Neural coding of language semantics and speaker identity during speech comprehension

Project Abstract Effective communication requires understanding both the meaning of spoken words (semantics) and the identity of the speaker. This integration of semantics and speaker identity is vital for social interactions, yet the neural mechanisms supporting this process are not fully understood. This project investigates how the anterior cingulate cortex (ACC), a region involved with both language and social cognition, contributes to the integration of semantic content and speaker identity during natural conversations and speech comprehension. We propose that the ACC encodes semantics and speaker identity using distinct but related neural population subspaces, allowing efficient generalization across different speakers while maintaining specificity. To test this hypothesis, we will record single-neuron activity in the ACC of patients with epilepsy during both passive listening to podcast stories and active participation in natural conversations. Our specific aims are to delineate how ACC neurons respond to different semantic categories and speakers (Aim 1) and to determine how neural populations align these semantic representations across different speakers using neural subspace analysis (Aim 2). Our findings will have important implications for basic science research and understanding of communication disorders. This study will provide new insights into the neural coding of language and social information, improving the development of assistive communication technologies and neural prosthetics aimed at individuals with language impairments, such as those with autism or aphasia. Collectively, this proposal aims to elucidate single neuron dynamics in real-life communication and will provide me with the skills needed to continue social communication and language neuroscience research as a principal investigator.