shook the boxes so the birds could hear that something was inside just one of them. The birds then guessed correctly which box had the food in it, walked over and tipped it over and ate their treat.
Then, in the next stage of the test, the researchers just shook the empty box, producing no sound. This time, the birds instantly figured out that the food must be in the other box and went to that one to retrieve the treat. That’s a feat that humans can’t handle until the age of three. Dogs and monkeys also fail this test. Until now, only great apes and humans were known to be capable of such inferential thinking. (Dolphins can almost certainly do it, but this hasn’t been conclusively tested.) Until now, only great apes and humans were known to be capable of such inferential thinking.
In yet a further stage of the test, the team tried tricking the birds. They attached tiny speakers to their wrists and played recorded sounds of boxes with food being shaken. But the birds could not be fooled. They picked the box with food in it only when the sound matched the exact sound of the boxes they had already recognized.
One other thing: The birds did better when the boxes were shaken side to side, rather than up and down. The scientists inferred from this (a bit of human inferential thinking!) that the up and down motions of the box distracted the birds because it’s similar to the way the birds normally bob their heads when interacting with one another.
Our ability to make logical inferences is considered as one of the cornerstones of human intelligence, fueling investigations of reasoning abilities in non-human animals. Yet, the evidence to date is equivocal, with apes as the prime candidates to possess these skills. For instance, in a two-choice task, apes can identify the location of hidden food if it is indicated by a rattling noise caused by the shaking of a baited container. More importantly, they also use the absence of noise during the shaking of the empty container to infer that this container is not baited. However, since the inaugural report of apes solving this task, to the best of our knowledge, no comparable evidence could be found in any other tested species such as monkeys and dogs.
Here, we report the first successful and instantaneous solution of the shaking task through logical inference by a non-ape species, the African grey parrot. Surprisingly, the performance of the birds was sensitive to the shaking movement: they were successful with containers shaken horizontally, but not with vertical shaking resembling parrot head-bobbing. Thus, grey parrots seem to possess ape-like cross-modal reasoning skills, but their reliance on these abilities is influenced by low-level interferences.