Hard-wired to learn by imitation

The box was painted black and had a door on one side and a bolt running across the top. The food was hidden in a tube behind the door. When they showed the chimpanzees how to retrieve the food, the researchers added some unnecessary steps. Before they opened the door, they pulled back the bolt and tapped the top of the box with a stick. Only after they had pushed the bolt back in place did they finally open the door and fish out the food.

Because the chimps could not see inside, they could not tell that the extra steps were unnecessary. As a result, when the chimps were given the box, two-thirds faithfully imitated the scientists to retrieve the food.

The team then used a box with transparent walls and found a strikingly different result. Those chimps could see that the scientists were wasting their time sliding the bolt and tapping the top. None followed suit. They all went straight for the door.

The researchers turned to humans. They showed the transparent box to 16 children from a Scottish nursery school. After putting a sticker in the box, they showed the children how to retrieve it. They included the unnecessary bolt pulling and box tapping.

The scientists placed the sticker back in the box and left the room, telling the children that they could do whatever they thought necessary to retrieve it.

The children could see just as easily as the chimps that it was pointless to slide open the bolt or tap on top of the box. Yet 80 percent did so anyway. "It seemed so spectacular to me," Mr. Lyons said. "It suggested something remarkable was going on."

...Mr. Lyons loaded a movie on his computer in which Charlotte eagerly listened to him talk about the transparent plastic box.

He set it in front of her and asked her to retrieve the plastic turtle that he had just put inside. Rather than politely opening the front door, Charlotte grabbed the entire front side, ripped it open at its Velcro tabs and snatched the turtle. "I've got it!" she shouted.

A chimp couldn't have done better, I thought.

But at their second meeting, things changed. This time, Mr. Lyons had an undergraduate, Jennifer Barnes, show Charlotte how to open the box. Before she opened the front door, Ms. Barnes slid the bolt back across the top of the box and tapped on it needlessly. Charlotte imitated every irrelevant step. The box ripping had disappeared. I could almost hear the chimps hooting.

Ms. Barnes showed Charlotte four other puzzles, and time after time she overimitated. When the movies were over, I wasn't sure what to say. "So how did she do?" I asked awkwardly.

"She's pretty age-typical," Mr. Lyons said. Having watched 100 children, he agrees with Dr. Horner and Dr. Whiten that children really do overimitate. He has found that it is very hard to get children not to.

If they rush through opening a puzzle, they don't skip the extra steps. They just do them all faster. What makes the results even more intriguing is that the children understand the laws of physics well enough to solve the puzzles on their own. Charlotte's box ripping is proof of that.

Mr. Lyons sees his results as evidence that humans are hard-wired to learn by imitation, even when that is clearly not the best way to learn.