Sequential contrast versus compensation for coarticulation in Japanese versus English

• Published in: The Journal of the Acoustical Society of America Vol. 119; no. 5_Supplement; p. 3423
• Main Authors: Mash, Daniel, Kawahara, Shigeto
• Format: Journal Article
• Language: English
• Published: 01-05-2006
• ISSN: 0001-4966; 1520-8524
• DOI: 10.1121/1.4786854

English listeners identify a stop ambiguous between /t/ and /k/ more often as ‘‘k’’ after [s] than after [sh] (Mann and Repp, J. Acous. Soc. Am. 69, 548–558, 1981). This bias could reflect compensation for coarticulation, which undoes the fronting of a stop expected next to [s], or instead sequential contrast, in which an intermediate F3 value sounds lower and more ‘‘k’’-like after [s]’s high-frequency energy concentration. Japanese may distinguish these accounts’ predictions. The Japanese sibilant fricative is pronounced [s] before [u] but [sh] before [i]. These vowels delete between two voiceless consonants, so [s] or [sh] can occur before [t] or [k] in surface forms. If Japanese listeners treat [s] followed by a stop from a [t-k] continuum as containing an intervening [u] and [sh]-[t-k] as containing an intervening [i], sequential contrast but not compensation for coarticulation predicts that they should identify more of the continuum as ‘‘t’’ after [s] because [u] concentrates energy at lower frequencies than [i]. However, if they instead judge the stop in terms of the fricative that occurs next to it on the surface, then they should respond like English listeners, and their behavior is predicted equally well by sequential constrast and compensation for coarticulation. [Work supported by NIH.]