Disrupter or enabler? Assessing the impact of using automatic speech recognition technology in interpreter-mediated legal proceedings

Covid-19加速了通信技术的使用。在多语言环境中，这增加了远程口译的使用，口译员通过数字平台与客户连接。虽然远程口译具有挑战性，但数字平台为协助口译员完成任务创造了新的机会。自动语音识别（ASR）被认为是一种最先进的技术，可以支持并彻底改变口译员的实践。ASR可以为口译员提供他们正在听和翻译的演讲的现场文本。这可能对法律环境中的口译员有实质性的好处。这是因为现在的法律程序通常需要远程口译，而在法庭和类似环境中的沟通可能高度不可预测，这对法律口译员构成了挑战，往往缺乏培训和准备时间。当远程工作时，口译员可能会面临额外的问题，例如音质差。然而，远程法律口译的研究尚未涉及到自动语音识别的使用。为了弥补现有研究的这一空白，本研究旨在评估ASR对口译员远程法律口译绩效的影响。具体目标是：首先，开发一个对ASR使用敏感的远程法律口译质量评估模型；其次，探讨自动语音识别在多大程度上提高远程法律口译员的工作绩效；第三，调查口译员如何适应这种新技术。为此目的，设计了一个两阶段的方法。首先，将进行全面的文献综述，以适应现有的质量评估方法来评估口译员在有和没有ASR支持的远程法律口译中的表现。其次，将通过多种资源收集经验数据，包括模拟法庭听证会和与口译员的反思会议。将对通过这些资源收集的数量和质量数据进行交叉分析，以得出关于第二和第三个目标的结论。

Covid-19 has accelerated the use of communication technologies. In multilingual settings, this has increased the use of remote interpreting, whereby interpreters are connected to clients via digital platforms. Whilst remote interpreting is challenging, digital platforms create new opportunities to assist interpreters with their task. Automatic speech recognition (ASR) has been proposed as a state-of-the-art-technology that might support and indeed revolutionise the practice of interpreters. ASR can provide interpreters with a live transcript of the speech they are listening to and interpreting. This could have substantial benefits for interpreters in the legal setting. This is because legal proceedings now often require remote interpreting while communications in court and similar settings can be highly unpredictable, posing challenges for legal interpreters often lacking in training and preparation time. When working remotely, interpreters may face additional problems, such as poor sound quality. However, the research on remote legal interpreting has not yet investigated the use of ASR. To address this gap in the existing research, this study aims to assess the impact of ASR on interpreters' performance in remote legal interpreting. The specific objectives are: firstly, to develop a quality assessment model for remote legal interpreting that is sensitive to ASR use; secondly, to explore to what extent ASR can improve the performance of remote legal interpreters; thirdly, to investigate how interpreters adapt to this new technology. To this end, a two-stage methodology has been designed. Firstly, a comprehensive literature review will be conducted to adapt existing quality assessment approaches to evaluate interpreters' performance in remote legal interpreting, with and without ASR support. Secondly, empirical data will be collected through multiple resources, including simulated court hearings and reflective sessions with interpreters. Quantitative and qualitative data gathered through these resources will be cross-analysed to draw conclusions regarding the second and third objectives.