人机协同翻译中学习者认知负荷的动态演变研究——基于汉语块状语英译的纵向追踪分析
A Study on the Dynamic Evolution of Learners’ Cognitive Load in Human-Machine Collaborative Translation—A Longitudinal Analysis Based on the English Translation of Chinese Chunk Expressions

本研究基于认知负荷理论，运用纵向追踪实验与多元数据采集方法，系统考察高校翻译学习者在人机协同环境下处理汉语块状语英译时的认知负荷变化轨迹及策略调适机制。研究发现：(1) 人机协同模式显著降低了学习者的内在认知负荷，但初期引发外在认知负荷的短暂升高；(2) 学习者经历“机械依赖–批判反思–策略内化”三阶段的发展历程，策略使用频次与使用效能呈现倒U型演进曲线；(3) 元认知调控能力的发展是认知负荷优化的关键，具备较强元认知监控的学习者能提前2~3周完成策略内化；(4) 基于认知支架设计的渐进式教学干预能有效促进学习者从表层模仿到深层理解的跨越。本研究在理论层面丰富了翻译教学中认知负荷的动态理论，在实践层面为高校翻译课程的人机协同教学设计提供了发展性评估框架和阶段性干预路径，对培养人工智能时代的翻译人才具有参考价值。
Grounded in cognitive load theory, this study uses longitudinal tracking and multi-method data collection to investigate how translation students adapt their cognitive load and strategies when translating Chinese chunks into English in a human-machine collaborative setting. Key findings show that: (1) Human-machine collaboration significantly lowers intrinsic cognitive load but causes a brief rise in extraneous load initially; (2) Learners progress through three stages-mechanical dependence, critical reflection, and strategy internalization with strategy use forming an inverted U-shaped curve; (3) Stronger metacognitive skills allow learners to internalize strategies 2~3 weeks faster; (4) Scaffolded teaching interventions help learners move from surface imitation to deep understanding. The study contributes to cognitive load dynamics in translation education and offers a practical framework for designing collaborative human-machine translation courses.