Abstract
This paper outlines a framework for the description of English intonation, and then reports on a study of “high falling” and “low rising” nuclear tones, based on measurements on a corpus of 150 utterances produced by three speakers. It is found (i) that under the conditions of the experiment, the various tokens of a given intonation type as produced by a given talker cluster closely around a single pattern; and (ii) that speakers differ one from another in consistent ways, suggesting that adequate normalisation procedures can be found. Synthetic speech approximations to the patterns are discussed.
Get full access to this article
View all access options for this article.
