Cross-language and individual differences in the production and perception of syllabic prominence. Rhythm-typology revisited.

点击翻译按钮获取中文摘要

The consequences of cross-language and inter-speaker phonetic differences in the prosodic structuring of words and utterances for the realization of word stress and ¿information¿ accents (topic and focus) will be analyzed in the controlled production of contextualised sentences and their reiterant equivalents. A number of languages, traditionally considered to be stress-timed (e.g. German, English, Russian, Bulgarian), will be compared to syllable-timed languages (e.g., French, Spanish, Italian) and languages which are not so readily categorized as either (e.g., Portuguese, Czech) as well as to Japanese (as the representative of moratimed languages). Structural differences: a) differences in the variability of syllable complexity, b) the use or not of distinctive vowel length, c) the phonological status and d) type of spectral ¿reduction¿ in unstressed vowels, e) the type of stress and accent system, it is hypothesized, constrain the exploitation of duration, F0, ¿loudness¿ and spectral distinctness in producing prominent syllables. Anatomical and gestural-pattern variation (e.g., vocal tract shape, degree of mouth opening) between individuals may also systematically affect the weighting of these parameters. Whether the perception of prominence differs correspondingly is not known. Trading relations between parameter weights, either within- or cross-language could maintain comparable prominence percepts. Signal-manipulation experiments will be employed to examine and quantify the relative perceptual weighting of the prominence-giving parameters across and within languagegroups. The results will have implications for theories of universality, rhythm typology, metrical phonology and, more practically, for interference theories in second-language learning and for speech synthesis.