9604359 Cook Birds behave as if the perceive, learn and act upon an object-filled visual world. The ultimate goal of this research is to understand how these small autonomous systems form accurate perceptions of the visual world and use this information to learn and predict information about the relations among objects and events. The proposed research examines these questions by looking at the visual and cognitive mechanisms of one highly visual non-mammalian system - the pigeon (Columba livia). Pigeons are ideal for comparative cognitive studies because the demands to minimize body weight for flight have caused birds to evolve small, compact, and powerful central nervous systems capable of exceptional visual perception and the learning of complex discriminations. The new research builds on the PI's prior NSF-supported research by examining the mechanisms underlying the visual discrimination and abstract conceptualization of same-different relations by this animal. The capacity to direct and recognize difference and similarity implications for our understanding of perception, conceptual behavior, intelligence-related behaviors, and their symbolic mediation by language. One powerful means of studying these relations has been the same-different judgment task. In this task, the subject is asked to respond "same" when two or more stimuli are different from the others. The current proposal follows up on the PI's recent success in producing this type of discrimination in pigeons, and is designed to gain a better understanding of the mechanisms underlying this important and little studied discrimination in animals. The aims of the new research are to analyze and separate the perceptual and conceptual mechanisms of odd-item and 2-item same-different discriminations using multiple types of stimuli. Besides its general implications for our understanding of the learning, study will address interaction of bottom-up, stimulus driven, perceptual processes and top-down, subject-driven, control process in the government of pigeon discrimination and choice behavior. Where appropriate, comparisons to the theories and findings of human and machine perception and conceptual learning and behavior will be made. Overall, this project will provide new evidence concerning the structure and processes of visual-based concept learning - in birds specifically, and in animals more generally.
