Vowel onset characteristics as a function of voice and manner contrasts in Persian coronal stops

Acoustic and electrolaryngographic analysis was carried out on vowel onsets after coronal stops produced by seven male native speakers of Persian (Farsi) in syllable initial position. In addition to voice onset time (VOT), we present measures of pitch (Fx), closed quotient (Qx) and spectral tilt (ST).

Results show that VOT distinguishes between voiced and voiceless stops, and between plosives and affricates. Pitch at vowel onset, as measured electrolaryngographically, distinguishes voiced from voiceless plosives but not voiced from voiceless affricates; it does, however, distinguish the plosives from the affricates. Closed quotient also distinguishes voiced from voiceless plosives but not voiced from voiceless affricates; it distinguishes /d/ from /dȝ/ but not /t/ from /tʃ/. The spectral tilt measure distinguishes voiced from voiceless stops but not plosives from affricates.

Closed quotient and spectral tilt were found to be closely positively correlated but with some evidence that they are at least partially independent. This independence may enable variation between whispery voice and breathy voice at the onset of vowels after phonologically voiceless stops, and variation between whispery voice and modal voice at vowel onset after phonologically voiced stops, particularly /dȝ/.