Some acoustic characteristics of the geminates in the Ikema dialect of Miyako Ryukyuan
Résumé
Ikema is an endangered Miyako Ryukyuan dialect spoken in the Okinawa islands of Japan,
having a set of voiced/voiceless geminates in word-initial and medial positions. Among its obstruents,
the geminates /tt ff ss vv zz tts/ appear in initial and medial positions (/tta/ ‘tongue’, /ffa/ ‘child’, /ssa/
‘grass’, /vva/ ‘you’, /zzu/ ‘fish’, /avva/ ‘oil’) and the geminate /dd/ occurs only in medial position
(/badda/ ‘side’). This paper is an acoustic description of some aspects of these geminates, as a followup
to the rt-MRI study by Fujimoto and Shinohara (2015), which showed articulatory differences
between voiceless and voiced, as well as singleton and geminate consonants. In this study, we analyzed
sets of real words uttered several times in isolation by five native speakers by using oscillograms and
spectrograms. First, we confirmed longer closure duration for plosives and affricates and longer
frication noise for fricatives in geminates as compared to singletons. Second, our acoustic analyses
found full voicing of voiced geminates in Ikema. This result seems to be a consequence of pharyngeal
expansion identified in the MRI study during voiced geminate constriction. Third, VOT was
systematically shorter for the geminate voiceless plosive (/tt/) compared to its singleton counterpart.
This corresponds to the faster transition from the geminate consonants to the following vowel found in
the MRI study.