This chapter presents a dynamical systems perspective on both facial expression specifically and emotion communication more generally. I have been developing this perspective in response to several interesting problems that have recently emerged in attempts to verify one currently popular theory of infant emotional expression. In this chapter, I will review these problems and discuss the limitations of extant theory. The dynamical systems view I will present is not a fully articulated alternative proposal. Instead, it is intended to represent a new direction for theoretical and empirical exploration in which solutions to the problems with the current theory may be found.

A Theory of Infant Emotional Expression

The currently popular view of infant emotional expression is most fully embodied in Izard's (1977, 1991) differential emotions theory. According to this theory, there is a species-specific set of human emotions that emerge during development according to a maturational timetable. In its original formulation (Izard, 1971; Izard and Malatesta, 1987; Izard et al., 1995), the theory proposes an innate concordance between infant emotions and a specified set of infant facial expressions (Izard, Dougherty, and Hembree, 1983). These expressions are direct readouts of their corresponding emotions, that is, they are automatically produced when the emotion is experienced and are not produced in other circumstances. Thus, facial expressions serve as veridical indices of infant affects and can be simply “read” by observers to determine the infant's emotional state.