9616342 LESLIE The human brain is sufficiently organized in the first year of life to support the emergence of higher cognitive functions. One key function is the recognition of physical objects in a changing environment. Existing research shows that infants have impressive abilities in this regard. This project will investigate the development of the cognitive mechanisms underlying this skill. The approach will draw upon recent advances in understanding mechanisms of visual attention in adults. A series of experiments will explore how infant attention picks out and "holds onto" physical objects under a variety of circumstances. The brain mechanisms underlying such skills are crucial for learning about the world, for learning verbal labels, and for developing numeracy. The ability to attend to physical objects requires that objects be distinguished from one another. One way this can be done is to recognize that two objects have distinctive features. Another way is simply to note that the objects simultaneously occupy different locations in space. These simple solutions do not always work (a) because objects can change the features they present (e.g., with a change in viewpoint) and (b) because objects can move in and out of view (e.g., by passing behind an occluder). For organisms that live in a complex, moving, three dimensional world, such complex changes are the rule rather than the exception. The problem that this creates for infants as they track physical objects can be likened to a radar operator who has to track "objects" on a radar screen. When two blips appear on different locations on the screen, the operator can register the number and location of the objects and track them continuously as they move across the screen. However, one of the blips might move off the edge of the screen and after a short period another blip might appear. Is this the same object as the one that disappeared? How many objects should the operator report, two or t hree? The infant brain must regularly face a similar dilemma because of gaze redirection, occlusions, and so forth. Physical objects in the world do not have transponders to identify themselves to the infant brain! Despite this, infants manage to identify and track physical objects, learn verbal labels for them, and even to enumerate small sets of objects. This project will investigate how infants first individuate, then re-identify physical objects that from time to time pass behind occluders. Experiments will measure how infants allocate their visual attention to physical objects based on both locational and featural information and on the integration of the two kinds of information. Experiments will also determine what limit there may be on how many objects infants can track at any one time. The results of these studies will reveal properties of early emerging, and thus basic, brain functions that support learning and general cognitive development. Longer term, such investigations will bear upon early brain and cognitive development and education issues as well upon the design of artificial systems that can behave intelligently in complex environments. ***
