Episodes of facial displays involving the zygomatic action (AU12: lip corner pulling or smiling) were selected from a large sample of children (n = 95) exposed to pleasant and unpleasant odours in the presence of an unfamiliar person in order to investigate potential differences in morphological, temporal patterning and social signal value of smiling.
In a first experiment, using the facial action coding system (FACS: Ekman & Friesen 1978), a considerable morphological flexibility of smiles was observed in relation to the subjects' hedonic experience. The facial configurations of smiling were formed by a number of actions in the upper (AU 4: brow lowering), middle (AU 9: nose wrinkling) and lower face (AU 14: dimpling, AU 15: lip comer depressing, AU 17: chin raising, AU 23: lip tightening) and the mouth was more often ‘closed’ in response to unpleasant odours. When exposed to pleasant odours, zygomatic action co-occurred more frequently with an opening of the mouth (AUs 25, 26, 27) or with a raising of the cheeks (AU 6). An analysis of the temporal patterning of zygomatic actions showed that they occurred more rapidly, dropped off the face less abruptly with a stepped decrease, were less smooth, and were often associated with shorter gazes directed toward the examiner in response only to unpleasant odours. These findings suggested that a number of subjects might exert some control on their smiling while confronted with a presumed social constraint, namely the smelling of unpleasant odours in the presence of an unfamiliar person.
In a second experiment, the communicative value of smiling was investigated in a real-time projection of 10 variants of smiling to a panel of receivers (n = 52). The Duchenne smile (AU 6 + 12 + 25) and smile with lips opening (12 + 25) provided more accurate information about the hedonic valence of the inhaled odour than did the other types of smiling. In contrast, the perceived valence of the facial displays simultaneously combining zygomatic action with muscular actions of the lower face (AUs 15, 17, 23) appeared more difficult to discriminate by untrained receivers. It was hypothesized that the senders displayed some forms of smiling possibly to mask their responsiveness to unpleasant odours in signalling ambiguous or incorrect information about their internal state to a recipient.