The Neural Coding of Speech Across Human Languages

Experimental approaches with significantly greater spatial and temporal resolution are necessary to directly resolve, both local- and population- level, the contrastive encoding of speech sounds. This study proposes an innovative methodological approach using customized intracranial high-density electrode arrays to record neural activity directly from nonprimary auditory cortex in patients undergoing clinical neurosurgical procedures (acute intraoperative and chronic extraoperative). This approach overcomes obstacles in traditional neuroimaging by offering high signal-to-noise recordings, unprecedented detailed spatiotemporal resolution, and a large number of simultaneously recorded cortical sites in awake, behaving subjects. The research study team will leverage the diversity of languages spoken by patients that are treated at the large volume epilepsy and tumor brain mapping programs at the University of California, San Francisco. They will examine cortical responses to speech stimuli (natural speech corpora and control tokens) in Spanish, Mandarin, and English speakers (monolingual and bilingual). They will also focus on encoding models in three fundamental domains of acoustic-phonetic cues that are present in all languages: pitch, formants, and amplitude envelope. The aims of this study seeks to determine how pitch cues encode lexical tone processing in Mandarin (Aim 1), the cortical representation of vowels in Spanish and English (Aim 2), and the encoding of the speech amplitude envelope in Spanish and English (Aim 3). Together, these aims will elucidate mechanistic principles of speech encoding in the human auditory cortex to understand what is shared and different across human spoken languages. Abnormalities in these fundamental processes have been implicated in a host of communication disorders such as dyslexia, developmental language disorder, central hearing loss, and aphasia. These results should heavily impact current theories of speech processing and, therefore, will have significant implications for understanding and remediating human disorders across different languages.