Neural rhythms as priors of speech computations

Endogenous rhythms of auditory neural circuits have a striking resemblance to the temporal modulations of incoming speech signals. Here, we show that these rhythms may serve as priors for speech recognition, encoding knowledge of speech structure in the dynamics of network computations. In a network of coupled oscillators, we find that speech is readily identified when characteristic frequencies of the oscillators match low-frequency circuit rhythms in the auditory cortex. When signal and circuit rhythms are mismatched, speech identification is impaired. Compared to a baseline recurrent neural network without intrinsic oscillations, the coupled oscillatory network has significantly higher performance in speech recognition across languages but not in the recognition of signals that lack speech-like structure, such as urban sounds. Our results suggest a central computational role of brain rhythms in speech processing.