This paper presents the recording paradigm and the perceptual evaluation of a corpus of 16 prosodic social affects performed by a set of 8 native American English speakers (5f, 3m). The social affects are defined according to given communication goals in predefined social contexts, such as varying the relative hierarchical relation between the speaker and the interlocutor. The prosodic and facial strategies are evaluated by native listeners, rating their performance for achieving the targeted communication goal. Variations in the prosodic and facial strategies observed are then described and discussed in light of Ohala's frequency code. By selecting the best performances, 15 social affects were analyzed. Given the dimension of dominance as a main aspect related to the observed pitch level, the complexity of expressions is reflected in the multiparametric nature of the prosody. Voicing strength seems to be an important part of the acoustic information. In addition, the visual expressions allow an efficient interpretation of prosodic communication goals. Individual strategies to perform these social affects are observed in the prosodic variations, and may be related to factors such as the extraversion of speakers, their gender, and intrinsic pitch.
Bibliographic reference. Rilliard, Albert / Erickson, Donna / Shochi, Takaaki / Moraes, João Antônio de (2013): "Social face to face communication — American English attitudinal prosody", In INTERSPEECH-2013, 1648-1652.