在人工智慧快速發展以及數位媒介重塑知識生產的環境下，外語文教育持續轉向素養導向的時代課題。本期學報以「跨域連線・AI 對話・永續學習：外語文教育的素養導向實踐」為專題主軸，共收錄七篇論文，其中五篇列為專題論文，分別從翻譯研究、法語學習與義大利語寫作評量探討人工智慧與外語文教育之關係；另收兩篇一般論文，內容涵蓋生成式 AI 融入西語多媒體創作、Spatial 元宇宙平台應用於日語自主學習，以及英美文學與拉丁美洲後殖民文學研究。整體而言，本期論文橫跨法語、義大利語、西班牙語、日語、翻譯研究、英美文學與拉丁美洲文學等領域，既呈現外語教學面對 AI 與新興科技時的實踐探索，也展現文學與文化研究在變動世界中對人性、歷史、信仰與權力結構的深層反思。
專題論文首先是吳敏華教授的〈人工翻譯與神經機器翻譯：〈米哈波橋〉的中文再現〉，本文以阿波里奈爾〈米哈波橋〉的中譯為個案，比較神經機器翻譯與人工翻譯在詩歌再現上的差異，指出機器翻譯雖具有速度與效率，卻難以充分掌握詩歌中的韻律、意象、文化脈絡與審美深度。此研究提醒我們，在人工智慧日益精進的今日，人類譯者的文學感受、詮釋判斷與創造能力，仍是跨語言文化傳遞中不可取代的核心。
薛芬妮教授的〈人類與人工中介：ChatGPT 作為法語作為外語（FLE）中的「更有知識的他者」〉則從維高斯基的近側發展區(ZPD)理論與「更有能力他者」概念出發，探討 ChatGPT 在法語書面產出學習中作為認知中介的可能。研究指出，AI 可促進反思性知識共構與學習者自主性，擴展學生的學習支持系統；然而，教師仍然是具有意圖性、社會情意支持與批判引導功能的重要中介者。此文為混合學習情境中人類教師與人工智慧之間的角色分工，提出兼具理論與實踐意義的觀察。
施喬佳教授的〈生成式 AI 與人類教師於義大利文寫作評量之比較研究：可信度、一致性和教學法相容度〉進一步將焦點放在外語寫作評量。本文比較 ChatGPT、Gemini、Claude 與人類教師在義大利文作文評分與回饋上的表現，指出生成式 AI 尚無法取代教師的專業判斷，但能在評分依據、診斷性回饋與錯誤模式辨識方面展現輔助潛力。此研究不僅回應教師工作負荷日益增加的現場困境，也為 AI 如何成為可信賴的教學夥伴提供具體參照。
簡藝珊教授的〈生成式 AI 多媒體創作專案於西語課程之實踐研究〉以專案式學習(PBL)為架構，探討生成式 AI 融入西班牙語多媒體創作任務的課堂實踐。研究強調，AI 的價值不在於取代學生的語言學習歷程，而在於透過「草稿—回饋—修訂」歷程中輔助語言形式調整與文法重點補強，再透過小組的文法解說表與同儕互評過程中觀察語言能力與素養行為之建構歷程；呈現PBL結構下結合AI工具之多媒體任務中如何實施，以及提升學生在創作過程中展現數位素養、合作溝通、自主學習與跨領域整合能力。
邱明麗教授的〈以 Spatial 元宇宙平台作為日語自主學習運用之探討〉則將外語學習延伸至沉浸式虛擬空間。作者將 Spatial 平台融入大學基礎日語與會話課程活動，設計「旅遊照片分享」「我喜歡的歌曲」「偏鄉教材製作」等任務導向主題，引導學生在 3D 虛擬空間中以日語進行內容建構、文化詮釋與成果展示。研究指出，Spatial 平台結合任務設計與教學鷹架後，可促進日語學習者在目標設定、資料蒐集、時間管理、歷程監控與自我反思中的自主學習(Self-Directed Learning, SDL)與自我調節學習(Self-Regulated Learning, SRL)，亦提醒教師須留意平台操作、素材整合與認知負荷等挑戰。
一般主題部分中，學者李寬〈對不完美的反認識〉轉向福克納小說中的美國南方白人男性形象，透過「不笑」與追求完美的情感依附，分析角色如何拒絕認識內戰後南方社會的變遷與父權家族的衰敗。本文以笑聲理論切入現代美國小說，展現文學研究對情感、權力與歷史創傷的細緻詮釋。
學者秦佾〈後殖民場域中的天主教神職人員——以阿奎達的《山上的狐狸與山下的狐狸》為例〉則結合社會學、神學與後殖民理論，分析神職人員如何在神聖與世俗、中心與邊緣之間扮演中介角色，同時也揭示天主教會歷史共業與本土宇宙觀之間難以化解的張力。
綜觀本期論文，AI 不再只是工具，文學也不只是文本；兩者皆成為我們理解人類學習、溝通、記憶與倫理責任的重要入口。從翻譯、寫作評量、生成式創作到元宇宙學習，從美國南方到拉丁美洲後殖民場域，本期作者共同展現外語文研究在跨域連線中的深度與韌性。謹向所有作者、審查委員與編輯團隊致上誠摯謝意，並期待本期成果能持續啟發外語文教育與人文研究在 AI 時代中的永續對話。
Against the backdrop of the rapid development of artificial intelligence and the reshaping of knowledge production through digital media, foreign language and literature education continues to face the timely challenge of shifting toward competency-oriented practice. With the special theme “Cross-Disciplinary Connections, AI Dialogue, and Sustainable Learning: Competency-Oriented Practices in Foreign Language and Literature Education,” this issue includes seven articles. Among them, five are featured articles that examine the relationship between artificial intelligence and foreign language education from the perspectives of translation studies, French language learning, Italian writing assessment, Spanish multimedia creation, and Japanese autonomous learning. The remaining two general articles cover research on British and American literature as well as Latin American postcolonial literature. Overall, the articles in this issue span the fields of French, Italian, Spanish, Japanese, translation studies, British and American literature, and Latin American literature. They not only present practical explorations of foreign language teaching in response to AI and emerging technologies, but also demonstrate how literary and cultural studies offer profound reflections on humanity, history, faith, and structures of power in a changing world.
The first featured article is Professor Min-Hua Wu’s “Human Translation vs. Neural Machine Translation: Rendering ‘Le Pont Mirabeau’ into Chinese.” Taking the Chinese translation of Apollinaire’s “Le Pont Mirabeau” as a case study, this article compares the differences between neural machine translation and human translation in the representation of poetry. It points out that while machine translation offers speed and efficiency, it remains difficult for it to fully grasp the rhythm, imagery, cultural context, and aesthetic depth of poetry. This study reminds us that, in an age when artificial intelligence is becoming increasingly advanced, the literary sensitivity, interpretive judgment, and creative capacity of human translators remain an irreplaceable core in cross-linguistic and cross-cultural transmission.
Professor Laurette Guinot Fanny’s “Human and Artificial Mediation: ChatGPT as a ‘More Knowledgeable Other’ in French as a Foreign Language (FLE)” begins with Vygotsky’s theory of the Zone of Proximal Development (ZPD) and the concept of the “More Knowledgeable Other” to explore the possibility of ChatGPT serving as a cognitive mediator in the learning of written production in French. The study shows that AI can facilitate reflective knowledge co-construction and learner autonomy, thereby expanding students’ learning support systems. However, teachers remain crucial mediators with intentionality, socio-affective support, and critical guidance. This article offers observations of both theoretical and practical significance regarding the division of roles between human teachers and artificial intelligence in hybrid learning contexts.
Professor Giorgia Sfriso’s “Human-GenAI Alignment in Italian FL Writing Assessment: Reliability, Consistency, and Pedagogical Implications” further shifts the focus to foreign language writing assessment. This article compares the performance of ChatGPT, Gemini, Claude, and human teachers in scoring and providing feedback on Italian compositions. It points out that generative AI is not yet able to replace teachers’ professional judgment, but it does demonstrate auxiliary potential in providing scoring rationales, diagnostic feedback, and identifying error patterns. This study not only responds to the increasingly heavy workload faced by teachers in actual teaching contexts, but also provides concrete references for how AI may become a trustworthy pedagogical partner.
Professor Yi-shan Chien’s “A Practice-Based Study on Integrating a Generative AI–Assisted Multimedia Creation Project into a Spanish Course” adopts Project-Based Learning (PBL) as its framework to investigate classroom practices involving the integration of generative AI into Spanish multimedia creation tasks. The study emphasizes that the value of AI does not lie in replacing students’ language learning processes. Rather, through the “draft–feedback–revision” process, AI assists students in adjusting linguistic forms and strengthening grammar focus. In addition, through group grammar explanation sheets and peer assessment, the study observes the construction of language ability and competency-oriented learning behaviors. It presents how multimedia tasks combining AI tools within a PBL structure can be implemented and how such tasks can enhance students’ digital literacy, collaborative communication, autonomous learning, and interdisciplinary integration during the creative process.
Professor Ming-Li Chiu’s “An Exploration of the Metaverse Platform Spatial as a Tool for Self-Directed Japanese Learning” extends foreign language learning into immersive virtual spaces. The author integrates the Spatial platform into activities in university-level basic Japanese and conversation courses, designing task-oriented themes such as “travel photo sharing,” “my favorite song,” and “teaching material development for rural education.” Students are guided to construct content, interpret culture, and present outcomes in Japanese within a 3D virtual space. The study indicates that when the Spatial platform is combined with task design and pedagogical scaffolding, it can promote Japanese learners’ Self-Directed Learning (SDL) and Self-Regulated Learning (SRL) in goal setting, information gathering, time management, process monitoring, and self-reflection. It also reminds teachers to pay attention to challenges such as platform operation, material integration, and cognitive load.
In the section of general articles, scholar Kuan Li’s “Unknowing Imperfections” turns to the representation of Southern white male figures in William Faulkner’s novels. Through the emotional attachment between “incapacity to laugh” and the pursuit of perfection, the article analyzes how these characters refuse to recognize the social changes of the postbellum South and the decline of patriarchal families. By approaching modern American fiction through theories of laughter, this article demonstrates a nuanced literary interpretation of emotion, power, and historical trauma.
Scholar Yi Chin’s “Catholic Clergy in Postcolonial Contexts: A Case Study of Arguedas’s The Foxes from Above and the Foxes from Below” combines sociology, theology, and postcolonial theory to analyze how Catholic clergy serve as mediators between the sacred and the secular, and between the center and the margins. At the same time, the article reveals the unresolved tensions between the historical complicity of the Catholic Church and indigenous cosmologies.
Taken together, the articles in this issue show that AI is no longer merely a tool, and literature is no longer merely a text. Both have become important entry points for understanding human learning, communication, memory, and ethical responsibility. From translation, writing assessment, and generative creation to metaverse learning, and from the American South to Latin American postcolonial contexts, the contributors to this issue collectively demonstrate the depth and resilience of foreign language and literature research in cross-disciplinary connection. I would like to express my sincere gratitude to all authors, reviewers, and members of the editorial team, and I hope that the work presented in this issue will continue to inspire sustainable dialogue in foreign language education and humanities research in the age of AI.