Osamu Takeuchi

Since the 1989 revision of the National Curriculum Standards, English teaching in Japanese junior high schools has focused on listening, reading, speaking, and writing (Ministry of Education, Cultu...

Piezoelectric braided cords composed of poly-L-lactic acid (PLLA) fibers can be integrated into textiles to detect mechanical signals without requiring sensors to be attached directly to the body. In this study, we investigated the potential of a bed-sheet-integrated piezoelectric braided cord sensor for unobtrusive physiological monitoring during sleep, using sleep apnea as an example of its possible applications. Finite element method (FEM) analysis was performed to qualitatively examine the directional characteristics of the piezoelectric response. Guided by these qualitative findings, a bed-sheet sensor with a multi-orientation stitching configuration was developed through iterative prototyping and experimental evaluation. Piezoelectric braided cord signals were recorded from 23 participants during their routine clinical polysomnography (PSG) examinations. The developed sensor enabled the detection of respiration- and pulse-related mechanical signals during sleep. During the initial analysis of data from a small number of these participants, we observed that a pulse-related periodic component remained detectable in the respiratory signal when respiratory motion decreased during apnea. Based on this observation, an apnea-related signal detection algorithm was developed and subsequently applied to the complete dataset of 23 participants. The sensor-based analysis was performed independently of the physicians’ PSG-based assessments. For the 18 participants with available physician-based apnea event assessments, the sensor-detected event counts were compared with PSG-based counts within 5-min intervals, without event-by-event temporal matching. These findings demonstrate the potential of the proposed sensing approach and provide a basis for future clinical investigations.

In recent years, research on language teacher well‐being (LTWB) has expanded rapidly. However, the conceptual relationship between general teacher well‐being (TWB) and LTWB remains insufficiently theorized, particularly in contexts where teachers simultaneously occupy subject‐specific and general education roles. Although existing frameworks, such as the Conditions‐Components‐Outcomes model, have contributed to structural dimensions of TWB, they have not explicitly focused on the dynamic processes through which subject‐specific and general‐teacher roles interact in shaping well‐being. This article proposes an ecological role‐interaction model of LTWB, conceptualizing well‐being as an emergent process arising from the interplay between subject‐specific and general teacher roles within multilayered institutional ecologies. Drawing on ecological and role‐based perspectives, it proposes that LTWB and TWB are neither hierarchically ordered nor fully separable constructs. Rather, they are dynamically shaped through ongoing role negotiation. The case of Japanese teachers of English provides a particularly valuable context for illustrating and refining the proposed model. By foregrounding role interaction and role negotiation within broader ecological systems, this model offers a new conceptual framework for understanding LTWB and provides directions for future research and empirical investigation in applied linguistics.

This study compared the linguistic resources writers use to express attitudes, signal certainty, and interact with readers, that is, stance and engagement patterns. We examined these patterns in large language model (LLM)‐generated essays and human learner essays across Common European Framework of Reference for Languages (CEFR) levels A2 through C1. We analyzed 592 human essays alongside 4724 LLM responses to identical writing tasks produced by six models. Task‐controlled mixed‐effects models revealed systematic divergences that exceed what raw frequencies suggest, indicating that structural proficiency can mask functional misalignments. At A2, LLMs overused attitude markers (71% higher than humans) while underusing engagement markers (22% lower). At upper levels, gaps widened further: self‐mention collapsed to 41% of human rates at B2 and 27% at C1, engagement markers fell to 27% at both B2 and C1, and hedges and boosters dropped to approximately 61% of human rates at C1. These patterns held across all six models, indicating that CEFR‐conditioned generation produces a functional profile that is consistent across models but mismatched to human writing: it associates lower proficiency with evaluative excess and higher proficiency with suppressed interactional marking. When LLM texts serve as models or feedback sources in language teaching, practitioners can expect stance and engagement profiles to diverge from human learner writing in level‐dependent ways.

This study investigates the linguistic forms university students notice through ChatGPT-generated feedback in second language (L2) academic writing and their subsequent application. While AI-mediated automated writing evaluation (AWE) shows promise, the transition from noticing to productive use remains under-explored. Qualitative analysis of student interviews, ChatGPT history logs, and draft revisions revealed that although participants noticed diverse academic vocabulary and complex syntactic structures, such awareness rarely transferred into independent use. Findings indicate a clear gap between receptive and productive knowledge, which may be hindered by cognitive and psychological constraints. High cognitive load, fear of making mistakes, and the pressure to prioritize task completion can often impede the retrieval and experimental use of newly noticed forms. Consequently, this study argues that providing AI feedback alone may be insufficient for long-term L2 development. Pedagogical interventions could be helpful to externalize the noticing process and provide scaffolds for retention. Specifically, utilizing ChatGPT’s dialogic features can facilitate hypothesis testing and the exploration of alternative expressions in a non-evaluative environment. This study is expected to contribute to the theoretical understanding of AI-mediated noticing and the practical design of individualized L2 writing instruction.

Large language models (LLMs) can generate text that resembles L2 learner writing at requested proficiency levels, but whether such texts adapt to communicative contexts as human learners do remains unclear. This study compared LLM‐generated and human second‐language writing across three task‐defined categories (informal–personal, semi‐formal–expository, formal–argumentative) and four CEFR levels (A2–C1) using Biber's multi‐dimensional analysis (MDA). Six instruction‐tuned models responded to Write & Improve prompts ( n = 592 human; n = 4724 LLM), with EFCAMDAT as a human baseline. Under CEFR‐only prompting, LLM outputs shifted on MDA dimensions tied to informational density and explicitness, but showed weaker, differently directed, or less variable differentiation on dimensions tied to persuasion and elaboration, especially in formal–argumentative tasks. The largest observed gap was at C1, where CEFR descriptors most emphasize register flexibility. LLM outputs were more uniform than human writing and clustered near the center of the human distributional space, showing less variation than authentic learner texts. The broad pattern recurred across all six model families. CEFR‐conditioned generation can reproduce some structural signals of proficiency while showing limited human‐like register differentiation in MDA space, a pattern we interpret as form without function.

Before the introduction of conceptual metaphors, traditional rote memorization tasks were the primary methods for teaching English phrasal verbs (PVs). Verb and particle combinations in idiomatic PVs were considered arbitrary, leaving learners with few options beyond memorizing lists. As a result, learners often avoided PVs in favor of single-word alternatives due to their perceived ambiguity. Research in EFL contexts later explored conceptual metaphors as a means of improving PV acquisition and reported greater effectiveness compared to rote memorization. However, subsequent reviews of the literature found that these results may reflect the inadequacy of traditional tasks rather than the superiority of conceptual metaphors. To address this issue, the present study replaced rote memorization with a more robust vocabulary learning technique: retrieval practice. Eighty-five non-English majors at a Japanese university participated in an online study during the Covid-19 lockdowns. Results revealed no statistically significant difference (p > .05) between the conceptual metaphor and retrieval groups, indicating that conceptual metaphors are not the only effective method for facilitating PV acquisition.

Learning vocabulary is central to learning a second or foreign language (henceforth L2). L2 vocabulary research shows how individual differences (IDs), such as metacognitive elements, motivational factors, and strategy use, contribute to vocabulary learning and acquisition. L2 vocabulary learning that incorporates IDs is often discussed in relation to the internalized vocabulary learning process, including how L2 learners acquire vocabulary by sustaining motivation, using vocabulary learning strategies, and regulating learning processes. The associations between the IDs and vocabulary learning have often been examined within the framework of self-regulated learning (SRL). However, discussions on how SRL processes are reinforced through co-regulated or socially shared learning with others and within social contexts remain scarce. In this article, we introduce a sociocultural perspective on strategic vocabulary learning. Specifically, we discuss how the internalized processes of self-regulated vocabulary learning can be socially constructed through co-regulation and socially shared regulation. The proposed framework has the potential to provide future research agendas and instructional practices in L2 vocabulary learning and teaching. We conclude by proposing directions for future research on the socially constructed nature of strategic vocabulary learning.

This study investigates how Large Language Models (LLMs) encode second language (L2) writing proficiency distinctions compared to human learners, focusing on the structural alignment between synthetic outputs and human developmental patterns. We analyzed CEFR-graded Write & Improve 2024 learner essays and matched LLM generations (A2–C1) using statistical models controlling for prompt and text length. With a CALF feature set (Complexity, Accuracy, Lexical Complexity, Fluency), level explained 2.9 % of variance in human texts ( R ² = 0.029) but 10.6 % in pooled LLM outputs ( R ² = 0.106), indicating clearer A2–C1 features in LLM-generated texts overall. LLM level effects were robust ( R ² range = 0.038–0.454), with sharper differences between neighboring CEFR bands than in learner essays (e.g., A2–B1 distance, D ² = 4.81 versus 1.07). Prompt-wise analyses found effects for all 10 prompts in the LLM set but only 2/10 in the learner set, with much greater separation than in the learner cohort. These results support CALF and CEFR construct validity: with prompt and length controlled, A2–C1 texts show ordered CALF gradients, especially for LLM texts across prompts. The sharper LLM separation likely reflects idealized, low-variance production; therefore, CALF captures core structural proficiency (syntax, lexis, accuracy, fluency) but not discourse-pragmatic qualities or human variability. The analyses provide level-calibrated evidence that LLM texts show clearer distinctions than learners’, under identical prompts, whose development is gradual and overlapping. This positions LLMs as both tool (exemplars, rubric calibration) and challenge (assessment validity and fairness), while offering SLA researchers insight into how proficiency constructs are encoded in human versus model-based writing.