Birdsong as a model for processing speech-like variation in acoustic signals

Abstract:Steve Nowicki, professor of biology, will describe two problems in understanding the perception and function of animal communication signals, using bird song as a model system. The first concerns the detection of natural perceptual categories in phoneme-like elements that comprise more complex signal elements such as a song phrase, analogous to the identification of phonemes that make up words in human language. The second problem is how to identify features in complex signals that vary extensively within individuals, across individuals in a given population, and across populations, but that the birds nonetheless use to identify population differences with a surprising degree of accuracy. Both problems have implications for understanding neural mechanisms responsible for the perceptual processing of natural communication systems, including human speech.refreshments served