The Paradox of Emotion Attribution

Abstract

People who explain why ambiguous faces are expressing anger perceive and remember those faces as angrier than do people who explain why the same faces are expressing sadness. This phenomenon may be explained by a two-stage process in which language decomposes a facial configuration into its component features, which are then reintegrated with emotion categories available in the emotion explanation itself. This configural-decomposition hypothesis is consistent with experimental results showing that the explanation effect is attenuated when configural face processing is impaired (e.g., when the faces are inverted). Ironically, although people explain emotional expressions to make more accurate attributions, the process of explanation itself can decrease accuracy by leading to perceptual assimilation of the expressions to the emotions being explained.