2024年度，人工智能与语言认知实验室在学校的大力支持下，取得了丰硕的科研成果，涵盖了多个领域，包括语言认知、人工智能在语言教育中的应用、翻译研究、以及多模态认知过程分析等。这些成果主要发表在Metaphor and Symbol、European Journal of Education、International Journal of Bilingualism、《外语教学与研究》、《外语教育研究前沿》等国际和国内重要期刊上，推动了语言学、认知科学和人工智能技术的深度融合，为国内外学术界的研究提供了新的视角和方法。
以下是2024年度高水平成果及其简介（按作音序排列）：
[1] 齐涛云 & 孙曙光. (2024). “师生合作评价”同传教学模式的实施原则与例析. 外语教育研究前沿(04), 65-72+96. （CSSCI）
摘要：师生合作评价（teacher-student collaborative assessment，简称 TSCA）同传教学模式对于提升大班同声传译教学效果具有重要作用。本文是TSCA同传教学模式建构的后续研究，聚焦该模式的微观操作层面：先基于教学实践论述了TSCA课前、课中和课后三阶段实施原则与同传课程的融合方式；再通过教学案例展示教学活动的设计与实施。
[2] 齐涛云 & 杨承淑. (2024). 同传认知过程研究的多模态路径. 语言与翻译, (03), 57-66. （CSSCI扩）
摘要：认知一直都是同传研究的热点议题，基于多模态语料库可以对同传认知的微观过程进行有效探究。本文试图对 同传认知过程研究的多模态分析方法进行初步构建。在该方法的操作过程中，首先要在语料中寻找认知标记， 然后分别从串行和并行两个方向通过多模态证据对认知标记产生的认知过程进行推断。串行和并行方向分析 分别依据Levelt提出的言语产出蓝图和Gile提出的同传认知负荷模型。利用本文所建构的研究方法可以揭示 同传策略或同传失误等背后的同传特有认知活动，这不仅可以进一步提高同传语料库描写和解释的充分性，对同传实践和同传教学也可以提供积极的启示。
[3] 孙中洋, & 张晓东. (2025). Overall inhibitory control modulated the comprehension of Chinese idiomatic items where minor glyph message pertains. Metaphor and Symbol. （SSCI）
Abstract: Idioms are the unity of figurative and formulaic expressions, with a variety of creative variants used in daily life. This study explored how Chinese readers made sense of Chinese idioms and their variants, with possible constraints from their strengths in overall inhibitory control (OIC). Our new attempt at a comprehensive measure of inhibitory control managed to predict participants' self-paced reading behavior in the context of idioms in Chinese script. Those with poorer OIC were assumed to show more difficulties in inhibiting the literal meanings of Chinese idioms as supportive contexts unfolded. Such constraints from OIC also resurfaced during the comprehension of Chinese variants that were likely treated as novel idioms.We thereby assumed a hybrid route for the comprehension of Chinese idiomatic expressions. Nonetheless, idiom familiarity or decomposability turned out to exert rather meager influence in this study.The current studyenriches the individual-difference approach to idiom comprehensionby further expounding how its account varies in an ideographicwriting system characterized by symbolic glyphs.
[4] 王敬, 司显柱 & 王丹旎. (2024). 政治文献翻译中MTI学生与专业译者的语言复杂度对比研究. 外语教学与研究, (06), 924-936+961. （CSSCI）
摘要：本研究基于 MTIC 和 PROC 两个可比语料库，从词汇复杂度和句法复杂度两个维度，探讨在政治文献翻译领域里 MTI 学生译文与专业译者译文的差异。研究发现：1）译者翻译水平是造成译文复杂度差异的关键因素；2）在词汇复杂度方面，MTI 学生译文在词汇难度及词汇多样性的 11 个指标上均显著高于专业译者译文，表明 MTI 学生倾向于使用复杂化和多样化的词汇来增强译文的表层专业性；3）在句法复杂度方面，MTI 学生与专业译者的译文在 T 单位长度、T 单位复杂度以及复杂 T 单位占比 3 个指标上有显著差异。MTI 学生在简化复杂句式和显化逻辑结构方面较专业译者有所不足。这些发现为提升 MTI 学生在政治文献翻译领域的译文质量提供了重要的指导和参考。
[5] 薛欣, 张超, 余琦斌, & 刘丛. (2024). 服务机器人更胜一筹?——餐盘浪费监督力度与服务提供者类型对自助餐顾客响应的影响. 旅游导刊, 8(4), 35.
摘要：餐饮企业监督顾客实现“光盘”，是防止食品浪费的重要举措，但难免会引起消极的顾客响应。在数智化时代，使用服务机器人执行监督职能，是否可以缓和自助餐顾客因餐盘浪费监督而产生的消极响应？本研究基于心智知觉理论，通过预实验和情景实验，在自助餐厅监督顾客餐盘浪费的情境下，探究服务提供者类型（服务机器人vs.人类服务员）与监督力度（大vs.小）对顾客响应的影响及其心理机制。研究结果表明：（1）相较于较大的监督力度，较小的监督力度增加了消极的顾客响应，具体表现为顾客态度更消极、负面口碑行为意愿更高；（2）服务提供者类型与监督力度对顾客响应产生了交互效应，当监督力度较小时，服务机器人相较于服务员减轻了消极的顾客响应，当监督力度较大时，服务机器人减轻消极的顾客响应的效应消失；（3）感知限制性在服务提供者类型和监督力度对顾客响应的交互影响中发挥了中介作用。本研究不仅丰富了数智化技术助推顾客减少餐盘浪费的理论研究，同时也为自助餐企业实施餐盘浪费监督时选择何种服务方式提供了实践启示。
[6] 张晓东, & 王伟. (2024). 词汇知识对笔译能力的影响. 外语教学与研究, 5, 781-792. （CSSCI）
摘要：翻译研究大多从本体出发对词汇进行考察，但关于译者词汇知识对翻译能力 影响的研究较少。鉴于此，本研究以某高校MTI 一年级研究生为受试，通过翻译测试、 译后口头报告、访谈，考察译者的接受性词汇知识和产出性词汇知识对笔译能力的影响 及其加工机制。研究发现：1）译者的接受性词汇知识和产出性词汇知识均能显著预测 其英汉、汉英笔译能力，预测量各不相同；2）在分级词汇中，接受性词汇 10,000、产出性 词汇3000 能够显著预测英汉笔译能力；汉英笔译能力的显著预测变量为产出性学术词 汇、接受性词汇10,000 及产出性词汇 3000；3）英汉笔译能力的直接影响因素为接受性 词汇知识，中介影响因素为产出性词汇知识；对于汉英笔译能力，两种词汇知识的作用机 制则相反。
[7] Zheng, Y., Kirk, I., & Waldie, K. (2024). Oscillatory features of German–English sentence processing: Evidence from an EEG study. International Journal of Bilingualism, 13670069241285332. （SSCI）
Aims and objectives: In this study, we investigated the neural oscillatory patterns occurring during sentence-level overt bilingual processing tasks on 46 German–English bilinguals with the intention of finding out if there were task-specific oscillations for this language pair.
Methodology: Four tasks, including Language 2 (L2) listening, Language 1 (L1) speaking, L2 shadowing (i.e., repeating a segment of an utterance), and backward interpreting (i.e., orally translating L2 sentences to L1), were implemented with the technique of electroencephalography (EEG).
Data and analysis: Data included demographic indices, neural oscillation metrics, and oral production scores from participants. Static spectral analysis, time–frequency analysis, principal components analysis, cross-frequency coupling, and inferential statistics were applied to the data.
Findings: The sensory task and articulatory tasks differed greatly in the delta–theta band and the beta–gamma band in terms of time–frequency dynamics. Analyses also revealed variations in a frontal/prefrontal theta component and a distinct gamma component across the four tasks. Moreover, we found a right fronto-temporal theta–gamma coupling unique to backward interpreting.
Originality: The task-wise comparison experiment paradigm has rarely been implemented in neuroimaging research on language, especially so in EEG research on bilingualism.
Significance: The findings provide novel insights for the understanding of language conversion mechanism and may benefit the training of professional interpreters.
[8] Zheng, Y., & Zhang, B. (2024). 25-year neuroimaging research on spoken language processing: a bibliometric analysis. Frontiers in Human Neuroscience, 18, 1461505. （SSCI）
Introduction: Spoken language processing is of huge interest to cognitive and neural scientists, as it is the dominant channel for everyday verbal communication. The aim of this study is to depict the dynamics of publications in the field of neuroimaging research on spoken language processing between 2000 and 2024.
Methods: A bibliometric analysis was conducted to probe this particular subject matter based on data retrieved from Web of Science. A total of 8,085 articles were found, which were analyzed together with their authors, journals of publication, citations and countries of origin.
Results: Results showed a steady increase of publication volume and a relatively high academic visibility of this research field indexed by total citations in the first 25 years of the 21st century. Maps of frequent keywords, institutional collaboration network show that cooperations mainly happen between institutions in the United States, the United Kingdom and Germany. Future trends based on burst detection predict that classification, Alzheimer’s disease and oscillations are potential hot topics.
Discussion: Possible reasons for the result include the aging of the population in developed countries, and the rapid growth of artificial intelligence in the past decade. Finally, specific research avenues were proposed which might benefit future studies.
[9] Zhou, C., & Hou, F. (2024). Can AI Empower L2 Education? Exploring Its Influence on the Behavioural, Cognitive and Emotional Engagement of EFL Teachers and Language Learners. European Journal of Education, e12750. （SSCI）
Abstract: Artificial intelligence (AI) is transforming L2 education, yet its specific impacts on English as a foreign language (EFL) teachers and language learners' engagement remain understudied. To address this deficiency, this study, grounded in Fredricks, Blumenfeld, and Paris's (Review of Educational Research, 74, 109) three-dimensional engagement model, explored the impacts of AI on the behavioural, cognitive and emotional engagement of EFL teachers and language learners through semi-structured interviews with 24 EFL teachers and 38 college language learners, followed by a thematic analysis with MAXQDA to uncover the effectiveness of AI. The study found that behavioural engagement showcased the integration of AI tools, highlighting increased frequency of use and their practical applications in enhancing language acquisition tasks. Cognitive engagement was marked by the recognition of AI capacity to augment teaching strategies and learning processes, although it also surfaced concerns about the potential overreliance on technology. Emotional engagement reflected a complex interplay of attitudes, with most informants viewing AI positively but acknowledging concerns about job displacement, and its impacts on emotions of students and teachers as well as the relations between them. The study concluded that while AI held promise for L2 education, the integration must consider its limitations and ethical implications. The research provided valuable insights for educators, learners, technology developers and policymakers, encouraging innovative practices and informed decision-making in L2 education.
人工智能与语言认知实验室
2024年12月26日