Thinking About Thinking in Computers, Robots, and People

* This material is based upon work supported by the National Science Foundation under Grant No. 0433653. Abstract Although much research has focused on emerging understandings of representation and mind, very little research has explored the adult understanding of different representational systems. The current experiments demonstrate that adults differentiate intentional and nonintentional representational systems when reasoning about them. In the first experiment, subjects were taught a novel feature of either a computer or a person, and asked whether the feature characterized other representational systems. Results indicate that when making inductions from a person to a computer, they do so more for nonintentional mental properties than intentional mental properties. In a second experiment we asked adults to make predictions about the behavior of a person, a computer, and a robot, and found that they assume that humans engage in more object/category related behaviors, and that computer systems engage in more spatial/featural behaviors. Combined, these data establish the existence of a distinct category of intentional mental entities, and give guidelines for its deployment in predicting a variety of representationally mediated behavior.