Perceiving Talking Faces

Dominic W. Massaro, UC Santa Cruz

4:00 pm Thu. Sept. 12 in 311 Psychology

Speech perception has been studied extensively in the last decades. We have learned that people use many sources of information in perceiving and understanding speech. This talk focuses on the important contribution of visible information given in the talker's face in face-to-face communication. Although the influence of visible speech is substantial when auditory speech is degraded, visible speech also contributes to performance even when paired with intelligible speech sounds. Our studies use a synthetic talking face to achieve control over the visible speech and to study those visible aspects that are informative. Our talking head can be heard, communicates paralinguistic as well as linguistic information, and is controlled by a text-to-speech system. Using this talking head, auditory and visible speech are manipulated independently of one another to uncover the fundamental processes involved in speech perception by ear and eye. Perceivers judge audible, visible, and bimodal speech under a variety of experimental conditions. The results from a plethora of studies are consistent with the Fuzzy Logical Model of Perception (FLMP). This model assumes three successive but overlapping stages of processing: evaluation, integration, and decision. At evaluation, each source of information provides independent and continuous support for linguistic alternatives. The multiple sources are combined at the integration stage, and an assessment is made concerning the relative goodness of match of the relevant alternatives. This assessment is then mapped into some type of behavior such as a categorization or a rating judgment, using a straightforward decision rule. The model makes the strong prediction that the contribution of one source of information is inversely related to its ambiguity. The FLMP outperforms other models that assume categorical perception, selective processing of one modality, auditory dominance, and averaging of the sources of information. Finally, it is proposed that the FLMP describes a universal law of behavior. Demonstrations of the talking head and various psychological phenomena will be will be provided.