My latest projects

• Initial clusters in Yine (‘Piro’).
  Word-initial clusters in Yine have been analyzed as involving an appendix structure. However, a corpus study reveals effects of the Sonority Sequencing Principle (SSP), which has traditionally been associated with complex onset representations. This raises questions about the relationship between the SSP and cluster representation—or, more broadly, between phonotactics and syllabification. In this case, phonetic cues can account for many of the data patterns, including some that the SSP cannot explain.
• Learnability of length-referencing alternation in Japanese and an artificial language.
  The unnatural alternation of Japanese /g/ nasalization conditioned by mora length is not internalized by native speakers, as evidenced by a wug test extending this pattern to novel words. [abstract and slides for AMP2023] A structure akin to it is also examined in an artificial language learning experiment, which demonstrates a universal learning bias against referencing length, or ‘counting.’
• Utterance-final high vowel diphthongization in Chongqing Mandarin.
  I analyze vowels such as /i#/[iɪ#] and /u#/[uʊ#] as the results of the interaction of two independent constraints: final lengthening and *HighHigh. [abstract and slides for MFM30] Dispersion Theory can predict the inventory of diphthongization processes of this kind.