English Native and Nonnative Speakers' Perception of English Reduced Word Forms with Reduced Vowels

It has been controversial whether phonetic-fine details of the lexical items are encoded and accessed in L1 or L2 speakers' lexicon and how the variants of the reduced word forms are recognized. To address these issues, in this study we carried out the perceptual word recognition experiment, i.e., AB identification test with English native speakers and Korean nonnative English speakers. Two types of synthesized listening stimuli were created from two types of the original word forms by shortening or lengthening the durations of the reduced vowels: words containing the reduced vowels and words without them in the first syllables. First, the results show that a wider variety of tokens of the variants of reduced word forms differing only in the durations of the reduced vowels were recognized with substantially high rates as those with the reduced vowels for English L1 speakers than for Korean L2 English speakers. These findings indicate that phonetic-fine detail information of the reduced word forms with respect to the vowel durations is represented and reconstructed in the lexicon for L1 and L2 speakers' lexicon although their representational robustness varies depending on L1 or L2 status. Furthermore, we show that the degrees of vowel reduction exert the differential effects on the reconstruction or identification of English words according to the word-initial C(V)C types and L1 or L2 status. It also turns out that depending on the presence/absence of the reduced vowels in the words, Korean L2 English speakers are less sensitive or oversensitive to the vowel duration when they recognize the words.