Humans learn about their world through both personal experience and social communication but complete information is rarely available. In fact, we are remarkably adept at drawing conclusions from only partial information (for example, a doctor diagnosing a disease or a detective solving a crime) and imagining alternate scenarios (an especially useful enterprise for scientists). Like humans, many animals are able to learn about the world through personal experience, as Pavlov showed in dogs and Skinner showed in rats and pigeons. In previous NSF-funded research, the principal investigator showed that rats have the capacity to go beyond direct experience and training and make causal inferences, that is, inferences about a cause-effect relationship they had never previously observed. Causal inferences were much like those of a scientist or a human child in that they originated from the rat's interactions with the world. Rats also act as if they expect a cause to be present even if it is not directly observable. Thus, rats exhibit at least some of the hallmarks of human reasoning. The current research project builds on these results by exploring other ways in which animals are able to reason about their world. Can a rat, like a doctor or detective, determine a cause-effect relationship from only partial and incomplete information? If so, what are the psychological mechanisms that support these causal inferences? To what extent can a rat reason about hidden events in order to solve ambiguous problems?
Answers to these questions can determine aspects of cognition that are uniquely human and aspects that humans share with other animals. Such knowledge may impact the development of artificial intelligence systems designed to make adaptive decisions and inferences based on limited prior knowledge. The principal investigator also serves as faculty mentor in two programs for high-school students in the greater Los Angeles area (drawn from diverse backgrounds) and will involve the students in this research.
