With National Science Foundation support, Dr. Elisabeth Selkirk and her colleagues will conduct three years of linguistic research on English, Japanese, and Tashlhit Berber. The project will investigate relations between prosody and semantically defined types of Focus. Prosodic variables include the presence and distribution of pitch accents, phonological phrasing, and relative pitch range. Semantic variables include whether a sentence involves a contrastive correction, a counter-assertion, or presentation of new information. Materials manipulating semantic variables will be read by actors and recorded. Computer-extracted pitch contours in the recorded materials will then be analyzed to establish correlations between semantic and prosodic properties. Listening experiments will test whether the prosodic contrasts that speakers produced as reflexes of different Focus types influence sentence comprehension. The cross-linguistic comparison tests for points of similarity in languages that differ in word and sentence prosody. While English makes no use of tones to differentiate lexical items, it does exploit tonal accents at the sentence level to mark Focus types. By contrast, Japanese makes no use of sentence level accents to mark Focus types, but it does differentiate lexical items with tone. Berber, unlike English and Japanese, indicates the contrastive status of a constituent with a special syntactic construction. The pitch contours of sentences with this construction will be compared to sentences containing other Focus types to see whether any special prosody occur when Focus is already indicated by other means. What's at issue is whether languages avoid redundancy between different systems for expressing the same semantic properties.
This cross-linguistic study of prosody and semantic Focus is significant for several reasons. First is the question of universal properties of language. The research asks whether an abstract phonology of Focus is common across these different languages (e.g., a system of phrasal stress prominence). Second, better understanding of how prosodic properties of a sentence influence its semantic interpretation or its appropriateness in a specific discourse would permit a deeper understanding of the interface between phonology and other domains of linguistic competence. In the practical realm, better descriptive generalizations concerning semantically or pragmatically relevant prosody would advance computer-based treatments of spoken language, uses of natural speech in research on human sentence processing, and the assessment of language impairments. The grant will also advance international collaboration, through cooperation with Dr. Abdelkrim Jebbour, a Berber scholar from Morocco. It will also contribute to the training of graduate students in the analysis of prosody and of syntactic, semantic, phonological, and phonetic properties that are relevant to prosody.
