Much research has been done on how children acquire names for things. Just look at any mother and child playing together and at some point the mother will ask the child what some object is. We ask children to show us their noses and cheeks; we ask them to show us trees and grass; shirts and pants; keys and bottles; juice and bananas; and just about everything else. The question is, when a child gives an object a name, what does the name mean to the child and how does the child go about giving objects their names?

    For a long time, researchers were convinced that perceptual features, most notably shape guided children's naming. We wanted to find out whether children's naming might be guided not by shape, but rather, by what an object can do. When you think about what an object can do, you are thinking about its causal powers--what kind of effects does it have on other objects or on people. We believe that causal powers motivate children's naming. The difficulty in testing whether shape or causal powers guide children's naming is that in the world, shape and causal powers are highly correlated. What an object looks like often is equal to what it can do.

    To test our hypothesis, we designed a device called the blicket detector, which would allow us to separate an object's shape from its causal powers. The device is a wooden box with a lucite top which can light-up and play music when activated by a remote control. When certain objects are placed on the detector we can activate its light and music, giving the appearance that those objects have the power to activate the device. This is a causal power--whether or not an object "turns the machine on."

    We gave three and four-year-old children sets of four objects. One at a time, each of the four objects was placed on the detector; two of the objects turn the machine on, and two of them did not. I would pick up one of the objects that turn the machine on and call it a blicket and then ask the children which of the three remaining objects was the other blicket. We presented children with three types of sets of objects. In one type of set, all the objects were perceptually identical. In this case, children easily chose the other object that they had seen turn the machine on. Their choices could have been based on perceptual feature or causal property. In the second type of set, the four objects were perceptually different. As in the first case, children easily chose other object that had turned the machine on, making choices based on causal property.

    In the third type of set, there were two pairs of perceptually identical objects. In our demonstration one from each identical pair turned the machine on. In this case, the perceptual feature is conflicting with the causal power of the objects. In contrast to the other two types of sets, we found that with this type of set children were equally likely to choose the other object that had activated the machine or the perceptually identical object. These results suggest that when causal property and perceptual features are equally evident, children are equally as likely to use causal powers as they are to use perceptual properties when naming objects.