Single-neuronal elements of speech production in humans

Abstract

Humans are capable of generating extraordinarily diverse articulatory movement combinations in order to produce meaningful speech. This ability to orchestrate specific phonetic sequences, their syllabification and inflection over sub-second timescales allows us to produce thousands of word-sounds and is a core component of language 1,2. The basic cellular units and constructs by which we plan and produce words during speech, however, remain largely unknown. Here, using acute ultrahigh density Neuropixels recordings in humans, we discover neurons in the language-dominant prefrontal cortex that encoded detailed information about the phonetic arrangement and composition of planned words during the production of natural speech. These neurons represented the specific order and structure of articulatory events prior to utterance and reflected the segmentation of phonetic sequences into distinct syllables. They also reliably predicted the phonetic, syllabic and morphological components of upcoming words and displayed a temporally ordered dynamic. Taken collectively, we show how these cells were spatially organized and how their activity patterns transitioned from articulation planning to production in real-time. We also demonstrate how they tracked the composition of phonemes during perception, and how they distinguished processes specifically related to speaking from listening. Together, these findings reveal a remarkably structured organization and encoding cascade of phonetic representations by prefrontal neurons in humans and a cellular process that can support the production of natural speech.