Critical language barriers: Understanding the role of online translation tools in high-stakes professional sectors

关键的语言障碍：了解在线翻译工具在高风险专业领域的作用

• 批准号: AH/X006506/1
• 负责人: Lucas Nunes Vieira
• 金额: $ 25.34万
• 依托单位: University of Bristol
• 依托单位国家: 英国
• 项目类别: Fellowship
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=AH%2FX006506%2F1
• 关键词: Critical; language; barriers; Understanding; role

Online machine translation tools like Google Translate are widely available and hugely popular. These tools are also socially and ethically complex. On the one hand, they support linguistic diversity and increase access to information. On the other, these tools can be inaccurate, so they are inherently risky. The low cost and convenience of this technology are nevertheless highly appealing, so machine translation is now present even in high-stakes contexts. Official reports show that these tools are used to interview refugees, to communicate with patients, and to obtain consent for police searches. The possible negative consequences of using the technology in these situations can vary from information security threats to malpractice or miscarriages of justice. Most human-centred machine translation research to date has overlooked these potentially serious implications and tended to focus on how the technology is used in language services or in academic settings. This project will help to address a growing need for research focused on situations where non-linguists and non-students use machine translation to overcome a language barrier. The project's aim is to improve the understanding of machine translation's social implications in high-stakes professional sectors with a view to shaping guidance and public policy. Machine translation use will be situated in the context of broader theoretical traditions involving intercultural digital ethics, technological mediation, and the social construction of technology. The project will draw on relevant concepts from these frameworks and address two empirical questions. The first question will explore how professionals who work in high-stakes environments characterise machine translation and its use. Lawyers, doctors, police officers, civil servants and a range of other professionals will provide important accounts of how they deal with language barriers in their work and of what they see as the challenges and opportunities of machine translation. The second question will examine the implications of machine translation use for synchronous (i.e., live) ethically charged conversations. It will explore multimodal methods of interacting with the technology (e.g., with speech-to-text and text-to-speech conversions) and their effects on the communication and the interlocutors. The analysis will interrogate how machine translation can change the understanding of translation (written) and interpreting (spoken) and explore points of convergence between these two fields. The methods include both quantitative and qualitative techniques, and will break new ground in translation studies in terms of sampling strategy and research design. The UK's Chartered Institute of Linguists is a Partner on the project. A Royal Charter body of over 7,000 members, the Chartered Institute of Linguists will act as the incoming secretariat of the All-Party Parliamentary Group on Modern Languages. Its mission is to promote the value of languages and language skills in the public interest. Our partnership will play an important role in achieving the project's aim. Together, we will work to co-produce outputs and communicate the project's results to policy making and influencing bodies including Parliament, think tanks, government agencies, and trade unions. The project's outputs include open datasets, a policy advisory report as well as a book on ethical implications of machine translation beyond use-case settings in education and the language industry.Machine translation is a tool with significant potential. Our intention with the project is ultimately to help this potential to be realised in ways that enrich cross-cultural communication, promote social justice, and protect individuals' right to information, which all require an evidence-based understanding of the role this technology plays in society.

在线机器翻译工具（例如Google Translate）广泛可用且非常受欢迎。这些工具在社会和道德上也很复杂。一方面，它们支持语言多样性并增加获取信息。另一方面，这些工具可能是不准确的，因此它们本质上是风险的。然而，这项技术的低成本和便利性具有很高的吸引力，因此即使在高风险环境中，机器翻译也存在。官方报告显示，这些工具用于采访难民，与患者进行交流并获得警察搜查的同意。在这些情况下使用技术的可能负面后果可能会因信息安全威胁对渎职或司法流产而异。迄今为止，大多数以人为本的机器翻译研究忽略了这些潜在的严重含义，并倾向于专注于该技术如何用于语言服务或学术环境中。该项目将有助于满足不断增长的研究需求，这些研究的重点是非语言学家和非学生使用机器翻译来克服语言障碍的情况。该项目的目的是提高对机器翻译在高风险专业领域的社会影响的理解，以塑造指导和公共政策。机器翻译使用将位于涉及跨文化数字伦理，技术调解和技术的社会建构的更广泛的理论传统的背景下。该项目将利用这些框架中的相关概念，并解决两个经验问题。第一个问题将探讨如何在高风险环境中工作的专业人员表征机器翻译及其使用。律师，医生，警察，公务员和其他专业人员将提供重要的说明，以应对他们工作中的语言障碍以及他们认为是机器翻译的挑战和机遇。第二个问题将研究机器翻译使用对同步（即实时）道德充电对话的含义。它将探索与技术交互的多模式方法（例如，通过语音到文本和文本到语音转换）及其对交流和对话者的影响。该分析将询问机器翻译如何改变对翻译（书面）和解释（口语）的理解，并探索这两个字段之间的收敛点。这些方法既包括定量和定性技术，又将在抽样策略和研究设计方面打破翻译研究的新基础。英国特许语言学家是该项目的合作伙伴。宪章学院是由7,000多名成员组成的皇家宪章机构，将担任现代语言全党议会组织的即将上任的秘书处。它的使命是为了公共利益而提高语言和语言技能的价值。我们的合作伙伴关系将在实现该项目的目标方面发挥重要作用。我们将共同致力于共同生产产出，并将项目的结果传达给政策制定和影响机构，包括议会，智囊团，政府机构和工会。该项目的输出包括开放数据集，策略咨询报告以及一本关于教育和语言行业用例设置的机器翻译的道德含义的书。MachineTranslation是具有巨大潜力的工具。我们对该项目的意图最终是为了帮助以丰富跨文化交流，促进社会正义和保护个人的信息权的方式实现这一潜力，这所有这些都需要基于证据的信息对这项技术在社会中扮演的角色有基于证据的理解。