Year: 1987
Type: conference-abstract
Abstract: Tags frequently occur in English text. Common examples include “‘Are you coming?’ he said,” where “he said” is an attributive tag, and “I bought a new car, Manny,” where “Manny” is a vocative tag. Because of the frequency of tags, it is important to have an adequate treatment of their intonation for more natural sounding text-to-speech synthesis. This paper examines the intonational properties of tags in American English and discusses their phonological characterization. Investigated instrumentally are the F0 patterns employed by speakers producing tags in question and statement contexts (“‘Are you hungry?’ he said” versus “‘I am hungry,’ he said”). A phonological analysis is proposed for such tags building on work by Pierrehumbert [“The phonology and phonetics of English intonation,” Ph.D. dissertation, MIT (1980)] and Beckman and Pierrehumbert [“Intonational structure in Japanese and English,” Phonology Yearbook 3, 255–309 (1986)]. Finally the incorporation of these results into the text-to-speech system under development at Bell Laboratories is discussed, and the resulting improvement in the quality of synthesis is demonstrated.
DOI: 10.1121/1.2024607
Authors: Richard W. Sproat, Shirley A. Steele
Institution: AT&T (United States)
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W2039651103