Digitizing the moving face during dynamic displays of emotion

Humans typically decode facial signals during dynamic interactions in which the face moves. In this study, we digitized real time video signals in ord er to examine movement asymmetries across the face during emotional and non emotional expressions. Forty dextral males were tested. For each expression . a 400 ms video segment was analyzed for changes in signal value (pixel in tensity) over consecutive frames. Thr upper and lower face regions were exa mined separately due to differences in the cortical enervation of facial mu scles in the upper (bilateral) vs lower face (contralateral). Results revea led distinctly different movement asymmetries over the lower and upper hemi face. In the upper face, more movement occurred over the right side for mos t facial expressions, regardless of emotionality. The latter finding questi ons the assumption that muscles of the upper face are symmetrical and/or bi laterally enervated in a symmetrical manner. In the lower face, negative ex pressions linked to fight-flight emotions (i.e. fear. anger) were associate d with greater left sided movement, whereas happiness tended to be associat ed with more right sided movement. No consistent pattern of movement asymme try occurred for nonemotional expressions. Although the valence-related mov ement asymmetries in the lower face are consistent with neuropsychologial m odels of emotional expressivity, it remains unclear whether they reflect ac tivation or inhibitory hemispheric mechanisms. Taken together, these data s uggest that multiple factors may contribute to expressive movement asymmetr ies of the face. (C) 2000 Elsevier science Ltd. All rights reserved.