Environment

Can Ravens Think Abstract Thoughts?

Animals do the most amazing things. Read about them in this series by Janaki Lenin.

Researchers at the University of Vienna tested the ability of crows to empathise with others. Credit: Jana Müller, Universität Wien

Researchers at the University of Vienna tested the ability of crows to empathise with others. Credit: Jana Müller, Universität Wien

Ravens know when they are being watched. They are careful to stash their scavenged booty when no other bird is looking. If they are careless, another might raid their cache when their backs are turned. But they can’t always avoid observers. Then, they use the cover of boulders and trees to hide from view. Should no visual barrier be available, they conceal their stash quickly and don’t go near the site lest they lead a snooping raven to it. If another raven ventures too close to their hoard, they lead them away. Scientists use this behaviour to open a window into their minds.

Few animals, like great apes, dolphins, and ravens, seem to have the ability to imagine what others might see, want, or think. This is called Theory of Mind. Such claims are controversial since the animal behaviour can have many causes. A new study claims to address the shortcomings of previous experiments and prove common ravens have a theory of mind.

Typically, scientists design experiments to show if animals can reason what others can see. For instance, in one experiment, they left a treat in a visible spot and hid another from a dominant chimpanzee. But both morsels were within the view of a subordinate animal. The junior preferred to grab the hidden morsel, demonstrating that it knew what the dominant one couldn’t see.

Or, in another case, rhesus macaques stole a grape that a human observer couldn’t see. But they didn’t when the grape was in plain sight. In both these studies, chimpanzees and macaques showed they were aware of what another could see.

Sceptics weren’t convinced. They said these experiments were designed or interpreted wrongly. Most of these animals can look in the direction another points. So they may just follow the other’s gaze without displaying any great cognitive powers. The experiments don’t differentiate between ‘behaviour reading’ and ‘mind reading.’

Any experiment seeking to prove what animals know has to rule out behaviour cues. For this reason, the animal version of theory of mind remained unresolved. Until it could be convincingly proven, the capacity for abstract thought is uniquely human. Researchers from the University of Vienna, Austria, and University of Houston, US, took up this challenge. They conducted a study on common ravens to show they are mind readers.

Ravens hid their food only well when dominant conspecifics were visible and audible at the same time. Credit: Jana Müller, Universität Wien

Ravens hid their food only well when dominant conspecifics were visible and audible at the same time. Credit: Jana Müller, Universität Wien

Common ravens scavenge across the Northern Hemisphere. Besides some mammals, it is the only species that can follow another’s gaze. By looking in the direction that another is seeing, the birds can spot a predator or observe where another raven hides its stash of food to steal it later.

Thomas Bugnyar has studied ravens for almost two decades. As adolescents, these birds hang out with their friends and fall out with them. They mate for life as mature adults, defend their territories from intruders, and raise successive generations.

“There is a time when who is in the pack, who’s a friend, who’s an enemy can change very rapidly,” said Bugnyar in a press release. “There are not many other species that demonstrate as much social flexibility. Ravens cooperate well. They can compete well. They maintain long-term, monogamous relationships. This all makes them a good place to look for social cognition, because similar social pressures might have driven the evolution of similarly advanced cognitive capacities in very different species.” Their social lives are similar to humans.

Bugnyar and his colleagues designed their experiment in such a way that their subjects couldn’t use others’ gaze to anticipate what they could see. Instead, they had to imagine what others could see.

The subjects were ten not-yet-mature hand-raised crows. The experimental setup at the Haidlhof Research Station, Austria, involved two adjoining rooms. Ravens in one room could look into the other through two windows when they were open. The research team gave cheese and dry dog food to a raven in one bare room that had no obstacles to block the other raven’s view. The bird hid its treat quickly since its neighbour was watching.

The researchers closed the window and played raven calls from the next room. Even though it may have thought a rival was in that room, it must have realised it can’t be seen. It took its time finding a satisfactory caching spot.

This is how ravens behave in the wild. If the others were out of sight, they could be put out of mind. Their behaviour could be used as a proxy for how they might’ve understood a situation.

The researchers drilled peepholes in the window shutters. They trained the ravens to look through these holes into the next room as their human trainers hid a cheese piece. The birds then had 30 seconds to find these pieces. Only if they passed this test could they participate in the study.

The study claims ravens can connect the acoustic information with their own experience mentally. Credit: Jana Müller, Universität Wien

The study claims ravens can connect the acoustic information with their own experience mentally. Credit: Jana Müller, Universität Wien

For the experiment, the team closed both window shutters but opened only one peephole. They gave a cheese piece to a raven in one room. To fool it into thinking it was being observed, the team played raven calls from the other room. It imagined its neighbour could watch it through the open peephole, and it hastily hid its treat.

When the peephole was closed, the raven acted as if no other bird was watching them in spite of the audio playback of raven calls. Using sounds and peepholes took care of the criticism of previous studies. “The results indicate that ravens mentally integrate information about the others’ audible presence and their own visual experience of seeing through the peephole, which fits one of the recent hypotheses of how a Theory of Mind could work,” says Bugnyar.

However, Robert Lurz, author of Mindreading Animals, and Professor of Philosophy at Brooklyn College, U.S., isn’t convinced. “Their test is quite clever, and I believe it has the potential to solve the logical problem with additional controls,” Lurz told The Wire. “But I am skeptical it can tell us whether ravens are mind-reading or behaviour-reading. The test does not rule out the possibility that the ravens in the window and peephole conditions are just responding to a common perceptual cue.”

Ravens can not only see each other through the open window, they can also hear each other. So even though the researchers didn’t playback raven calls when there was a visible rival, sound was a common factor to both open window and peephole situations.

“Both the live raven and the speaker are producing the same types of sounds,” says Lurz. “Since the open window and peephole conditions are perceptually similar, you would expect the ravens to behave in similar ways in both conditions – which they do.”

Lurz also points out that the study doesn’t make a distinction between looking and seeing. “Looking and seeing, as is well known, are not the same thing. Seeing is a subjective mental state, while looking, what researchers call ‘line of gaze,’ is an objective spatial vector from a subject’s eyes to objects before the eyes. Many objects that we look at in our daily lives cannot be seen. For example, dust particles and white rabbits camouflaged in snow.”

During the peephole training sessions, ravens learn they can look into the adjacent room. Based on this experience, they infer that any bird in that room can see them, say the researchers.

Lurz contends the subjects may learn to both look and see into the adjacent room. “The researchers assume the ravens are learning to see and not just look. But they have no control for this. So it is quite possible – and not altogether unlikely, given that ravens are known to follow and represent others’ line of gaze – that the ravens simply learn that the peephole affords looking at or line of gaze into the adjacent room. This knowledge would allow them to infer that another raven can look through the open peephole condition, just like a raven in the open window condition. Thus, since the ravens think the same sort of thing is happening in these two conditions, they should behave in the same sort of way, which they do. But this isn’t theory of mind, since looking is not a subjective mental state like seeing.”

Until researchers come up with an experiment that rules out these variables, the jury on whether animals can have abstract thoughts is still out.

The study was published on February 2, 2016, in the journal Nature Communications.

Janaki Lenin is the author of My Husband and Other Animals. She lives in a forest with snake-man Rom Whitaker and tweets at @janakilenin.