The overall goal of the present grant proposal is to characterize the process by which listeners recognize words in fluent speech. The proposed research examines the nature of the competitor set that is activated during word recognition in contnuous speech, focusing on lexical neighborhoods, sub-phonemic acoustic/phonetic variation and embedded lexical competitors. A novel program of research on perceptual learning will also be initiated. Eye movements will be monitored as participants follow spoken instructions to pick up and move (with a mouse) line drawings of concrete objects on a computer monitor (e.g., "pick up the candy. now put it above the circle".) These studies will provide the finest grained information to date about the time course of lexical activation in continuous speech. The availability of an explicit linking hypothesis means that these data can be used to compare and contrast models that are explicit enough to generate predicted activations over time for lexical candidates. The proposed research will help bridge the gap between research on spoken word recognition research and research on speech perception and it will provide important new information about the effects of perceptual learning on lexical processing. The results will be important for models of spoken word recognition, and will also inform models of the effects of neurological damage on spoken word recognition and models of lexical development in children. Moreover, because the eye movement paradigm does not rely on meta-linguistic judgments that are difficult for these populations, it can be easily extended to empirical investigations of brain-damaged patients, young children, and infants.