AVSP 2003 - International Conference on Audio-Visual Speech Processing

September 4-7, 2003
St. Jorioz, France

Auditory-Visual Speech Perception Development in Japanese and English Speakers

Kaoru Sekiyama (1), Denis Burnham (2), Helen Tam (2), Dogu Erdener (2)

(1) Future University Hakodate, Japan
(2) University of Western Sydney, Australia

Development of auditory-visual speech perception was investigated in a cross-linguistic developmental framework, using the McGurk effect. Stimuli consisting of /ba/, /da/, and /ga/ utterances were presented to participants who were asked to make syllable identifications on audiovisual (congruent and discrepant), audio-only, and video-only presentations at various signal-to-noise levels. The results of Experiment 1 with 24 adult native speakers of English and 24 of Japanese supported previous reports of a weaker visual influence for Japanese participants. Experiment 2 was a short version of Experiment1 in which 16 Japanese and 14 English language 6-year-olds, and new groups of 24 Japanese and 24 English adults were tested. The results showed that the degree of visual influence was low but statistically equivalent for Japanese and English language 6-year-olds, and that there was a significant increase in visual influence over age for the English but not the Japanese language groups. Nevertheless, both the Japanese and English language groups showed an increase in speechreading performance (in the visual-only condition). It appears that the developmental increase of speechreading performance is related to the increase of the size of the visual influence in the English language participants, whereas such a straightforward relationship is not the case for the Japanese participants.


Full Paper

Bibliographic reference.  Sekiyama, Kaoru / Burnham, Denis / Tam, Helen / Erdener, Dogu (2003): "Auditory-visual speech perception development in Japanese and English speakers", In AVSP 2003, 43-47.