Word-initial Voicing Contrast in Khuzestani Arabic Stops

This research generally aimed to gain a better understanding of the stop voicing contrast in Khuzestani Arabic (KhA) by focusing on VOT and vowel onset f0. The potential effects of voicing status, place of articulation, vocalic context, and gender on the word-initial VOT and f0 measurements were investigated. 15 females and 15 males were asked to repeat 31 authentic words three times out of context with no marked intonation pattern. The results showed significant differences between VOT and f0 values as a function of voicing status, among others. In addition word-initially, the statistical analysis indicated significant place-dependent and vocalic-related VOT variations in the context of VOICELESS stops, whereas for f0 the effects of gender, place of articulation, and vocalic context were significant both at the VOICED and VOICELESS levels. Word-initially, in KhA /b, d, ɡ/ are produced with voicing lead, while their VOICELESS counterparts /p, t, k/ have long lag, and the two stops with no VOICED cognate /tˁ, q/ make use of short lag. Overall, KhA exhibits a two-way laryngeal contrast system, and similar to Swedish the two opposite ends of VOT continuum are utilized. Logistic regression models were executed separately for both genders to predict the levels of voicing status and emphaticness. The results revealed that in voicing contrast more weight was given to VOT compared to f0. Lastly, Pearson’s correlation showed a significantly positive linear relationship between the two acoustic parameters in marking the voicing feature of VOICELESS plosives.