Effects of the Stress Cues on English Word Recognition by Korean Learners of English

Abstract

This study investigates the effects of the stress pattern (trochaic vs. iambic) on English word recognition by Korean learners of English. Speech segmentation, more specifically word recognition, has been known to be affected by the language-specific cues. In the present study, stress is of main concern as a word recognition cue because English and Korean have different prosodic characteristics in terms of stress. While stress has a contrastive function and seems to constrain lexical access in English, Korean is known to have no stress system on the word level and little has been studied about its role in Korean word recognition. As a result, three research questions were proposed as follows: (a) can Korean learners of English use an initial-stressed syllable of a word as a cue in word recognition of an English connected speech?

(b) are there any differences in the performance by Korean learners of English at different proficiency levels?

(c) is there any interaction between the stress pattern and the other factors including the syllable count and the word class?

In order to answer these questions, the word spotting task was conducted with a total of 42 university students, who were born and raised in Seoul or Gyeonggi-do province. They were divided into two groups, the advanced and the intermediate-low group. The participants were asked to detect a real English word from a stream of nonsense syllables. To examine the effects of the stress pattern and its interaction with the other factors including syllable count and word class, the materials used in the task were carefully selected, including disyllabic and trisyllabic nouns or verbs. The task was followed by a word knowledge test containing the list of the target words in order to confirm that they target words were highly familiar to the participants.

The findings of the study suggest that the Korean learners did not seem to use the trochaic pattern to set a word boundary. They responded faster and more accurately to the target words with the iambic stress pattern. Furthermore, there was no difference between the groups, suggesting that the L2 prosodic cues like stress are hard to be acquired. The other factors of concern in the present study such as syllable count and word class did not show any interaction with the stress pattern. In conclusion, Korean learners of English do not seem to use the frequent distributional stress cue in English when recognizing a word from a sequence of nonsense syllables. These findings not only provide some understanding on the speech segmentation by Korean L2 listeners, but also shed light on the significance and necessity of the instruction on stress in teaching English listening.