Programme

Abstract

A recent empirical study found that the European Commission (EC)’s public communications uses very complex language, specialized jargon, and a nominal style that obfuscates political action, excluding many people from democratic processes due to insufficiently accessible information. Accessibility is an underestimated asset for democratic participation, consequently, a lack of accessibility to information directly leads to the exclusion of several people from democracy.

Persons who encounter difficulties making sense of content are a diverse group of individuals with varying degrees of reading, writing, and understanding abilities, including people with intellectual disabilities, older adults who struggle with long and complicated documents, or low-literacy people. The inclusion of people with language barriers in democracy is a sine qua non to guarantee full legitimate democratic outcomes.

The EC has recently imposed new requirements across the European Union regarding the accessibility of some products and services for people with disabilities following the ratification by the EU of the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD). One way to guarantee that people have equal access to information is through the production of information in some form of easy language (e.g., plain language or easy-to-read).

However, making information accessible to all requires considerable efforts and expertise, which sometimes limits the production of or the translation into easy language.

In this talk I will describe the work carried out in the context of the Horizon Project iDEM to facilitate democratic participation for people with language barriers. The talk will highlight the creation of language resources within the project as well as the implementation of text simplification services integrated into a mobile application to facilitate participation in deliberative, democratic processes.

Speaker biography

Horacio Saggion is chair in Computer Science and Artificial Intelligence at the Department of Information and Communication Technologies, Universitat Pompeu Fabra (UPF), Barcelona, Spain.

He was appointed by UPF in 2010 as a Ramón y Cajal research fellow from the Spanish Ministry of Science and he was promoted to full professor in 2021.

Horacio is director of the TALN Natural Language Processing Group where he works in several areas of Natural Language Processing (NLP) automatic text summarization, text simplification, NLP for Sign Languages, information extraction, figurative language, sentiment analysis and related topics. His work combines symbolic and statistical techniques.

Since his arrival he has obtained funding to carry out his research from different national and international organizations (Ministry of Science, Spanish National  Research  Agency, FP7, CIP, H2020, Horizon Europe). He also collaborates with the industry through contracts, projects and Doctoral PhDs.

Horacio has published over 200 works in journals and peer-reviewed international conferences having also organized several workshops in the area of NLP for accessibility.

He is the project coordinator of the Horizon Europe funded project iDEM - Innovative and Inclusive Democratic Spaces for Deliberation and Participation - and Principal Investigator of the IDEAL project - Inclusive Democratic Engagement and Language Technologies in Europe, and co-PI in the AI-BOOST project.

Abstract

Traditionally when lawyers drafted documents, every word was chosen with intent, tested against precedent and weighted for consequence. In law, this relationship between word and consequence is meant to be controlled: this word, in this context, under this jurisdiction, does this legal work.

When legal language is precise, it is often because someone has previously fought over every word. Sometimes lawyers employ deliberate vagueness or ambiguity because they too serve a purpose. Generative AI, though, operates on a different logic. It does not (and cannot) know the law. It predicts language. To use a gambling metaphor, it cannot see its own hand. It has no idea whether it is holding a winner or a busted flush. Winning or losing matters because legal language is not low-stakes text. How it is drafted, translated, summarised or filed can affect someone's claim, defence, contract, immigration status, benefit entitlement, financial exposure or right of appeal.

This talk asks what happens when the deterministic language of law meets the probabilistic language of generative AI, and why translation is where that collision becomes hardest to ignore. Drawing on legal theory, translation studies, knowledge communication theory, risk theory and human-AI interaction research, it examines questions such as how and why we might be ‘gambling’ with the language of the law, why we might not perceive it as risk, and why we might continue even when we do. And what the stakes are for legal language, for law as a system, and for the people whose lives depend on getting it right.

Speaker biography

John O’Shea LL.B (Hons.), LL.M has 30 years of experience translating legal documents from Greek into English. He works with international investors, multinationals, and leading law firms in Greece and Cyprus on major court cases and key transactions.

His research interests include legal risk and translation, translator liability, neural machine translation and genAI in legal contexts, and Greek–English legal translation, human-AI interactions, and high-risk AI systems and their use in law. His work has been presented at international conferences and published in journals and book chapters.

He is involved in a long-term project translating Greece’s Codes into English, including the codes of criminal and civil procedure and the code of private maritime law. He has taught legal translation courses, delivered CPD internationally and remotely, and is currently Chairperson of FIT Europe.

Abstract

For several academic disciplines, AI is still a disruptive force. For Translation Studies, it is a regular visitor in our homes since, at least, 2017. Some of us are accustomed to it, but when we interact with it, it still causes wonder, estrangement and stress. In this panel, we want to reflect on what is happening to the research that we do in Translation and Interpreting Studies after we opened our doors to AI.

There are many questions to ask regarding the transformative influence of AI on research. Is AI an adequate instrument to be used in Translation and Interpreting Studies research? Or is it similar to a massive and heavy telescope with tiny defects in its lenses? Are we promising the discovery of new planets which disappear after the lens is polished? Is it an instrument with hefty costs, which requires careful handling, demands constant attention and powerful skills of rigorous interpretation of the data it produces? This instrument became itself an object of research. AI produces an immense amount of language data that researchers are increasingly focusing on and analysing. Furthermore, AI is also assisting us with the outputs of our research. A major question here is: can we still research anything that is related to language and translation that is not somehow under the influence of AI?

This panel will explore AI not simply as a new technological development, but as a critical challenge and an opportunity for Translation Studies research. Our discussion will focus on how research in our field can best serve people and society in the age of AI; how it can continue to contribute meaningfully to the advancement of knowledge in times of transition and transformation; how translation and interpreting researchers can engage with AI in ways that are both intellectually rigorous and socially responsible; and how these considerations can inform the research perspectives and methodologies that we adopt in studying AI and its impact on translation and interpreting practices.

Panellists

• Sabine Braun (CTS Surrey)
• Miguel-Ángel Jimenez-Crespo (Univ. Rutgers – US)
• Omri Asscher (Bar-Ilan University – Israel)
• Sara Palmer (CTS Surrey)
• Christopher Mellinger (UNC Charlotte – US).

Chair

Félix do Carmo (CTS Surrey).

Abstract

Translation and interpreting research is under increasing pressure to demonstrate not only academic excellence, but also social relevance, professional impact and public value. At the same time, the environments in which multilingual and accessible communication take place are being transformed at extraordinary speed by platformisation, automation and AI, alongside changing labour conditions and shifting expectations around language, accessibility and global communication. In this context, important questions emerge:  What kinds of research does the language services sector need today and in the future? Which research questions matter most in the current moment? And how can academic research remain both intellectually rigorous and grounded in professional realities while retaining its critical and independent voice?

This panel brings together voices from academia, professional practice and the wider language services industry to reflect on these questions. Rather than treating research as something produced only within universities and later transferred into practice, the panel will conceptualise research as a shared space of dialogue, negotiation and mutual influence between researchers, practitioners, businesses, institutions and language service users.

The discussion will also examine how research can contribute to demonstrating, articulating and advancing the economic, social and cultural value of translation, interpreting and other practices of multilingual and accessible communication in a rapidly changing world. In doing so, we will ask where current research is succeeding with this and which aspects urgently require greater scholarly attention, what kinds of evidence are needed, and which forms of expertise and experience may currently be overlooked. We will also revisit the broader question of how different stakeholders can work together more effectively in shaping appropriate research agendas. And, ultimately, we will consider what kinds of research society needs from translation and interpreting studies in the years ahead.

Panellists

• Dom Hebblethwaite, Head of Membership & Ventures, Chartered Institute of Linguists
• John Anthony O'Shea, Founder of Jurtrans, Chairperson of FIT-Europe, member of EU's Language Industry Expert Group
• Carla Para Escartin, AI Data Research Director, Toppan Digital Language
• Sara Robertson, Chief Executive, Institute of Translation and Interpreting
• Félix do Carmo, Senior Lecturer in Translation and Natural Language Processing, Lead of Professional Development, Centre for Translation Studies, University of Surrey
• Raisa McNab, Chief Executive, Association of Translation Companies.

Chair

Sabine Braun, Director, Centre for Translation Studies, Co-Director, Surrey Institute for People-Centred Artificial Intelligence.