Do proposed facial expressions of contempt, shame, embarrassment, and compassion communicate the predicted emotion?

Abstract

Shame, embarrassment, compassion, and contempt have been considered candidates for the status of basic emotions on the grounds that each has a recognisable facial expression. In two studies (N=88, N=60) on recognition of these four facial expressions, observers showed moderate agreement on the predicted emotion when assessed with forced choice (58%; 42%), but low agreement when assessed with free labelling (18%; 16%). Thus, even though some observers endorsed the predicted emotion when it was presented in a list, over 80% spontaneously interpreted these faces in a way other than the predicted emotion.

Keywords:

• Facial expressions
• Free labeling
• Forced choice
• Response method
• Emotion

Acknowledgements

This study was funded by a grant from the National Science Foundation (0421702).

We thank Maria-Paz Rodriguez, Stephanie Hadley, Jessica Terwilliger, Jeehye Choi, Kerrie Pieloch, and Matthew Williams for their help with data collection and coding.

Notes

¹In pilot testing, several subjects indicated that they were unfamiliar with the meaning of the word contempt and, therefore, failed to choose it in the forced-choice task. In order to alleviate this problem, “scorn” was included in the forced choice list as a synonym for contempt. Both scorn and contempt were scored correct for the contempt face.

²In Haidt and Keltner's (1999) free-labelling method, participants were asked to explain what “happened to make the person feel this way” (p. 236) but were not required to label the emotion. Only after data collection did the authors find that a large proportion of the participants spontaneously provided an emotion label for the faces.