Perception of a Japanese moraic obstruent in an identification task

Native Japanese speakers perceive a moraic obstruent when the closure duration between successive moras is longer than a normal obstruent. However, it has not been well clarified how the perception of the moraic obstruent relates to the closure and neighboring moras’ duration. To investigate this point, a perceptual experiment was conducted with stimuli consisting of two‐mora nonsense syllables with a closure between the moras (/bipa/, /guku/, /kuku/, /kuto/, and /tapi/). The closure duration was shortened or lengthened in 10‐ms steps. The duration of the first mora was modified by changing its vowel duration to 50%, 100%, and 150% by the STRAIGHT method. Forty native Japanese speakers identified whether the nonsense syllable contained a moraic obstruent with the 2AFC method. Results showed that the perceptual boundaries of the moraic obstruent were 182 ms (SD=13.2 ms), 204 ms (SD=14.7 ms), and 222 ms (SD=16.6 ms), respectively, with 50%, 100%, and 150% vowel durations in the first mora, and that a logist...