| |
Abstract:
Studies showing that late-learners of a language often reach
native levels of proficiency with some skills but rarely do with
others, suggest there is considerable variability of
experience-dependent change (plasticity) in different subsystems of
language. It has been suggested that speech segmentation, breaking
continuous speech into units, is one of the less modifiable
subsystems, and that non-native speakers may be unable to use
language specific cues like the typical "strong-then-weak"
stress-pattern found in English. This study addressed the
questions: (a) Is semantic, syntactic and stress-pattern
information actually used during segmentation by native speakers?
and (b) Which of these cues are used by people who learned English
late in life? English sentences containing varying amounts of
semantic, syntactic and useful stress information were presented as
continuous speech. Native English speakers and native speakers of
Japanese who learned English after the age of twelve indicated
where target phonemes fell within words in the sentences. Results
from native English speakers indicated stress-pattern, syntax and
semantics were used during segmentation and that people used the
information flexibly. Results from native Japanese speakers
indicated they were capable of using the English stress-pattern to
segment speech even though this information would not be useful for
segmenting their native language. Furthermore, it was shown that
native and non-native speakers utilized semantic information
similarly, but the non-native speakers did not use syntactic
information as efficiently as native speakers. These results
demonstrate marked differences in developmental plasticity of
different subsystems within language.
|
