Ultrasound and acoustic analyses of the English lateral
Joo Kyeong Lee (University of Seoul)
Abstract
The current work attempts to investigate Korean learners’ production of English /l/
with a closer look at articulatory and acoustic substances. Ultrasound imaging of the
tongue was used to examine the coordination of the target gestures to compare
articulatory differences between clear-[l] in onsets and dark-[ɫ] in codas, and between
native speakers and Korean learners. Acoustically, F2-F1 values at the midpoint of /l/
and normalized duration of /l/ were analyzed together. In the experiment, 8 Korean
learners and 3 native speakers participated in ultrasound and audio recordings. The
results show that smoothing splines of the tongue mostly overlapped between onsets
and codas in the case of Korean learners. This suggests that Korean learners’
production of onset /l/ was not distinctive from that of coda /l/. The acoustic measures
of F2-F1 were consistent with the articulatory configurations; F2-F1 values were not
significantly different between onsets and codas, while native speakers’ F2-F1 values
of onset /l/ were significantly greater than those of coda /l/. Moreover, Korean learners
produced a significantly shorter /l/ in both onset and coda position than native speakers.
Overall, Korean learners were not able to articulate the darkening gestures for codas,
but they implemented clear-[l] as it occurs in coda position in Korean. The onset
produced by Korean learners was significantly shorter than that of native speakers.
Tongue tip raising was appropriately executed, but it seemed to be short of temporal
sustainability. This suggests that Korean learners produced the tap [ɾ] rather than [l] in
onset position.
Keywords
clear-[l], dark-[ɫ], ultrasound, acoustic, F2-F1, duration, Korean learners