This article reports on a longitudinal interview study that focused on high-achieving Japanese university students and their development and changes of ideal L2 self (Dörnyei, 2009) and intrinsic motivation (Deci & Ryan, 1985, 2002). Semi-structured interviews were conducted with approximate half-year intervals for two years. Participants were two of the interviewees in Takahashi (2014), who were contrasting in terms of the vividness of their ideal L2 selves. The results indicated that although each of the interviewees seemed to perceive English as relevant for their future, their ideal L2 selves developed in distinct ways, and the intensity of studying English also varied. In crystallizing their ideal L2 selves, what seemed to play an important role was their possible career plans related to their majors. Furthermore, one of the interviewees not only developed clear ideal English selves but ideal L3 (French) selves, helping him decide to study in France for a year. These results indicate that for university students, future career aspirations might play an important role in developing elaborate ideal L2 selves, and when supported by both future career goals and intrinsic motivation, the development of an ideal L3 self is also possible.

View full abstract

Performance assessment attracts rising attention in recent English language education, but little has been introduced to English language program in Japan. This is because such assessment imposes severe burdens on instructors: conducting interview, rating students' performance, and analyzing and delivering scores. An automated scoring system for students' performance can be a solution to this problem. To present the solution, we developed an automated scoring system for speech by Japanese learners of English and examined the practicality of the system in an English language program. The system assesses the students' responses to the discourse completion tasks that were created to elicit the expressions learned in the lessons. The responses are represented as vectors applying the bag-of-words model, and the system scores the response using Support Vector Machine and Naive Bayes Classifier. The accuracy of these two prediction methods is estimated using cross-validation. The exact agreement is 74% between our system and the human scores.

View full abstract

It is known that many of us hear an “inner voice” during silent reading. We call this auditory processing “phonological coding,” and its processing time is estimated by a simple linear equation: [120×Sy + 80×Cn] millisecond; where Sy is the number of syllables and Cn is the number of consonants in a target text. Subtitles on TV or videos are made for viewers to read out while the speech action is on the screen. For example, BBC’s guideline says, “The recommended subtitle speed is 160-180 words-per-minute,” and TED’s guideline says “Keep the subtitle reading speed at a maximum of 21 characters / second”. In this report, we compared TED Talks’ subtitle length in time and their predicted duration time for phonological coding calculated by the linear equation. The average length of fifty thousand TED Talks’ subtitles included in two hundred videos is measured to be 2.95 second, and calculated by the equation to be 2.88 second. Quantifying the difference between two groups of measured and calculated values, Cohen’s d effect size is 0.07. These findings suggest the appropriateness of the equation to predict the time for phonological coding, and also might suggest that TED viewers would be able to read out the subtitles just before they vanish away.

View full abstract

Aligning linguistic representations at each level (e.g., sound, syntactic, semantic, etc.) is important for not only L1 speakers but also L2 learners to achieve smooth communication (Pickering & Garrod, 2004). At the syntactic level, previous studies revealed that learners’ repetitive use of the same structure with written primes (syntactic priming), the preference of Prepositional-Object (PO) targets compared to Double-Object (DO), and the lowest priming with uncontrolled spoken primes (Morishita, Satoi, & Yokokawa, 2010; McDonough, 2006; Morishita, 2011a). However, the effect of modality has not been clarified. Thus, the emergence of syntactic priming by presenting both spoken and written primes needs to be elucidated. The current study conducted a picture description task with controlled spoken primes to investigate whether the differences of output modality (i.e., spoken or written) and verb (i.e., same or different between primes and targets) affect Japanese EFL learners’ syntactic priming. The participants described pictures in either forms after listening to the primes with a PO, DO, or filler. The results show that a priming effect was observed in both modalities, indicating learners successfully understood the primes. Therefore, syntactic structures were activated by the spoken primes, and it enforces syntactic processing when learners encounter the same structures again.

View full abstract

Two experiments investigated whether low- and intermediate-level Japanese EFL learners can successfully construct syntactic and semantic structures after listening to and reading object relative clauses (ORCs) and passives. Experiment 1 investigated whether 120 learners utilize animacy when comprehending ORCs. After listening to or reading sentences, including 20 stimuli with the manipulation of animate or inanimate sentential subjects, the participants answered comprehension questions. The results showed that 70 low-level participants achieved only 50% accuracy for animate noun questions but a higher rate for inanimate ones, possibly guided by their world knowledge, thus illustrating the participants’ lack of ability in constructing syntactic structures. To clarify whether they constructed syntactic structures online, Experiment 2 utilized a passive/active sentence–picture verification task with the manipulation of animacy for two nouns. After listening to or reading 12 stimuli and 48 fillers, 48 participants performed correctness judgments by matching pictures. The accuracy rates and response times were compared to examine the influence of animacy, syntax, proficiency, and modality. The low-level participants, who were poorer at listening than reading, showed a lower accuracy rate for animate than inanimate passives and a slower latency for animate than inanimate actives, which revealed their difficulty in constructing syntactic structures.

View full abstract

The effects of international online video talk in a language exchange situation on Japanese EFL (English as a foreign language) college students taking a teacher training program are discussed in this paper. Twenty-eight college students who were taking a teacher training program in Tokyo and 14 university students who were taking a Japanese language course in Melbourne, Australia, participated in an international exchange project through online video talk. The effects of this international exchange experience were researched, with special focus on the willingness of the Japanese students to communicate, their motivation to learn English, and their perspective on teaching English in high school. The video talk situation contributed to increasing their willingness to communicate, because they could have stronger feelings of friendship for their partners due to the visual information received through the video screen. Also, the language exchange situation led them to feel more motivated to acquire fluent English communication ability, because the high Japanese proficiency of some of their partners stimulated them to make more effort to learn English. Furthermore, they found the international exchange experience through video talk very impressive and came to think that they should use it in their future English teaching in high school.

View full abstract

The primary objective of this paper is to investigate the influence of pedagogical intervention on the L2 motivation of Japanese university students. It consequently examines the sensitivity of the new questionnaire developed by Agawa and Takeuchi (2016c). A quasi-experimental study was conducted to compare two types of instructions: conventional instructions used in test-preparation courses (contrast group [CG]) and instructions based on self-determination theory (SDT) (treatment group [TG]). In the CG, 23 students worked on tasks similar to TOEFL and IELTS questions. In the TG, 24 students received instructions designed to satisfy their needs for autonomy, competence, and relatedness. The newly developed questionnaire based on SDT was conducted at the beginning and end of the academic year. The results indicated that, in the TG, the students’ needs were more satisfied after the intervention. They also showed that, after the treatment period, their self-determined forms of motivation significantly increased. Meanwhile, in the CG, neither the degree of students’ needs satisfaction nor motivation intensity showed a significant difference. These results suggest that fulfilling the needs could help enhance Japanese EFL learners’ motivation. The results also demonstrate the sensitivity of the new questionnaire in measuring changes in the needs fulfillment degrees and L2 motivation intensities among Japanese university students.

View full abstract

TED Talks are widely used for English Education today. One practical reason is that English subtitles are available for free. There is a guideline for creating subtitles so that the audience of the TED video can easily understand the message in terms of the human cognitive load of recognizing alphabet letters at one time during watching video talks. In this study the authors attempted to analyze the display time of subtitles from 2,143 TED Talks videos which include 577,457 subtitles in order to clarify the characteristics of their distribution and mean time length. We then discuss the result compared with the above guideline for the limitation of linguistic information in the subtitles. The result shows that the mean length is 2.8 seconds with 1.0 standard deviation and a logarithmic normal distribution. The distribution indicates a similar phenomenon to our previous study which analyzed the length of breath groups observed in an English movie corpus named Seleaf. We also discuss the educational use and precaution of TED subtitles in the language classroom with ICT settings.

View full abstract

This research focuses on the effect of first language (L1) on discourse marker (DM) use in the second language (L2). Some researchers suggested that non-native speakers may use English DMs under the influence of their L1. However, there have been only a few empirical studies addressing this issue. The current research consists of two cross-linguistic studies to rectify the inadequacies in the literature. The first study conducted a contrastive analysis using English-Japanese parallel corpus data. The result revealed a complicated relationship between English DMs and their Japanese equivalent expressions. Most English DMs correspond to a wide variety of Japanese DMs, while some Japanese DMs correspond to different English DMs. Based on these 166 correspondences, in the second study, both English and Japanese speech data collected from picture description tasks were analyzed in terms of DM use. The results suggest that Japanese learners’ L1 use may have an influence on the frequency of some markers such as and, so, and but, leading them to overuse these items. These findings may be part of the evidence of L1 effects on DM use in L2 learners’ speech and contribute to identifying the features of their acquisition of DMs.

View full abstract

This study investigated how on-task behavior of EFL learners is associated with their partner’s and their own motivational dispositions during a communicative task. A total of 114 university students were randomly assigned to pairs and engaged in a decision-making task. Learners’ on-task behavior was measured by counting the total number of English words produced and turns taken during the task (i.e., task engagement variables). Their motivational dispositions during the task were measured using a 5-point Likert scale questionnaire. Cluster analysis was performed to classify learners into high motivation and low motivation learners. Correlation analyses were used to examine the relationships between their trait and state motivational variables and on-task behavior. The results showed that for both high motivation and low motivation learners, state motivational variables correlated with their task engagement variables. The result also demonstrated that only when highly motivated learners were paired with highly motivated learners, task engagement significantly and positively correlated with their partners’ motivational dispositions. This study concluded that for successful completion of an interactive task, a partner plays a significant role only for highly motivated learners, and spontaneous scaffolding is not expected to happen between learners with different levels of motivation.

View full abstract

This article describes an empirical study that investigated the efficacy of the peer-review process on students’ academic papers. The author implemented peer-review (PR) activities in a Project-based English Program during the fall semester of 2015, and discusses the findings of the study. The participating students of the study were 49 sophomores from the College of Life Sciences. The participants were provided with four PR worksheets to make constructive suggestions or commentary regarding the fundamentals of an academic paper: the organization of second-language texts and logical development of ideas/arguments. During this interactive activity, students discussed how their texts should be organized and developed. They revised their texts as needed, and submitted the completed papers at the end of the course. The author and one independent rater assessed participants’ texts before and after the PR process, using the writing rubric constructed for this study. The results of this evaluation showed positive increase in average mean between the pre- and post-PR scores. Students who successfully improved their papers during the process showed a better understanding of the necessary elements of an academic paper, and made their logic clearer to the audience. Therefore, the author concludes that the PR process did enhance the improvement of students’ revisions.

View full abstract