Key points are not available for this paper at this time.
Past studies (Vance, 1979; Kawahara, 2012) of rendaku, a morphonological alternation in Japanese, have produced conflicting results about the sensitivity of Lyman’s Law to a locality effect in nonce words. In a large-scale forced-choice experiment with 72 stimuli, our analysis of the responses from 180 native speakers of Japanese shows that for many speakers, Lyman’s Law is indeed sensitive to a locality effect: in nonce words, the blockage effect of rendaku by Lyman’s Law tends to be stronger when the blocker consonant is in the second syllable than when it is in the third syllable. This finding supports Vance’s original insight. Then, to explore why Kawahara’s study failed to find a locality effect, we replicated it with a larger number of speakers, and found some evidence that the locality effect is identifiable in a naturalness judgment experiment as well.
Kawahara et al. (Thu,) studied this question.