Researchers trained a group of kea at Willowbank Wildlife Reserve in Christchurch to operate a touchscreen laptop with their tongues. Photo / Amalia Bastos
A group of kea have been trained to use touch screens as part of a new study into how they interpret virtual and real-world environments.
The study, by the University of Auckland, shows the alpine parrot treats virtual and real-world environments as continuous and equivalent.
To achieve this, researchers trained a group of kea at Willowbank Wildlife Reserve to operate a touchscreen laptop with their tongues.
They then presented them with a series of tasks that took place either fully in the real world, fully on-screen, or with a mixture of both.
First, the kea saw a real ball placed onto a seesaw which tilted so the ball would roll into one of two real boxes. They correctly indicated which box they thought the ball was in by touching it with their beaks.
They then performed the same task when all elements were replaced with virtual ones on a screen. Again they correctly indicated which box the ball was in.
To test whether they expected physical events taking place in a virtual environment to be continuous with the real world, they were presented with a version of the task which had the same virtual animation of the seesaw as before, but it now appeared to deposit the ball into one of two real boxes placed in front of the screen.
The kea continued to select the box the ball was seemingly deposited into, which suggests they expected events on-screen to continue into the real world.
An author of the study, Patrick Wood, said training the birds to operate touchscreens was an interesting challenge.
"A parrot's beak is a lot like your fingernail: it won't activate a touchscreen. So, we had to teach them to lick the screen with their tongues. Once they acquired this skill, they quickly gained confidence using the touchscreens and they really seem to enjoy it, too," he said.
The researchers also presented the parrots with an additional experiment that pitted real and virtual objects against each other.
This confirmed their findings were not due to simpler explanations, such as selecting the box tokens moved closest to, and that the birds showed no preference between real and virtual objects when they were directly compared against each other.
The findings are in contrast to a recent study which found 19-month-old human infants did not expect the real and virtual worlds to interact and therefore did not expect a virtual seesaw to deposit a virtual ball into one of two real boxes.
Lead author Amalia Bastos said the findings open new possibilities for scientific investigations into kea behaviour and intelligence that could rely on this technology.
"Our study validates the use of virtual reality and tasks blending the real and virtual worlds for use with this species. This potentially has implications for other parrot species as well," she said.
"However, further work is needed to determine whether kea with extensive experience of screens might begin to dissociate the real and virtual worlds, and what types of experiences might shift their understanding of screens closer to that of human infants."