PhonDi Discussion Group

Ian Calloway: "Power Priming and Speech Perception"

Abstract
English listeners use information about the speaker's gender in categorizing sibilants as /s/ or /ʃ/. This study investigates whether self-perceived power influences how listeners process this social information, and, in turn, how listeners categorize sibilants. Self-perceived power, the perceived capacity to control the resources or punishments of others, is associated with the style of processing of social information an individual tends to adopt. The percept of high-power listeners tend to depend more on attributes associated with a social category, rather than attributes unique to the speaker, which may conflict with the category assigned to the speaker (Fiske & Neuberg 1990). I expected that when a listener is presented with conflicting information about speaker gender (e.g. a female face but a male voice), the percept of the low-power listener will be sensitive to both sources of information, while that of the high-power listener will be sensitive to just one.

In this study, Participants were primed for a high or low degree of self-perceived power (as in Galinsky et al. 2003) and completed a forced choice identification task. During each trial, they were presented with an image of a face and manipulated auditory stimuli, which ranged from "sigh" to "shy"; the participants indicated hearing "sigh" or "shy"'. For a given participant, the gender of the face was paired with either the same or the opposite gender voice for the duration of the experiment. Responses were significantly influenced by the gender of the speaker's voice, whether the gender of the face and that of the voice matched, and the power priming group to which the participant was assigned; in most instances the direction of the influence matched predictions. I also frame these results with respect to other studies on the relationship between power and linguistic perception, research on individual differences, and research on language change.