Student Theses and Dissertations

Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

RU Laboratory

Freiwald Laboratory


Facial expressions constitute a fundamental class of social communication signals in primates. Nonverbal expressions shared between individuals can instantly and effortlessly communicate endogenous affective state, the valence of external stimuli, as well as more complex concepts such as intentions, social rank, and receptivity to future social interactions. All of this information is transmitted through the fine, dynamic arrangement of dozens of small facial expression muscles. Commiserate with their importance, the muscles of facial expression are under direct cortical control from descending projections to the facial nucleus. These direct corticomotoneuron projections are unique to primates and originate in multiple distinct cortical face representations, including primary (M1), ventral premotor (PMv), and cingulate motor (M3) cortices. Much like their human counterparts, macaques use facial expressions to navigate both cooperative and competitive interactions with conspecifics. Yet despite their integral role in social communication, the neural bases for facial expression production are not well understood. Neither the relative contributions of each of these areas, nor the local codes supporting facial expression production, have been systematically explored. To this end, we first identified these three cortical face motor representations via fMRI, before targeting them for electrophysiological recording with chronic electrode arrays. Subsequently, we recorded from many single cells in each region simultaneously, while subjects produced naturalistic facial expressions. These expressions included socially-meaningful expressions elicited through a number of visuosocial stimuli, in addition to non-socially meaningful, volitional facial movements, as well as periods of rest. We found widespread electrophysiological activation of single cells in all cortical face motor regions during naturalistic facial expression production, during both volitional and socioemotional facial expressions. While individual neuron responses were highly heterogenous in terms of tuning, latency, and shape, neural populations within each cortical region supported linearly discriminable representations of expression type. Thus, all cortical face motor regions contained unique representations for facial expressions. However, this was achieved via regionally distinct coding regimes, with primary motor cortex enacting two temporally-distinct population codes, and premotor cortex, a single temporally-stable code. Furthermore, the dynamic structure of neural activities was not preserved across expressions, and this was not due to non-overlapping subpopulations of neurons, each specialized for one expression type. Instead, each expression was accompanied by a unique temporal evolution of population activity, formed by neurons which flexibly recombined their activities. At the level of the entire cortical face motor system, each expression engaged a unique combination of neural modes, which themselves contained a mixture of temporal and expression-specific features. Within motor cortex, expressions with similar kinematic features engaged a subset of shared neural modes, perhaps reflecting the direction of moment-by-moment movements -- a feature not observed in the dynamics of premotor or cingulate cortex. Indeed, we found that primary motor cortex contained substantial information regarding the continuous kinematics of the face during naturalistic facial expression production. Together, these results reveal the first cortical control mechanisms for naturalistic facial expression production in primates and generate new hypotheses for testing how motor programs are instantiated in a distributed system, which interfaces with various domains of social cognition.


A Thesis Presented to the Faculty of The Rockefeller University in Partial Fulfillment of the Requirements for the degree of Doctor of Philosophy

Available for download on Monday, March 24, 2025

Included in

Life Sciences Commons