‘Something old, something new, something borrowed’: Past, present, and future applications of Translation Quality Assessment

The advancements in computing and artificial intelligence (AI) alongside the growing need for rapid and effective communication on a global scale are reshaping the landscape of translation, impacting the process, the product, and our understanding of translation. This is especially true for hybrid scenarios, which involve flexible equivalence between the source and target texts, e.g., in the translation of news, popularisation of scientific texts, transediting, and transcreation. These changes also affect existing Translation Quality Assessment (TQA) paradigms, whose scope, aims, and applications should evolve alongside these developments. This paper engages in a reflective analysis of the historical evolution, present state, and potential future professional and didactic applications of TQA. After reviewing past paradigms, the paper looks into its contemporary applications within professional and educational settings. Special attention is given to the ISO 5060:2024 and its role in shaping TQA, as well as to the integration of computer-assisted revision in translator training. The paper then discusses how current TQA approaches can be adapted to hybrid settings and how they can expand their scope and applications to interact with new technologies. This is illustrated through two applied scenarios, one regarding the summary translation of a medical academic paper for non-expert readers, and the other involving the analysis of machine translation (MT) outputs to assess their quality for post-editing purposes, improving them through AI. The paper thus exemplifies how TQA can be adapted to hybrid translation scenarios and shows how error analysis can serve not only to assess MT quality, but also to determine whether and how to improve its output through automation.