ISCA Archive SpeechProsody 2012
ISCA Archive SpeechProsody 2012

English lexical stress and spoken word recognition in Korean learners of English

Jeonghwa Shin, Shari R. Speer

Two experiments explore how Korean-speaking L2 learners of English process English lexical stress during spoken word recognition. Korean doesn't employ lexical-level prosodic distinctions like English lexical stress, but it has phrase-level prosodic structure ((T) HLH), with the initial tone determined by the phonation type of phrase-initial sound. Results from eye-tracking and gating experiments showed slower processing times and lower accuracy for Korean learners' processing of English lexical stress information during word recognition compared to native English speakers. Processing difficulty was greater for iambic than for trochaic words, and time-course data indicate that processing difficulty increased between initial contact with the target word and the completion of word recognition. Post-hoc analyses of the eye tracking data also revealed that Korean L2 learners' word recognition was facilitated when the stress pattern matched the Korean tonal pattern induced by the word-initial phonation type contrast. Taken results together, the study implies that Korean L2 learners' processing of English lexical stress is modulated by both the processing strategy and constraints from perceptual experience with L1 phrase-level prosody.

Index Terms: lexical stress, phrase-level prosody, L1 transfer