One of the crucial aspects of vocal processing in the brain is the isolation of auditory features that identify an individual speaker or defined auditory objects like vowels and consonants. Hearing damage that compromises this ability cannot be remedied through traditional means; in particular, hearing loss that occurs early in life leads to a failure to develop the appropriate mechanisms for vocal processing, and neither those mechanisms nor their development is well understood. Songbirds use vocalizations to identify individuals, and many species can continue to learn new associations throughout life. Song recognition behavior has received particular attention in European starlings (Sturnus vulgaris), whose social nature makes them amenable to operant conditioning paradigms in which subjects learn to distinguish between songs from different individuals. Recent findings have identified a forebrain region, the caudomedial mesopallium (CMM), which contains neurons that develop selectivity for songs learned in the training procedure. Thus, starlings represent an attractive system in which to study how such selectivity arises, and the long-term objective of this research is to develop rigorous statistical models of CMM receptive fields (RFs). These models will shed light on the neuronal circuitry and mechanisms underlying auditory processing, and on how neuronal representations of auditory information are used by animals in making decisions.
The specific aims of this proposal support this goal by examining how two distinct elements of starling song contribute to song recognition behavior. Starling songs comprise a series of discrete stereotyped elements called motifs that are sung in a non-random pattern, and individual motifs are composed of a unique combination of spectrotemporal features. This proposal describes experiments that test the relative contributions of sequence and feature composition under the hypotheses that (1) motif selectivity arises at both the behavioral and physiological level from sensitivity to a combination of specific features and (2) motif selectivity at the behavioral and physiological level is sensitive to temporal context. This research will contribute to treatments for hearing loss and language deficits by deepening our understanding of how the brain processes complex sensory information. The methods it develops for analyzing auditory signals may also be useful in the creation of auditory prosthetic devices. ? ? ?
