Impacts of aging on suprasegmental and segmental encoding of vocally-expressed confidence in Wuxi dialect

While literature demonstrated that speakers encoded high vs. low confidence in prosodic cues and that older and younger adults differed in speech-related cues, it is unknown how speaker confidence is encoded at phonetic level and how aging could influence the vocally-expressed confidence. This study aimed to explore the effects of aging (by comparing middleaged vs. elderly speakers) on prosodic and phonetic features in encoding vocally-expressed confidence in Wuxi dialect. Vowels were spoken in confident, unconfident and neutral tone of voice by two age groups of native Wuxi dialect speakers using a standard elicitation procedure and were rated for their perceived confidence by an independent group of native listeners. The results showed that vowels perceived to be confident were associated with larger F1, F2, mean f0, mean intensity and intensity range and smaller f0 range. Elderly speakers raised their F1 and range of f0 when speaking. Furthermore, age affected patterns of vocally-expressed confidence in all formant and prosodic (f0 and intensity) features. These findings demonstrated important mechanisms on (1) how Wuxi-dialect speakers encoded confidence-related information by altering pattems of formants, fundamental frequency and intensity in speech vowels; and (2) how middleaged and elderly speakers could differ in using prosodic and segmental cues to encode different levels of confidence.