We employ mathematical and computational approaches to investigate auditory and vestibular systems. To be of use, a mathematical model must make experimentally testable predictions and it should be clear which features of the model are critical for correspondence with experiment.
In general, the requirements of prediction and correspondence pose a significant challenge for modeling in view of the complexity of biological systems and the evolution of many solutions to each environmental problem. We tackle these issues using comparative modeling. By comparing the structure of and predictions arising from different models, we learn which features of a system are responsible for particular experimental observations. Using this approach we improve our understanding of contemporary experimental observations, make experimentally testable predictions, and motivate new experiments (see figure for an example). Progress in the treatment of dysfunction can be accelerated by understanding the mechanisms underlying normal and impaired function.