Intelligent Birds & What Experiments Show About Them

In an Oxford experiment, a crow named Betty bent straight wire into a hook and used it to lift a small bucket of food from a tube. The experiment showed how one bird could shape an unfamiliar material into a tool that fit the task. Field researchers have also documented New Caledonian crows shaping hooks from plants.

Kea watched people reach into jars holding different proportions of rewarding and unrewarding tokens. The parrots tended to choose the hand that had sampled from the better odds. They could also take account of a barrier that made some of the tokens in a jar unreachable.

After learning spoken labels for numbers, an African gray parrot named Alex was asked about groups of objects. When the requested number was absent, he sometimes answered “none,” a word he had learned in another context. The researchers described it as evidence of a zero-like concept in this individual.

Researchers watched captive scrub-jays recover worms and peanuts they had hidden. The birds preferred the worms after a short delay but shifted toward peanuts after enough time for the worms to spoil, once they had learned the worms’ shelf life. The choice combined what the birds had hidden, where it was, and how long it had been there.

In one study, trained ravens chose a tool or a token they could use later to get food. Some of the delays ran overnight. Choosing the useful object when the chance to use it was still hours away is consistent with planning, though researchers debate how much practice and learned associations explain.

In one experiment a box required two tools: one to puncture a barrier and another to reach food. Some Tanimbar corellas, also called Goffin’s cockatoos, carried both tools together to the box when the barrier was present. They were less likely to bring the set when only one tool would do.

Researchers let nutcrackers hide seeds at sites in a room and later return to search for them. Even after 285 days, the birds recovered more caches than researchers expected from random searching. That is a measure of long-term place memory under controlled conditions.

In a British woodland, researchers taught a few great tits to open a feeder by sliding its door either left or right. After the birds returned to their groups, the favored technique spread locally. The feeder worked both ways, so neighboring groups could settle on different ways of getting the same food.

Trained pigeons learned to peck in response to slides of paintings by either Monet or Picasso. They could often make the same distinction with paintings they had never seen during training. The experiment measured how well the birds learned and generalized visual categories.

Scientists recorded activity in the hippocampus while chickadees hid and retrieved seeds. Distinct patterns appeared during individual hiding events and reappeared when the birds retrieved those seeds. The researchers compared the patterns with barcodes, which could help distinguish one visit from another at the same location.

Pinyon jays watched unfamiliar birds interact with individuals whose rank they already knew. Their responses in later encounters suggested they had used that information to judge where a newcomer stood relative to themselves. The setup let researchers test what the observer had learned before its own encounter.

Three- and four-day-old domestic chicks watched familiar objects disappear behind two screens. When researchers moved some objects between screens, the chicks tended to approach the screen concealing more objects. The task tested whether they could keep track of small sets after the objects had gone out of sight.

Captive rooks put stones into a tube of water to raise a floating worm within reach. When they could choose stone sizes, they preferred larger ones, which displaced more water. The experiment showed how the birds could adjust their actions to the job the stones had to do.