Project Overview
Vocal learning in songbirds requires a cortical brain area, called LMAN, which generates song variability that is required for learning. We hypothesize that this variability is generated from ‘random’ internal circuit dynamics within LMAN. In this view, LMAN may be a circuit specialized for the generation of neuronal variability. One candidate for the generation of such variable activity is high-dimensional chaos, the generic dynamical state of a balanced E-I network. Another possibility is that LMAN activity results from traveling waves, similar to those seen during development in the retina.
In order to distinguish these different hypotheses, we are using a combination of optical and electrophysiological methods to analyze the spatial patterns of activity in LMAN during singing. In collaboration with Professor Adrienne Fairhall at the University of Washington, we have begun to explore the theoretical implications of these different models of neuronal noise generation.
