Can English perceivers match Cantonese auditory and visual prosody?

The prosody of an utterance can be varied by changing F0, duration and amplitude. Such changes are typically accompanied by variation in the talker’s face/head motion (visual prosody). For native language utterances, people can match auditory and visual prosody accurately. We tested whether English perceivers can do this with an unfamiliar language, Cantonese, which differs from English specifically with regard to suprasegmental properties (e.g., different rhythm type; use of lexical tone). These differences may make extraction of prosody difficult, because they distract English perceivers and/or because they affect the way prosody is realized. However, AV cues for prosody may be similar across languages and sufficiently salient to overcome the suprasegmental differences. We tested native AustralianEnglish participants (N=27) with 50 Cantonese sentences spoken as questions, narrowly focused or broad focused utterances by two native Cantonese talkers. Participants completed a same-different matching task for auditoryauditory (AA); visual-visual (VV) and auditory-visual (AV) pairs. Each pair type consisted of the same sentence and talker, but different tokens. Matching performance was above chance for all conditions: AA > AV = VV. Results are discussed in terms of how auditory and visual prosody is conveyed and how this may be affected by language properties.