Audiovisual perception of congruent and incongruent spanish vowels in unimodal and bimodal conditions

This paper proposes a preliminary approach to the study of the integration of auditory and visual signals in the process of audiovisual speech perception of Spanish vowels. Following the experiment of McGurk and Macdonald (1976), the purpose of this study is to find out whether incongruent cues in audiovisual speech perception can alter the identification of the auditory stimulus, and to analyze whether this bimodal integration process causes a perceptual result other than the one that occurs in each channel separately. These results will allow us to know the extent to which the visual signal affects this class of stimuli and conditions, and determine, on the other hand, if the findings reflect differences in sensibilities in regards to the visual channel depending upon the speaker’s gender. For this purpose, 28 subjects (12 men and 16 women) had to identify the five vowels of Spanish distributed at random into three blocks: a) 25 audiovisual cross combinations between all vowels (thus in congruent and incongruent audiovisual conditions); b) 10 visual stimuli distributed in two series (5 x 2); and c) 5 auditory stimuli. Along with this, the participants had to indicate the degree of confidence of their response. Results showed, among other findings, that the presence in this kind of incongruent visual stimuli affects the perception of the auditory signal and results in perceptual fusions. It also demonstrates that visual information alone is not sufficient for discrimination, although important differences can be found depending on the presence of visual cues, such as rounding or vowel openness. There are also significant differences in visual perception and the variability of responses between men and women.