The acoustic realization of the Korean sibilant fricative contrast in Seoul and Daegu

The neutralization of / $s^h$ / and / $s^*$ / in Gyeongsang dialects is a culturally salient stereotype that has received relatively little attention in the phonetic literature. The current study is a more extensive acoustic comparison of the sibilant fricative productions of Seoul and Gyeongsang dialect speakers. The data presented here suggest that, at least for young Seoul and Daegu speakers, there are few inter-dialectal differences in sibilant fricative production. These conclusions are supported by the output of mixed effects logistic regression models that used aspiration duration, spectral mean of the frication noise, and H1-H2 of the following vowel to predict fricative type in each dialect. The clearest dialect difference was that Daegu speakers' / $s^h$ / and / $s^*$ / productions had overall shorter aspiration durations than those of Seoul speakers, suggesting the opposite of the traditional "/ $s^*$ / produced as [ $s^h$ ]" stereotype of Gyeongsang dialects. Further work is needed to investigate whether / $s^h/-/s^*$ / neutralization in Daegu is perceptual rather than acoustic in nature.