Infants’ facial input from 6 to 17 months rarely includes hypothesized expressive emotion prototypes

The facial configurations that infants encounter in typical interactions with caregivers and family members form a basis for their understanding of emotion. Studies to date have described or quantified these natural exposures to facial information, but descriptions of the emotional content of that input are still needed. We accomplished this by leveraging an existing, longitudinal corpus of video-recordings of 6- to 17-month-old infants’ interactions with close others. We found that infants were generally exposed to far more partial than full faces, and that exposure to full faces, while low, increased over the time periods observed. Exposure to facial configurations hypothesized to serve as expressive prototypes for categories of emotion (e.g., happiness, anger ) was extremely rare. When hypothesized expressive prototypes did occur, it was often in unexpected contexts, suggesting that they did not necessarily serve as a cue to the presumed emotional meaning. Lastly, descriptive analyses of the facial movement data revealed that infants’ input was dominated by a few, mostly simple facial configurations that did not correspond fully or uniquely to hypothesized category prototypes. These results extend evidence that emotions are expressed heterogeneously in daily life, such that infants must learn much more than one-to-one mappings between facial configurations and context/behavior to acquire expressive vocabularies for emotion.