Pre-task planning is one of the widely used approaches to improve second language (L2) fluency in task-based language teaching (Ellis et al., 2020). However, the effect of the language used in pre-task planning on L2 speech fluency remains underexplored. The current study employed a between-participants design to compare the effects of planning languages at three levels: L1 planning, L2 planning, and no planning (a control group), across a narrative task on four measures of L2 speech fluency among 84 Chinese EFL learners. The study also investigated the moderating role of L2 proficiency. Results revealed that L1 planning significantly improved speech rate and marginally reduced between-clause pauses, while L2 planning showed weaker, marginally significant effects. Neither language of planning condition significantly affected within-clause pausing or repair fluency. In addition, within-group correlation analyses revealed that higher L2 proficiency in the L1 planning group was strongly correlated with all fluency measures. Pedagogical implications are discussed in terms of how to optimize L2 speech fluency in different planning languages across different L2 proficiency levels.

The present study investigates GenAI-mediated vocabulary learning strategies (VLS) among Chinese university English learners by integrating measurement validation and person-centered analysis. A sample of 593 Chinese EFL students completed a 24-item questionnaire assessing four strategy dimensions: AI-assisted lexical discovery, AI-supported lexical elaboration, AI-mediated metacognitive regulation, and AI-based lexical activation. Confirmatory factor analysis supported a correlated four-factor structure with strong reliability and acceptable convergent validity, although high inter-factor correlations suggested closely integrated constructs. Latent profile analysis identified three learner groups, including low-to-moderate, moderate-to-high, and high strategy users. Profile comparisons revealed that higher-use groups reported greater AI experience, higher self-rated proficiency, and stronger vocabulary-related performance. No significant differences were found for gender or year of study. The findings extend VLS research into AI-mediated contexts, highlighting the importance of strategic engagement and supporting a reconceptualization of vocabulary learning as a dynamic human–AI interactional process.

As one of the pioneering applications of machine learning for quantitative data analysis in applied linguistics, this study investigates the predictive power of linguistic features on English as a Foreign Language (EFL) writing quality. To overcome the limitations of traditional linear regression,  specifically multicollinearity and non-normal data distributions, we employed a Multilayer Perceptron (MLP) neural network to analyse 96 essays from Chinese secondary school students. The model evaluated the relative predictive importance of various lexical, syntactic, fluency, and accuracy features. Demonstrating exceptional fit, the neural network accounted for approximately 80% of the variance in overall writing scores with no evidence of overfitting. Sensitivity analyses revealed that lexical diversity was the most robust predictor of writing quality (100% normalised importance), followed by grammatical accuracy (60.4%). Conversely, syntactic complexity, text length, and lexical sophistication exhibited comparatively minimal influence. These findings underscore the critical role of vocabulary diversity in evaluating adolescent EFL writing, while successfully establishing neural networks as a powerful, innovative methodological tool for future applied linguistic research.

This article describes what could be in a course in training learners how to learn. It covers four strategies, repeated spaced retrieval, learning though use, deliberate learning, and mnemonic devices. Although the main focus is on language learning, the strategies can also be applied well beyond that, and examples are provided from learning vocabulary (learning items), learning economics (subject/content knowledge), and learning to drive a car (skill). The strategies are based on three principles of learning – focus of attention, quantity of attention and quality of attention. The article includes detailed instructions and tasks covering learning, application and analysis for running a workshop on learning how learn that can be used for a short intensive course or for including as a regular part of a teaching program. The tasks are ready to use with possible answers provided.

At one point considered to have been the most densely populated place on earth, nearly thirty years after its demolition, Hong Kong's Walled City of Kowloon continues to be a topic of significant cultural memorialization and lively academic discussion. Yet despite the researchers' best efforts, its origin and early history continue to be riddled with fundamental inaccuracies. Tracing back to key primary sources, this article upholds that the Walled City was selectively built by the Chinese on empty government land between late 1846 and 1847, systematically arguing that it was therefore not an architectural palimpsest with an undue “centuries-old” history. Seeking to bring clarity as to “why” predominant academic discourses to date have differed, the article scrutinizes two of the most influential published narratives on the topic by Elizabeth Sinn and Julia Wilkinson, demonstrating how a fundamental lack of integration of key primary sources has resulted in the current state of knowledge. Closely aligned with and committed to efforts to decolonize the historical record, the article also proposes that such sources should be made more accessible for proper study, in the hope that narratives about the Walled City's origin and early history take account of the perspectives of those who built it.

普通話句末“的”的句法功能問題受到學界的廣泛討論。粵西廉江、茂名方言句末“的”形式為“個[kɔ³³]”,相當於普通話的“的₃”,且與“的₃”發展呈現高度平行。粵西方言句末“個”具有二重性,既能夠分析作結構助詞,又帶有廣義的傳信語氣,在表達斷言 確認時,句末可以出現兩個“個”。兩個“個”共現正是“個”由結構助詞演變為語氣助 詞的中間階段,共現現象不僅存在於粵西粵方言,還見於江西安遠（孔田）客家方言。這 一方言事實坐實了“結構助詞 > 語氣助詞”的演變過程,為普通話“的”等相關問題的討 論提供了重要線索。

嚴州（建德）方言的詞素“把”（po55）既可作為給予動詞亦可作為賓語標記使用（曹志耘 2017）。前一種情況下,po55相當於普通話中的三價動詞“給”,其中物體轉移的方式在語義上是非確定的,而後一種情況下,po55用於標記其後名詞的不同題元角色,類似普通話的“把”。Po55的多功能性源於Chappell（2007）所定義的語法化路徑,即從給予/幫助類動詞到賓語標記。通過與“把”及“把”字句的深入對比,我們認為po55後名詞必須經受特定形式的影響,並且總是與因動詞短語所示事件的實現而引出的特定結果狀態相聯繫。在Randall（2010）連接理論框架的基礎上,我們提出了對po55的基於致使性的統一處理。這使我們最終有理由相信,致使性構成了po55內在概念結構的核心部分,且po55詞彙化為嚴格意義上的致使單位,儘管它具有語法上的雙重特性。

龔岩（生於1974 年）因其「太極合一靈修」而成為當今聞名全中國的靈修導師。受過其靈修培育的修女、修生、神父及平信徒人數之多,令人驚歎。這是一種結合現代心理學與中國傳統太極哲學概念的靈修方式。「太極合一靈修」也為聖三論、基督論和神學人學提供了新的詮釋。在本文中,我們將介紹龔岩如何把中國太極哲學、現代心理學、神學和基督徒靈修銜接起來,以系統化其靈修,從而分析他的創新理論和實踐對基督徒神學和靈修傳統有何啟發。