During an early epoch of development, the brain is highly adaptive to the stimulus environment. Repeatedly exposing young animals to a particular tone, for example, leads to an enlarged representation of that tone in primary auditory cortex. While the neural effects of simple, single-frequency tonal environments are well characterized, the principles that guide plasticity in complex tone environments, as well as the perceptual consequences of cortical plasticity, remain unclear. To address these questions, this dissertation documents the neural and perceptual effects of simple and complex manipulations to the early acoustic environment.

First, I show that rearing rat pups in a multi-tone environment leads to complex primary cortical representational changes that are related to the statistical relationships between experienced sounds. Specifically, tones that occur together within short temporal sequences tend to be represented by the same groups of neurons, whereas tones that occur separately are represented separately. This suggests that the development of primary auditory cortical response properties is sensitive to higher-order statistical relationships between sounds.

The observed neural changes are accompanied by perceptual changes. Discrimination ability for sounds that never occur together within temporal sequences is improved. Heightened perceptual sensitivity is correlated with heightened neuronal response contrasts. These results suggest that early experience-dependent neural changes can mediate perceptual changes that may be related to statistical learning.

Finally, I develop and experimentally test a model of the relationship between cortical sensory representations and perception. The model suggests that cortical stimulus representations may function as the neural representation of previously encountered stimulus probabilities, and makes predictions about how changes in these representations should affect perception within a statistical inference framework. Preliminary behavioral results support the model predictions, suggesting that one function of early experience-dependent plasticity may be to internalize stimulus distributions to shape future perception and behavior.