Dating back to Darwin’s original treatise on emotions in humans and in animals (Darwin, 1872/1997), theories of emotion—from classic to contemporary—are built on the premise that emotions, and our expression and perception of them, serve functional purposes. Testing these theories, however, has relied almost exclusively on verbal labeling tasks (again dating back over 150 years to Darwin and extending into the present). Of particular relevance to the current review is that functional theories of emotion expression perception contend that expressions inform specific types of perceiver behaviors and outcomes. These perceiver behaviors and outcomes are central to any functional theory. Yet these same theories rarely measure adaptive behaviors or other outcomes in the perceivers. For reasons expounded upon in this paper, we propose that this approach limits the conclusions that can be drawn. Even in paradigms not employing verbal labeling, facial expression stimuli are generated through verbal labeling methodology. This results in a set of expressions necessarily reflecting lay conceptualizations of emotion expressions, which then drives the science of facial expression perception. This approach is limiting as lay conceptualizations are subject to change over time and thus make it prone to missing the immutable processes and mechanisms underlying expression perception that are central to our science. Finally, we offer some initial recommendations for how we, as researchers in the domain of facial expression perception, can begin to move beyond the use of verbal labeling in our stimulus generation and research paradigms.
Adams et al. (Mon,) studied this question.