摘要
本研究通过实验和访谈,对比职业译员与机器同声传译的质量(忠实性、地道性、可用性)和过程(工作方式、认知限制、问题机制)。研究发现,对于场景变量和源语变量差异较大的两个实验材料,译员与机器同声传译质量有所不同:忠实性上,对材料1机器表现更佳,材料2表现接近;地道性和可用性上,材料1无显著差异,材料2译员表现更佳。译员口译过程中的困难和问题源于认知精力不足和/或认知能力不足,机器口译则主要源于认知能力不足。本文据此认为译员与机器同声传译之间存在较大的合作互补空间,人机协同(机助人译、人助机译)未来可期。
Experimental and interview data were collected to compare the quality(fidelity,idiomaticity,and usability)and processes(operational modes,cognitive constraints,and failure mechanisms)of Chinese-English simultaneous interpreting by professional interpreters and a machine.The results indicate mixed-quality performances in two speech clips in which the situational and linguistic variables distinctly differ.On fidelity,the machine achieved better results in Clip 1 and similar results in Clip 2 as compared to the interpreters.On idiomaticity and usability,there was no significant difference between the machine and the interpreters in Clip 1,whereas the interpreters performed better in Clip 2.The interpreters’cognitive resources and capability deficiency accounted for interpreting difficulties and failures,whereas cognitive capability was the major issue for the machine.The study identified opportunities for mutual learning and collaboration between the machine and interpreters(machine-aided human interpreting and human-aided machine interpreting).
作者
卢信朝
LU Xinchao(Graduate School of Translation and Interpretation,Beijing Foreign Studies University,Beijing 100089,China)
出处
《外语教学与研究》
CSSCI
北大核心
2022年第4期600-610,F0003,共12页
Foreign Language Teaching and Research
基金
北京市社会科学基金项目“多语双向跨模态翻译语料库的研制与应用”(19YYB011)
中央高校基本科研业务费专项资金项目“英汉汉英会议同声传译语料库建设”(2021JS002)
北京外国语大学卓越人才支持计划资助。
