Form and function in facial expression of emotion
Adam K. Anderson
University of Toronto

Facial expressions are instrumental in externalizing one’s internal emotional state and thus in regulating social interactions. However, facial expressions not only “express” emotions but may also have adaptive functions for the sender beyond their established communicative value. Over 130 years ago, Darwin hypothesized that emotional expressions originated in a less appreciated functional role, to modify preparedness for perception and action. We examined whether the origin of facial expressions associated with different emotional states may serve to alter the function of sensory systems whose receptors reside on the face in the service of regulating sensory input. We present evidence for this thesis through a convergence of methods, including visual statistical modeling of expression appearance, visual and olfactory psychophysics, eye movement recording, and structural MRI. Our results provide evidence for two of Darwin’s evolutionarily derived principles of emotional expressive behavior: 1) the principle of “serviceable associated habits” or “function”—whereby specific facial expressions originate in patterns of movement serving adaptive functions for the expresser, and 2) the principle of “antithesis” or “form”—whereby emotions with opposite functions are opposites in physical configuration. These convergent sources of evidence demonstrate that facial expressions are not arbitrarily shaped social signals, but may have differentiated from an underlying origin as sensory adaptations.

Our second thesis is despite these origins, facial expressions have likely been co-opted, maintained, and further shaped for their more primary role as social signals. Evidence for this exaptation or co-opting of an original sensory opposition for the purposes of social communication comes from two sources. First, using psychophysical scaling and face adaptation procedures, we show how oppositions in facial expression structure are mirrored in the representations of facial expression space supporting expression recognition. Specifically, we suggest facial expression recognition rests upon high-level opponent coding of two fundamental dimensions underlying expression space, similar to low-level opponent coding found in color perception. Second, in a case study of the exaptation of the sensory regulatory function of facial expressions toward social communication, we show how expressions of social and moral disgust in response to unfair treatment might originate from a primitive facial response to the rejection of bad tastes. This provides evidence for the oral origins of moral disgust. In sum, through examination of the putative sensory origins and present day communicative functions of emotions this research program demonstrates new insights into the architecture of affective representations.