Teaching tone and intonation with the Prosody Workstation using schematic versus veridical contours

Prosodic features of speech (e.g., intonation and rhythm) are often challenging for adults to learn. Most computerized teaching tools, developed to help learners mimic model prosodic patterns, display lines representing the veridical (actual) acoustic fundamental frequency and intensity of the model speech. However, a veridical display may not be optimal for this task. Instead, stereotypical representations (e.g., simplified level or slanting lines) may help by reducing the amount of potentially distracting information. The Prosody Workstation (PW) permits the prosodic contours of both models and users’ responses to be displayed using either veridical or stereotypical contours. Users are informed by both visual displays and scores representing the degree of match of their utterance to the model. American English-speaking undergraduates are being studied learning the tone contours and rhythm of Chinese and Hausa utterances ranging in length from two to six syllables. Data include (a) accuracy of mimicking of the models’ prosodic contours, measured by the PW; (b) quality of tonal and rhythmic production, judged by native speaker listeners; and (c) learners’ perceptions of the ease of the task, measured by a questionnaire at the end of each session.