How Does the Brain Predict Who's Speaking?

While sitting in a noisy environment you may have trouble understanding your conversation partner. However, listening to a familiar voice of a friend may be easier compared to listening to an unfamiliar voice. Past research studies support this phenomenon with evidence of stronger speech-evoked brain activity while listening to a familiar speaker. Additionally, previous studies have developed the concept of the Predictive Coding Theory. This theory states that the brain predicts what will occur in its sensory environment based on its internal representations of the world. The incoming sensory inputs update the predictions. There has been an emphasis on the physiological evidence of neurocognitive processing in this theory. One in which is dominated by top-down processing, and the other uses sensory sampling. In this study, we claim that gamma band oscillations use sensory sampling and beta band oscillations use top-down processing and predicting. However, there is no evidence on what neural substrates are involved while tracking a speaker’s identity in noise. Our goals were to investigate the cortical oscillations produced and their location in the brain when there is a speaker identity cue versus no cue. We measured cortical EEG data of 13 normal-hearing participants in speech-in-noise. The speaker identity cues increased beta band oscillations in the inferior frontal gyrus region of the left hemisphere. However, without a speaker identity cue, greater gamma band oscillations were found in the supratemporal gyrus in both hemispheres. The results had significant differences amongst both conditions. With these results, we support the Predictive Coding Theory and the physiological evidence of cortical oscillations. While the brain tracks a familiar voice in noisy environments it uses predictive top-down processing indicated by beta waves. However, if the brain cannot predict the speaker, then sampling of auditory features occurs, indicated through gamma oscillations. This allows the brain to be open to all potential voices.