The purpose of this study was to investigate how Korean stops were manifested under three different hyper-articulated speech conditions such as clear speech, prosodic domain-initial position, and focused position in Kyungsang Korean. In the previous s ...
The purpose of this study was to investigate how Korean stops were manifested under three different hyper-articulated speech conditions such as clear speech, prosodic domain-initial position, and focused position in Kyungsang Korean. In the previous studies with tonal and non-tonal dialect speakers of Korean, it has been found that Seoul dialect speakers use both VOT and onset f0, while Kyungsang speakers make use of VOT to distinguish the three-way contrast of Korean stops (Lee and Jongman, 2012). Given that Seoul and Kyungsang speakers use VOT and onset f0 in a distinct way to signal the laryngeal contrast of Korean stops, it was expected that they showed different strategy in the use of phonetic properties of Korean stops to enhance the contrast in the three different hyper-articulation speech conditions.
Ten female speakers participated in the study. All subjects had lived and been educated in Kyungsang dialect and their parents spoke the same dialect. Each of two bilabial aspirated and lenis stops, /p/ and /pʰ/ were embedded in disyllabic words including vowels /i, a, u, e/. The target words were placed in initial position and medial position of intonational phrase (IP) under focused position and unfocused condition. Then subjects were asked to read the test sentences with two different speaking styles such as casual speech and clear speech.
Each participant was recorded using a digital recorder (Roland R-05), and the recorded material was sampled at 44,100 Hz with a 16-bit quantization. The voice onset time (VOT) and vowel duration, peak of fundamental frequency (F0) and the first formant and the second formant of vowels were measured using Praat.
The results of current study revealed that the VOT of the aspirated and lenis stops were increased in the IP-initial position than in IP-medial position. The F1 and F2 of vowels, /i, a, u/ were enhanced in the IP-initial position than in the IP-medial position. Compared with the results of aspirated stops, the difference of VOT duration of the lenis stops was relatively great between the two different prosodic positions. However, the vowel durations following the aspirated stops were found to be significantly greater in the IP-initial position than in IP-medial position while lenis stops did not show vowel difference between the IP-initial position and the IP-medial position.
Under the different speaking styles, the vowels after the aspirated stop and lenis stop were longer in the clear speech than in the casual speech. It was also found that the F2 of /i/ and F1 and F2 of /a/ were enhanced in the clear speech.
When the two Korean stops were produced in the focused and unfocused condition, the F0 peak was found to be significantly greater in the focused condition than in unfocused condition. In addition, vowel space expansion was also found in the focused condition: F2 of /i/, F1 and F2 of /a/, and F1 of /u/ and /e/ were increased in the focused condition.
In conclusion, the present study provided evidence showing that Korean stops were differently manifested under each hyper-articulated speech condition and that the enhancing strategies were found to be different between the aspirated and lenis stops of Korean. Compared with the results from Seoul dialect, Kyungsang dialect speakers did not show consistent interaction effects across the different hyper-articulated conditions.