Presentation: Sarah
Summary: Yu-chiao
The study aimed to investigate the function of
language experience on word recognition in foreign-accented speech. Two
experiments were conducted to examine
the priming effect of canonical, genuinely-accented and arbitrarily-accented
pronunciation on recognition of words. The first experiment recruited three
groups of listeners with different language backgrounds, including 48 native
listeners of English, 32 Italian learners of English, and 48 native Dutch listeners
proficient in English. Visual target words were 40 mono- and bisyllabic English
words with a short vowel /ɪ/, and 40 words with a long vowel /iː/. Auditory
prime words were threefold, containing a prime with the correct vowel length, the
incorrect vowel length, and an unrelated one. Additional 80 words without /ɪ/
and /iː/ were chosen as filler targets. Participants were instructed in
English, and were asked to judge whether the word shown on the screen was an
existing English word or not. RTs were submitted to analyses of variance to
examine the effects of version (canonical vs. variant), prime form (genuine vs. arbitrary), and relatedness
(related vs. unrelated). Results showed that native English listeners recognized
the canonical version the best. The genuine form (lengthening of /ɪ/) showed a
facilitating priming effect compared to unrelated primes, whereas the arbitrary
form (shortening of /iː/) did not. Italian listeners only showed the main
effect of relatedness, indicating no priming advantage of either the canonical
version or the genuine form. Dutch listeners performed similarly to English
listeners in terms of recognizing the canonical version the best. However, both
genuine and arbitrary forms exhibited
facilitating priming effects. In order to find out the possible reasons attributing
to the different performances between English and Dutch listeners, a second
experiment was conducted, attempting to investigate whether the difference was
due to more exposure to genuine Italian accents of English listeners or due to
a greater perceptual flexibility of Dutch listeners from their muitilingual
experience. Experiment 2 was thus the Dutch version of Experiment 1, conducted
with 32 Dutch participants. The result suggested that the recognition of Dutch
words could be facilitated regardless of whether canonical forms or variant
forms were heard. Such results supported the view that Dutch listeners had a
greater perceptual flexibility arising from their muitilingual experiences.