Do you know what I’m thinking? Probably not, but you will have a pretty good idea that I have my own thoughts, perceptions and intentions. This “theory of mind” is something that we take for granted, yet the ability to step into somebody else’s shoes is what has helped to make us such a success as a social species. Is it a trait only humans have?
Over the next few columns in the Uniquely Human series, I will be exploring the mind-reading abilities of humans and animals in all forms. We start with one of the most basic aspects of theory of mind: the ability to deceive somebody. Successful deception, after all, requires you to imagine what somebody else would think. The question is, are we the only species capable of such trickery?
To find out, we first have to look to experiments conducted more than 30 years ago. In 1978, psychologists David Premack and Guy Woodruff wondered if chimpanzees have theory of mind. At the time, researchers thought that deception might be a good way to get at the question, so Premack and Woodruff tried to see if chimpanzees could be taught to deceive fellow chimpanzees. If a snack was hidden in one box, could they learn to trick a competitor into searching in the second box by pointing at it? The researchers only managed to teach two of four chimpanzees to do this, and it took them a whopping five months to get it right. It wasn't the most convincing evidence.
Open and shut case? Not so fast. Chimpanzees have trouble comprehending the pointing gesture in the first place. Pointing with their hands and fingers is not part of their own behavioral repertoire. It simply wasn't a fair test.
Then, in 1987, science writer Virginia Morell witnessed a now-classic example of apparent chimpanzee deception while visiting Jane Goodall at Gombe National Park in Tanzania. The two were hiding in a feeding station, a spot where Goodall and her team sometimes handed bananas out to passing chimps. A high-ranking male called Beethoven ambled by, along with Dilly, an orphan who Beethoven had adopted. Beethoven immediately let out a food call, standard behavior for a chimpanzee. After promptly devouring the whole bunch – chimpanzees tend not to share food, even with their infants – Beethoven settled down for an afternoon nap, leaving a hungry Dilly to groom him. Goodall then made eye contact with Dilly and held up a single banana. "It was as if a signal passed between her and Dilly," Morell recounted on the website Slate this year. "Dilly did not utter a food cry, as chimpanzees normally do, but simply watched as Goodall placed the banana outside on the ground. Then Dilly quietly made her way to the fruit, downed it in three bites, and just as quietly returned to the snoring Beethoven."
Not only did Dilly appear to infer Goodall's intention in leaving the banana for her, but she also understood that Beethoven would gobble up the banana himself if given half a chance. By stifling her food call, she could keep Beethoven in the dark, none the wiser about her illicit snacking.
Trick of the mind
Still, at the time it was verboten for researchers to infer emotion, intention, or motivation within an animal. Goodall would describe the behavior "as if" it was deceptive.
And it certainly was a stretch to say that this meant chimpanzees had theory of mind. Scientists needed empirical evidence derived from controlled experiments before they could make any firm conclusions.
One way that researchers tried – experimentally – to get at the question of chimpanzee theory of mind in the subsequent years was “gaze-following”. Could a chimpanzee tell what you're looking at by following your gaze? In 1996, psychologists Daniel Povinelli and Timothy Eddy gave some juvenile chimps a test: the apes were given a choice to ask for food either from a human who could see them clearly, or from to a human whose eyes were hidden, such as by wearing a bucket on his head. In most cases, they begged indiscriminately. It seemed that the young chimps didn't care whether a human was actually able to see them while begging for food in all but the most obvious cases.
If chimps couldn't even understand what others could and couldn’t see, then it seems unlikely that they could deceive, let alone attribute more sophisticated goals and intentions to others. How could the story of Dilly and Beethoven be explained? Like the experiments of the 1970s, the reason could be that Povinelli and Eddy had thrust their chimpanzees into a strange situation. The experiments had arbitrary conditions that were perhaps not obvious to the chimps, such as the rule that they were only allowed to make one choice per trial. From the chimp's perspective, it could actually be quite reasonable to beg from everyone until someone hands over a tasty treat.
The flaws of such experiments led evolutionary anthropologist Brian Hare to develop a more naturalistic, species-appropriate test for chimpanzees. That meant he had to think like a chimpanzee. What he devised was a clever situation in which a subordinate chimp could compete for food with a more dominant chimp.
In his experiments, Hare set up enclosures containing two chimps, one at either end. He placed food in the centre. Thanks to well placed barriers, sometimes both chimps could see the food; sometimes only one.
In one instance, Hare allowed a dominant and a subordinate chimp to watch as he put food in the middle. However, the food was obscured from the dominant one once it was placed down.
As is typical for chimpanzees in this sort of scenario, the subordinate all but ignored the food, leaving it for the dominant. Subordinate chimps know better than to take food from dominant group members, just as Dilly knew not to let Beethoven catch her eating that banana. Even though the dominant couldn’t see the food, it knew where it was.
Then Hare added a twist: when the dominant chimp was replaced with a second dominant who hadn't seen the food – all she could see was opaque barriers – the subordinate had no problem gobbling it down. The conclusion? Chimpanzees don't just know what others see; they also know what others know.
The so-called "Hare task" has since been adapted and modified for a wide range of animals. A clever experiment in which rhesus monkeys could steal food either from a silent box or from a box outfitted with bells showed that they anticipated what others would and would not hear. Another set of studies demonstrated that ravens, too, know what others do and do not see. The food-caching corvids were more likely to hide their caches in a spot hidden from the view of others. If, while caching, they discovered a potential raider watching them, the ravens usually picked up and tried to hide their cache elsewhere. Together, these experiments demonstrate both the ravens' gaze-following abilities and their tactical deception skills.
How about our pets? Well, researchers know dogs can be sneaky. In one experiment, dogs were instructed not to take food from boxes, a few of which were rigged with noisy bells. Yet when a researcher wasn’t looking, the dogs would steal a meal, and would deliberately avoid the boxes with bells to avoid detection.
The evidence gathered so far suggests that many clever animals are capable of deception, and some can make basic predictions about the "knowledge states" of others. They can predict what others can see, and sometimes what others can hear, and they can use that information for their own benefit. They know who is knowledgeable and who is naive. Perhaps we should give their mind-reading talents more credit.