This study examined the perception of lexical accent in Tokyo Japanese. The most salient acoustic cue in perceiving lexical accent in Tokyo Japanese is said to be a steep fall in f o from an accented syllable into the following one. At the same time, lexical accent is known to trigger a post-lexical process called downstep, lowering in peak f o and compression in the f o range, in phrases that follow. However, it is not clear if listeners use downstep in perceiving lexical accent. Even if they do, cue weighting between a steep f o fall immediately following accented syllables and downstep which occurs in the following phrase in perceiving lexical accent is unknown. To investigate how native Japanese listeners use these two cues in perceiving lexical accent, using a minimal pair of words that differ only in accentedness, 1) a steep f o fall after lexical accent and 2) downstep were independently manipulated in gradual steps to mimic the f o contours of accented and unaccented phrases and the stimuli created were presented to Japanese listeners. The results indicate that Japanese listeners use not only a steep f o fall of lexical accent but also downstep in perceiving lexical accent.
Sugiyama et al. (Wed,) studied this question.