For Japanese university students, not good at understanding spoken English, one of the major challenges is that they cannot segment a speech stream and recognize words/phrases because what they hear does not match the mental acoustic image they have of them, and they are too unfamiliar with connected speech features. The speech segmentation and word recognition, components of bottom-up processing, are even harder if they are unaware of rhythmic differences between English and Japanese, subconsciously applying the Japanese rhythm structure when processing spoken English. This presentation proposes that connected speech training with explicit instruction of L1-L2 rhythmic differences should be incorporated in Japanese EFL classes to help less competent listeners to overcome such bottom-up listening challenges. The need of tuning up their mental acoustic image of words/phrases and building automaticity in using skills should also be addressed there.

The study to be reported here was conducted in the 2021 fall semester, with 31 university students enrolled in a freshman required English course. They were placed in the second-highest of five classes based on an in-house placement test in April and stayed in the same class in the fall semester. Based on observation of their performance in the spring semester, most of them had CEFR-B1-level reading skills at the beginning of the fall semester. However, a test to check their bottom-up listening skills in October (the pre-test in this study) showed that dealing with connected speech (even in A1/A2-level materials) was challenging for many, though they gained some familiarity with spoken English and its rhythm through the previous semester.

In order to train their bottom-up listening skills, more focus was placed on connected speech training in the fall semester. New additions were 1) self-grading VOA listening exercises focusing on connected speech and rhythm (used with another set of VOA quizzes for comprehension and vocabulary/grammar); 2) instructional videos (used with connected speech training through songs); 3) feedback videos (in place of written feedback) to their reaction papers in an attempt to raise metacognitive awareness. During the 13-week semester, before the class, the students worked on two VOA quizzes and song listening. In class, they received feedback on their reflections, lectures on the week’s song and VOA story, oral practice opportunities, and further exercises. Then they were given automaticity/fluency building activities via Kahoot!/Quizlet Live and speech production. Finally, they reflected on their learning and performance to submit the week’s reaction paper. In Week 13 they took the same bottom-up listening skills test (the post-test) and answered a post-course survey.

Compared to the pre-test, the mean correct answer rate in the post-test increased remarkably, from less than 60% to nearly 80%, suggesting that the students were now much better able to segment spoken English and recognize linked words, reduced syllables, and assimilated sounds. According to the post-course survey, all were satisfied with their learning in this course and nearly 90% of the students felt their English improved, especially in listening.