Generalized discrimination of positive and negative facial expressions by 7- and 10-month-old infants

Pamela Michel Ludemann, Purdue University

Abstract

A series of experiments was conducted in which the contributions of expression features vs. familiarity to early categorical recognition of emotion was examined. Using an infant-controlled habituation procedure, 7- and 10-month-olds were tested on their ability to generalize their discrimination of facial expressions across five models. The expressions posed varied in terms of their hedonic value (positive or negative) and category fidelity (prototypic or blended expression). In Experiments 1, 2, and 3, the prototypic expressions happy, surprise, fear, and anger were used in the stimulus sets. In Experiment 4, blended expressions, rated as depicting positive or negative emotions, were presented to the infants. Finally, in Experiments 5 and 6, responses to the prototypic and blended expressions were contrasted. Overall, these experiments showed that 7-month-old infants categorically recognized only happy expressions. In contrast, 10-month-olds demonstrated generalized discrimination of anger and, possibly, happy expressions and were able to discriminate these expressions from fear and surprise, respectively. Further, these older infants appeared to recognize the underlying hedonic value shared by the prototype and blended expressions. At both ages, infants failed to respond consistently to surprise and fear expressions. They also failed to include hedonically similar expression blends in the expression categories formed following habituation to prototypic expressions. The pattern of results suggests that the infant's ability to categorize facial expressions may be broadly influenced by familiarity. That is, positive expressions may be more readily categorized than negative ones and this difference may be mediated by experience. Through playful interaction, positive facial expressions (e.g., happy) may become meaningful to young infants. Similarly, for older infants, gains in self-locomotor skills may increase the likelihood of their exposure to negative expressions (e.g., anger). This, in turn, may increase the familiarity of negative displays. However, within a memory task, the presence of expression-specific features seems to determine the inclusiveness of the categories formed. Narrowly-defined stimuli sets (e.g., prototypic expressions) appear to promote the formation of clear-cut category boundaries. In contrast, less-well defined stimulus sets (e.g., blended expressions) seem to encourage boundary flexibility.

Degree

Ph.D.

Advisors

Nelson, Purdue University.

Subject Area

Developmental psychology

COinS