Crossing Gender Borders: Bidirectional Dynamic Interaction Between Face-Based and Voice-Based Gender Categorization

The processing of voices and faces is known to interact, for example, when recognizing other persons. However, few studies focus on both directions of this interaction, including the influence of incongruent visual stimulation on voice perception. In the present study, we implemented an interference paradigm involving 1152 videos of faces with either gender-congruent or gender-incongruent voices. Participants were asked to categorize the gender of either the face or the voice via key press. Task (face-based vs. voice-based gender categorization task) was manipulated both block-wise (relatively low executive control demands) and in a mixed block (relatively high executive control demands due to trial-by-trial task switches). We aimed at testing whether and how gender-incongruent stimuli negatively affected gender categorization speed and accuracy. The results indicate significant congruency effects in both directions – gender-incongruent visual information negatively affected voice categorization time and errors, and gender-incongruent voices affected visual face categorization. However, the former effect was stronger, supporting theories postulating visual dominance in face-voice integration. Congruency effects, which were not significantly reduced over the course of the experiment, were larger under high executive control demands (task switches), suggesting the availability of fewer attentional resources for incongruency resolution. Overall, voices generally appear to be processed in conjunction with facial information, which yields enhanced processing for more authentic voices, that is, voices that do not violate face-based expectancies. The data strengthen theories of face-voice processing emphasizing strong interaction between both processing channels.