They’re Going Ape for Computer Training

Keo, a 47-year-old male chimpanzee at Lincoln Park Zoo, paced the floor of his nonpublic living quarters one recent afternoon, clearly annoyed with his keepers.

In his rolling gait, he would stride up to a glass wall, stop and glare at the humans on the other side. He was supposed to have been at work on his computer at 1:30 p.m., but now it was 1:40 p.m., and the door to a small adjoining room with the computer was locked so he couldn’t enter.

Behind the glass, Steve Ross, the zoo’s supervisor of behavioral and cognitive research, was a little annoyed too. He was waiting for two visitors, 15 minutes late, who were coming to see how he has been training Keo and other apes to use a touch-screen computer in order to measure their cognitive abilities.

The work Ross and his colleagues are doing is part of a movement in American zoos to hire their own scientists and allow others to use zoo animals for sophisticated studies that used to be done almost exclusively in university settings.

Among the experiments being conducted at Lincoln Park is long-term research on using computers to “talk to the animals.” The idea is to get the apes to learn to use computer programs to communicate preferences on food, activity and living space. More broadly, Ross said, the work should add to the scientific literature on how and to what extent apes are able to think and perceive the world.

Keo’s annoyance at being delayed suggests he is keenly aware of time, even though he can’t tell time, Ross said. He and other apes judge the time of day with uncanny accuracy and are able to anticipate their scheduled activities, such as feeding and training sessions.

“Keo seems to gauge the time of day starting from the time keepers arrive every morning,” Ross said.

That kind of obvious intelligence is driving zoos to develop projects like the computer study.

When Lincoln Park opened its $25.7-million Regenstein Center for African Apes in 2004, it was designed to handle such research. Besides housing the apes in natural habitats for public viewing, the building houses the Lester E. Fisher Center for the Study and Conservation of Apes, a zoo-funded research component that employs Ross.

Some of its experiments are conducted in public view, such as an artificial termite mound where visitors can sometimes watch as chimps and gorillas use sticks, or try to use them, as tools to retrieve favorite snacks.

To avoid distractions, the computer experiment is conducted in nonpublic areas like Keo’s quarters and the basement holding areas the apes go in the morning while keepers clean their living spaces.

Keo was the first ape to be trained to use the screen. Now every Monday through Friday at 1:30 p.m., he and another chimp in his troop, 20-year-old Vicky, wait anxiously for the researchers to arrive. Usually Keo does the first session, then Vicky.

Currently Keo is doing a task Ross calls “match to sample.” Keo squats in front of a 42-inch computer screen that places a 2-inch-high photo icon -- the face of a chimp he has never met -- somewhere on the screen. If he touches it, little edible balls called Primatreats, in pina colada and banana flavors, roll out of a slot at the bottom of the screen as a reward.

Moments later, two chimp face icons flash on the screen, the one he just touched and one of another chimp Keo has never met. If he touches the image of the first chimp, he gets a reward. If he touches the second image, he gets nothing.

“In each session, Keo has 30 opportunities -- called trials -- to win a reward for touching the right face,” Ross said. “He has 10 minutes to get through each trial. Then we cut the experiment off for that day.

“Keo has gotten so good at it, he usually does all 30 in 2 1/2 minutes.”

The exercise with the faces is just for training, getting Keo used to the idea that the icons mean something, that there are right and wrong answers.

“In the next three to four years, as more of our apes become accustomed to the screen, we can devise programs for them that we can use to ask them questions about their world and how they perceive it. In a fashion, it might give them a way to talk to us.”

He said he and colleagues hoped to devise preference tests so the apes could say what foods they liked the most, which areas in the ape house they most enjoyed living in and how they reacted to individual keepers or visitors.

“For food preferences, for instance, we might put various food items on the screen, and they’ll receive the item they select. Once they understand that they get the item they choose, that should tell us a great deal about their food preferences.”