Abstract—As we are often blind to our conversational lapses
and shortcomings, the presenter will discuss the issue of
dysfluency based on L2 interactions by Japanese speakers
based on the JUSFC2018 corpus. The study’s first aim was to
examine if the number of words and mean length runs
increased with proficiency, as represented by TOEIC scores
(Group 1: scores 150-370; Group 2: scores 371-570; Group 3:
scores 571-770). The second aim was to compare the dialogic
fluency of each group of Japanese EFL learners with that of
native speakers to identify significant differences regarding
speaking rates, as well as acoustic, lexical and syntactical
dysfluency. Results showed that the number of words only
increased in the second range, before dropping in the most
proficient range; likewise, mean length runs (MLRs) showed
an increase from 11.2 syllables from Group 1 to 30.2 syllables
in Group 2, before dropping in Group 3 to 9.7. Concerning
possible differences in the number of words, Kruskal-Wallis
tests showed that there were statistically significant differences
in speaking rates, cross-talk pausing, the total amount of
silence, the percentage of silence, length of pauses, and the use
of L1 among the three groups of EFL learners and native
speakers. The post hoc tests of pairwise comparisons revealed
that native speakers differed from all three EFL groups. The
speaker will also discuss the issue of production, in particular
how individuals can be more aware of their fluency to provide
more meaningful, fluent and productive interactions.
Index Terms—Fluency, shyness, proficiency, hesitation
phenomenon.
The author is with Kyushu Institute of Technology, Japan (e-mail:
long@dhs.kyutech.ac.jp).
[PDF]
Cite:Robert W. Long II, "Lessons Learned: Insights from Japanese L2 Conversations," International Journal of Languages, Literature and Linguistics vol. 6, no. 1, pp. 69-75, 2020.