The primary auditory cortex (AI) of mammals shows several superimposed functional organizations when explored with simple signals such as pure tones. Basic spatial organizations have been described now for stimulus frequency, bandwidth, spectral envelope, frequency modulations, and binaural interaction. The consequences of these organizations for the cortical representation of complex signals, in particular of elemental speech signals, is of special interest since similar principles may provide the basis for the perception and categorization of speech in humans. The representational principles will be explored with elemental speech signals in AI of naive cats and cats that have acquired high behavioral affinity to the signals. The behavioral relevance of the studied complex signals will be established by engaging the animals in a psychophysical task of signal discrimination and generalized classification. The task will be performed under varying stimulus and environmental conditions such as by using different stimulus intensities and different levels of background noise. These manipulations are designed to force the animal to utilize more generalized classification schemes that operate independent from signal level and bachground conditions, thus approaching human discrimination and classification abilities. Determining the cortical representation of elemental speech sounds with distinct phonetic features in naive and highly trained animals will illuminate basic attributes of complex signal representations as well as more refined attributes after learning to discriminate and classify the signals. The emergence of refined spatial-temporal patterns of cortical activity by learning-induced plasticity provides a basic model of speech representation. The proposed model can potentially be extended to developmental aspects of early cortical processes for speech sound representation and can be applied to the exploration of the basic auditory processes underlying the perception and categorization of distinct phonetic features.
Showing the most recent 10 out of 72 publications