• Angelika Zirker

  Associate Professor of English Literature, University of Tübingen, Germany

• Regine Tobias

  Chief Library Director, University of Tübingen, Germany

  The role of the library at the university

• Sibylle Meissner

  Managing Director, Tübingen School of Education, University of Tübingen, Germany

  Tübingen - Our profile and the role of collections in academia

• Anja Schreiber, Alex Klein

  ​"Making Openness Work (So You Don’t Have To)": Bridging Open Science and Higher Education with the ZOERR in times of AI

  The ZOERR – the central open educational resources repository for institutions of higher education in the state of Baden-Württemberg - can be understood as a service provider for Open Science activities at universities.
  ZOERR has created an infrastructure for publishing teaching/learning materials under an open license. These OER are explicitly modifiable and flexible in many respects. They can be dynamically adapted to the respective context and different target groups as well as current research results and teaching/learning concepts (e.g. project-based learning).
  ZOERR is a service that covers several dimensions resulting from the high flexibility of OER. A portfolio of services has been developed to support the production, use and sustainability of OER. The ZOERR actively ensures quality standards in matters of technical, legal, didactic and formal (including metadata) requirements.
  In addition to the pure provision tasks of a repository, the ZOERR portfolio includes consulting, production, quality assurance, public relations, networking, community building and engaging with emerging topics (like AI) in the OER universe. This puts ZOERR in a strong position to address the challenges posed by the tension between open science, OER and OEP on the one hand and commercial AI products on the other.
  OER producers today often see no problem in producing open material with commercial AI tools. ZOERR is trying to raise awareness of the problems that arise from this paradox. 

  Michael Derntl

  The Role of the University of Tübingen's Digital Humanities Center in Promoting Access, Sustainability, and Digital Competencies

  The Digital Humanities Center at the University of Tübingen supports sustainable research practices through digital infrastructure, data stewardship, and training. This poster presents three key areas of contribution: promoting open access to research data collections via the university’s institutional research data repository, including FAIR data publishing support and long-term curation; enhancing access to museal object collections through creation of 3D models and provision of virtual worlds; and fostering resilience through educating students and training early and mid-career researchers in transdisciplinary digital competencies, such as through the university's Data Literacy Certificate. These examples demonstrate how a central digital humanities unit can strengthen the accessibility, preservation, and responsible use of diverse academic collections.

  Kevin Körner, Glaucia Peres da Silva, Esther Fink, Ann-Kathrin Steiner, Luke West

  Exploring Virtual Worlds Together: The Multiplayer Serious Game Platform Graveler for Multimedia Knowledge Transfer in GLAM Contexts

  The digitalization of objects administered by Galleries, Libraries, Archives, and Museums (GLAMs) enables modern and versatile ways of managing collections. At the same time, new digital tools are emerging that allow GLAMs to present and interact with their digitized objects in more immersive ways than would be possible with their real-world counterparts.
  On our poster, we present Graveler – a tool developed at the University of Tübingen as an interdisciplinary project involving the M.A. specialization in Digital Humanities at the Faculty of Humanities, the Global Awareness Education program, and the Center for Teaching and Learning in Higher Education. Graveler enables the creation of browser-based, virtual, and interactive role-playing game worlds with an appealing 2D aesthetic.
  GLAM institutions can embed their digitized objects into these virtual worlds using various media types such as images, audio, and video, as well as integrate external sources like websites, databases, or 3D viewers (e.g., Sketchfab). In addition, the worlds can include customizable role-playing elements such as scripted or AI-supported non-player characters, collection and combination quests, and minigames. Finally, Graveler supports multiplayer functionality and spatial chat, allowing groups to explore virtual environments collaboratively and engage in discussions within the game.

  Kevin Körner, Stefan Krmnicek, Luca Dreiling

  Immerse in Your Collections: Virtual Reality Case Studies from GLAM Institutions Using the ExPresS XR Framework

  In recent years, virtual reality (VR) has emerged as an immersive medium that offers Galleries, Libraries, Archives, and Museums (GLAMs) entirely new ways of presenting their digitized collections and exhibits in interactive and inclusive formats to broader audiences. To simplify the development of VR exhibitions, the M.A. specialization in Digital Humanities at the Faculty of Humanities and the Ancient Numismatics Research Center at the University of Tübingen developed the ExPresS XR framework. It enables the code-free creation of virtual reality exhibitions, including core elements such as individually designable and interactive showcases that can display media such as 3D models, images, text, audio, and video. Moreover, the framework offers additional features such as freely configurable minigames, data collection functionalities, and AI-based avatars, all of which can be integrated without the need for repeated implementation.
  On our poster, we present the core functionalities of ExPresS XR and discuss two case studies developed with it. The first, Temple Tax and Doves, features a 3D visualization of Herod's Temple and playfully conveys knowledge about the temple's economy, including 3D scans of contemporary coins. The second, Hetepheres Tomb VR, depicts the excavation of the tomb of Queen Hetepheres I – mother of Pharaoh Khufu, builder of the Great Pyramid – and illustrates the scientific process of reconstructing her throne at Harvard University. The exhibition further presents 3D models based on objects recreated from the debris discovered after the tomb’s opening.

  Dorothee Huff, Mireille Murkowski, Fabian Schwabe

  Travelling through Data: Julius Euting's Oriental Diaries in the Digital Age

  Julius Euting (1839-1913) was a Swabian scholar and traveller to the Orient. He studied theology and Oriental languages at the University of Tübingen, where he later worked as a librarian. In 1871, his career took him to Strasbourg, where he became an honorary professor of Semitic languages in 1880 and director of the university library in 1900.
  His estate at the University Library of Tübingen includes 26 diaries and ten sketchbooks, which were digitized between 2011 and 2013 and have been publicly accessible ever since. In 2013-2014, the diaries from his trip to northern Syria (1890) were published as a digital edition with an interactive map by the then eScience-Center at the University of Tübingen. Over time, this edition became increasingly difficult to maintain and was replaced in 2025 by a version built using standard JavaScript libraries. The new edition, created by the Digital Humanities master's specialization and the Digital Humanities Center, now provides the complete text of the journey and the excavation at Zincirli, directed by Felix von Luschan.
  The diaries offer a variety of starting points for further historical research, for example, on the history of Oriental studies, 19th-century travel literature, or the cultural history of the Middle East.

  ​Steffen Patzold, Dorothee Huff, Corwin Schnell

  From fragments to findings – Research infrastructures for digital humanities

  Libraries house many untapped treasures within their historical collections. This poster aims to show how digital humanities tools can be used to unlock these treasures and open up new research opportunities, which are supported by up-to-date research services at the University Library of Tübingen.
  Like many other libraries, the University Library of Tübingen holds boxes containing fragments of manuscripts and early printed books. These fragments either entered the library in their current state or were removed from the bindings of books at the University Library. Some consist only of small strips bearing a few individual letters, while others comprise of several leaves. In this way, texts have been preserved that would otherwise have been lost today.
  This poster presents a workflow that starts in the magazine of the University Library and leads to different kinds of representations and options to reuse in digital space. First, the physical object must be converted into a digital object through digitisation. Subsequently, these digital surrogates are made available online and can furthermore be used to virtually reassemble dismembered books. Then, the text is made machine-readable using automatic text recognition. This process provides the foundation for digital editions with linguistic and/or semantic annotations (e.g. in TEI-XML) as well as a wide range of analyses with AI.
  The University Library of Tübingen is committed to using new technological methods to catalogue, explore and make its documents available for research. It strives to do this in close cooperation with scientists in order to create synergies for mutual benefit with the library developing and supporting innovative infrastructures to use for its own holdings as well as making these available as research services. Current developments prepare the promotion of services for scholarly editions using the TEI Publisher, for long-term preservation of data and for AI-based document processing on campus. 

  Thorsten Trippel

  Text+ in Practice: FAIR Access to Language Corpora and Lexical Resources in a Federated Infrastructure

  Text+ is a consortium within the German National Research Data Infrastructure (NFDI), that focuses on language- and text-based research in the humanities and social sciences. Consisting of over 40 partners, Text+ is built on a federated architecture and offers interoperable services for the discovery, preservation, and reuse of diverse data, including digitised library holdings, spoken corpora, lexical resources and scholarly editions. Based on the FAIR principles and aligned with European infrastructures such as CLARIN, DARIAH, and EOSC, Text+ promotes a sustainable, inclusive, and ethically grounded research culture.
  Partners contribute substantial data. For example, the Tübingen Treebank Collection includes TüBa-D/Z and TüBa-D/DP, which provide richly annotated German corpora for multi-level linguistic analysis. Tools such as Tündra, which was developed for these treebanks, enable complex syntactic queries and are also used for Universal Dependencies treebanks in other languages. This supports advanced research in computational linguistics and digital philology. Lexical resources such as GermaNet and the accompanying tool Rover offer semantic networks.
  These resources are integrated into the Text+ Registry and accessible via Federated Content Search (FCS), which enables cross-institutional discovery and analysis. The poster illustrates how the FAIR principles are realised through metadata harmonisation, persistent identifiers, authority files, and standards-compliant formats. It showcases the Text+ Data Space as a collaborative infrastructure that combines local expertise with scalable, sustainable access to high-quality language data.

  Anne Kremmer, Annika Vosseler

  Tracing Origins and Confronting Legacies. Provenance Research at the Museum der Universität Tübingen MUT

  The Museum of the University of Tübingen MUT serves as an umbrella organisation for more than seventy university collections drawn from all faculties of the University of Tübingen, encompassing a diverse range of scientific, artistic, and cultural holdings, as well as ancestral human remains.
  Together, these collections reflect centuries of academic inquiry, collecting practices, and scientific exchange. Provenance research at the MUT investigates the origins, acquisition histories, and movements of these holdings to reconstruct their object and collection biographies and to illuminate the historical, social, and political contexts within which they were incorporated into the university. Within this framework, the MUT pursues a dual approach: a permanent position dedicated to provenance research across the university’s collections, and project-based research that focuses specifically on ancestral human remains from colonial contexts.
  This is illustrated by the recently completed joint project Precarious Provenance as well as ongoing research into ancestral human remains held within certain university collections. This structure enables the MUT to engage critically with questions of ownership, ethics, and historical accountability, while contributing to the rehumanisation of individuals whose remains were removed from their original contexts. By embedding provenance research as a sustained institutional commitment rather than an episodic undertaking, the MUT promotes transparency, critical reflection on the legacies of scientific collecting, and evolving ethical standards. In doing so, it positions itself at the forefront of provenance research among university museums and provides an example for how academic collections can critically engage with their histories and contribute to ongoing dialogues on restitution, repatriation, and reconciliation.
  www.unimuseum.de

  ​Anita Dreher, Claudia Schulte, Julian Dörenberg

  Core Facility Biobank Tübingen - High-quality biosamples and data for research excellence

  The Core Facility Biobank – The Bänk – is a core facility of the Medical Faculty at the University of Tübingen. It is currently under construction. Prospectively, it will provide researchers in Tübingen and beyond with high-quality biological samples of human or animal origin. These samples are processed using standardised methods and stored long-term under optimal, quality-assured conditions. This supports all areas of research, with the ultimate goal of advancing health-related knowledge. Although the biological samples themselves are not data, they bear the potential to generate valuable new information for biomedical research. However, associated data are essential; without comprehensive information such as genetics, clinical records, and lifestyle details, samples cannot be fully utilised for scientific purposes.The Biobank will actively retrieve, combine, curate, and standardise data from various clinical and basic research sources, to ensure that they meet the demands of modern biomedical research demands. This process often involves biobank linkage, connecting biological samples - such as blood, tissue, or DNA - with extensive associated data using unique identifiers. This integration creates powerful research resources by harmonising diverse data types, managing consent, and ensuring data security. Researchers can then link samples to detailed participant information for complex health studies, thus eliminating the need for direct patient contact.Access to new samples and data, as well as existing resources, is vital for successful research. Biobanks are referred to as the "gold of the 21st century", as they provide access a wide range of well-characterised samples that would otherwise be difficult or impossible to obtain. Globally, researchers often face challenges in acquiring sufficient and varied samples, and biobanks serve as crucial repositories that enable scientific progress. For many researchers, biobanks can mark the difference between a stalled project and a breakthrough, making them indispensable to high-quality biomedical research.
  https://www.medizin.uni-tuebingen.de/de/medizinisc...

  Edda Schwarzkopf, Carola Lorea

  The OMnibus of MANTRAMS. A mobile-first, immersive digital archive of multimedia material on mantras in Global Southern Asia

  The ERC-funded research project "MANTRAMS - Mantras in Religion, Media, and Society in Global Southern Asia" is jointly hosted by the University of Tübingen, the University of Vienna and the University of Oxford and currently stands in the first quarter of its six-year project runtime (2024-2030). It aims at producing a global history and anthropology of mantras past and present, combining historical-philological methods, material culture studies and sonic ethnography with unprecedented interdisciplinary collaboration. Researchers and local collaborators collate, collect and produce large bodies of multimedia material in many languages and scripts, with a particular focus on highlighting vernacular religious practices of marginalized communities. The OMnibus of MANTRAMS is the core digital research output, navigating FAIR and CARE principles for Indigenous Data Governance, local protocols of access and representation, EU Data Protection legislation as well as participatory processes of co-creating academic knowledge to build epistemic justice. Establishing a low-tech online presence as a de-colonial collection practice is the foundation of this progressive research design. This poster will present the variety of materials and sub-collections that are part of the MANTRAMS data corpus. It will include community-based design approaches and steps taken towards a sustainable End-of-Life stage for this rich yet easy to digest website narrating the research and the material for and with communities in the Global South.

  Martin Fassnacht

  Tübingen Specialist Information Services: Innovative infrastructures for specialised research

  The three specialist information services based at Tübingen University Library – FID Theology with the Index Theologicus (IxTheo), FID Religious Studies with the Religious Studies Database (RelBib) and FID Criminology with the Criminology Database (KrimDok) – have established themselves as central hubs for their respective disciplines. Through national and international cooperation, the Tübingen FIDs combine technical innovation, high professional quality and consistent service orientation to create unique offerings for their research communities worldwide. The poster presentation highlights five strategic milestones that all three FIDs are pursuing together:

  1. Metadata and reference bibliography
  The three databases systematically record tens of thousands of new titles each year and make them available in professional search systems as comprehensive bibliographic references. Semi-automatic cataloguing procedures for electronic and printed journal articles ensure high-quality indexing of large quantities of material. The metadata is primarily hosted in the German library network, but can be reused worldwide and also found via Google Scholar.

  2. Open Access publication services
  The Tübingen FID operates platforms for open access journals in their respective subjects. New additions include scholarly-led offerings for open access Sammelschriften with professional design templates and technical support. Comprehensive secondary publication services have already made tens of thousands of articles available in open access, including copyright advice and complete publication lists. The self-archiving services are open to all researchers worldwide in the respective subjects.

  3. Personalised services for individualised research services
  Thousands of scientists have personal pages in the three databases, many of them with a complete scientific oeuvre and a high output of open access secondary publications. The personal pages enhance the visibility of researchers and at the same time create sustainable integration into the respective subject infrastructure. In future, authors will be able to help design and expand their own pages – an innovation that transforms the Tübingen FID from a pure research platform into a personalised research service.

  4. FID databases as research analysis tools
  The three databases are developing into instruments for analysing research trends and developments in their respective disciplines. The systematic collection of metadata over centuries enables scientific analysis of publication dynamics, thematic priorities and international research collaborations.

  5. Further development towards digital humanities and AI
  A needs analysis has confirmed the necessity of systematically recording DH/AI-based theological research. IxTheo is therefore one of the first specialist information services to establish its own collection for digital/computational humanities and digital theology, which documents projects, publications, research data and code. This is accompanied by regular 'DH Lunchtalks' with international partner networks, which intensify professional exchange.

• Samuel Wagner

  Vice-President, University of Tübingen, Germany

• Richard Ovenden

  Richard Ovenden, OBE, Head of GLAM, Bodley’s Librarian and the Helen Hamlyn Director of the Oxford University Libraries, UK

Does Reading Matter (Anymore)?

• Convener

  
  Ann Okerson

  Director, Offline Internet Consortium, USA

• Joy Connolly

  President, American Council of Learned Societies, USA

  The Arts of Attention and Invention

  The corporate forces of news media and social media present human reliance on AI as, if I may use the phrase, a fork in the road. Some will say yes to Large Language Models and join in the imminent transformation of work, knowledge, or human life; others will say no and become nostalgic defenders of a lost age. "There is no alternative," as Margaret Thatcher liked to say of capitalism. I will talk about the ways humanists are confronting LLMs in real life - with nuance, with alternatives, and with keen awareness of our need to articulate the value of the skills imparted by reading and writing that LLMs cannot replace, including sustained attention and creative invention (the first step in forming an argument in ancient rhetorical education). It is time to meet the TINA rush to incorporate AI into every classroom and research project with serious arguments about the value of reading and writing as tools for thinking and community building.

• Lukas Griessl

  Post-Doctoral Researcher, University of Tübingen, Germany

  I knew that reading the texts wouldn't do me much good anyway.": Ethnographic Perspectives on Generative AI and Changing Practices of Reading in Academia

  The rise of generative AI (GenAI) tools like OpenAI's ChatGPT in various aspects of academic life has set in motion profound transformations of many everyday academic practices. One domain particularly affected by GenAI is the way we explore knowledge and engage with academic literature, especially the practices of searching for, reading, and working with texts. Drawing on empirical data collected within the project "Hybrid Epistemic Practices" at the University of Tübingen, which ethnographically examines how the introduction of GenAI is transforming the qualitative social sciences and humanities, this presentation provides ethnographic perspectives on these ongoing transformations. In dialogue with the panel's question, "Does Reading Matter (Anymore)?", we turn to how students and academic staff actually use GenAI in their everyday academic lives and how they describe and reflect on the changes introduced by GenAI. The material is based on a large set of ethnographic interviews and AI-based media diaries in which students and staff shared their concrete practices, feelings, and reflections on how generative AI reshapes, among other things, their engagement with academic texts and reading practices at the University of Tübingen.

• Marjorie Levine-Clark

  Professor of History and Associate Dean for Strategic Initiatives & Community Relationships, University of Colorado, Denver, USA

  From Books to Bits: Reimagining Victorian History for Today's Students

  At the University of Colorado Denver, our mission is to "make education work for all," striving to reach a diverse student body of about 50% first-generation college students and 50% students of color. These students are pulled in multiple directions by jobs and families and distracted by social media, and, increasingly, by artificial intelligence. This paper uses my spring 2026 revision of "Victorians and Victorianism," a course I last taught 20 years ago, as a case study to explore how we might respond to these pressures. Although it is disheartening to move away from book-length readings in an intimate, once-a-week history course, my pedagogy has been revitalized through considering how to use "bits" rather than books to develop students’ confidence and curiosity as readers. Through a variety of short primary sources, academic articles, and excerpts, students confront the interpretive challenges of analyzing novels, government documents, images, journalism, maps, and historians' work – learning what each type of source can and cannot tell us about the past. Simultaneously, the course treats AI itself as an object of study: students analyze the limitations and distortions in LLM-generated responses and complete handwritten in-class syntheses each week to ensure their historical thinking remains grounded in their own reading and interpretation. In scaffolded final projects, students reflect directly on how studying Victorian history sheds light on a contemporary issue that matters to them – an approach that encourages intellectual investment and hopefully reduces reliance on AI shortcuts.

• Andrea Bonaccorsi

  Professor of Economics and Management at the School of Engineering, University of Pisa, Italy

  The knowledge of humanities and the role of books. For a new epistemological reflection

• Karla Pollmann

  ​President, University of Tübingen, Germany

Collection as Data

• Convener

  
  Angelika Zirker

  Associate Professor of English Literature, University of Tübingen, Germany
  Convener

  
  Natalie Maag

  Head of the Library of the German Literature Archive, Marbach, Germany

• Hanna Fischer

  Professor of German Linguistics, Research Center Deutscher Sprachatlas, Philipps-Universität Marburg, Germany

  From Historical Survey Sheets to Computational Dialectology: The Wenker Collection as Research Data

  This paper presents a major historical linguistic collection within German dialectology: the more than 57,000 survey sheets ("Wenkerbögen") compiled for Georg Wenker's Sprachatlas des Deutschen Reichs in the late nineteenth century. These documents constitute one of the earliest large-scale, systematic dialect surveys in Europe, yet their analytical potential remained limited for decades due to their handwritten format. Recent digitization efforts have made the entire collection available as high-resolution scans, forming the basis for a dedicated web application designed to transform this corpus into usable research data. At the core of this application is a citizen-science workflow through which volunteers collaboratively transcribe the historical questionnaires. This approach not only accelerates the creation of machine-readable data but also enables transparent quality control. The app provides map-based access that visualizes the spatial distribution of the sheets. Furthermore, the Wenkerbögen are linked to the linguistic-geographical information system SprachGIS and, through this connection, are embedded within the Text+ research data infrastructure, enabling interoperability with other resources and services.The presentation demonstrates how this collection—long difficult to access—has been transformed into a dynamic data resource. Several recent research outcomes illustrate the novel analytical possibilities arising from the digitized and transcribed Wenker materials, including linguistic variables previously not considered as well as cultural and socio-demographic information that has not yet been systematically explored. These enriched data support new computational approaches to dialect change. By framing the Wenker collection as "collections as data," the talk highlights both the methodological challenges and the emerging research opportunities that result from the convergence of heritage collections, citizen science, and digital humanities.

• Gaetan Rappo

  Associate Professor of Religious Studies and Japanese History, Faculty of Culture and Information Science, Dōshisha University Kyoto, Japan

  Data Mining and Text Reuse in the Collections of Japanese Buddhist Texts

  This presentation outlines the current state of a computational workflow designed to analyze Japanese religious corpora dating from the 9th to the 16th centuries, specifically Buddhist and Shinto collections. The ultimate objective is to provide a clearer, long-term picture of doctrinal evolutions that transcends traditional sectarian boundaries. The methodology is designed to be flexible, often beginning directly with the processing of manuscripts to create digital editions grounded in Text Encoding Initiative (TEI) guidelines. Utilizing tools such as the Versioning Machine to visualize textual variants, this philological foundation supports subsequent data mining tasks. The analysis combines text mining techniques such as topic modeling with vectorization strategies for text reuse detection. By synthesizing these approaches, the project aims to establish a framework for mapping the complex intertextuality and shifting doctrinal landscapes within Japan's historical religious texts.

• Sarah Schnitzler

  PhD candidate in Literary and Cultural Theory and a Researcher in Fan Fiction at the DLA in Marbach, Germany

  Adding extra tags to prevent future confusion: Fan Fiction Paratext and/as Library Metadata

  Fanfiction as a medium is characterised by its sprawling, ephemeral nature in the digital space. Fans who produce and consume fanfiction have themselves implemented a system of metadata, specifically designed for the needs of transformative fan activities, to allow for maximum accessibility of their content. The poster child of this networked metadata system is indubitably the platform Archive of Our Own (AO3). Designed to fit the specific needs of its userbase, the AO3 has standardised a process of tagging and categorising a work of fanfiction from the moment of uploading. Any work is thus catalogued within the digital archive according to a set of suggested criteria, as well as any additional information a writer may feel the need for. Made possible by the archive's volunteer group of so-called "tag wranglers," who continuously order and connect the tags in use as the archive expands, the interface is highly searchable and accessible. The AO3's metadata system thus allows not only for free browsing according to an infinite combination of criteria, but even for serendipitous discovery.For any efforts to archive works of fanfiction, such as the collection initiated by the library of the German Literature Archive Marbach (DLA) in October 2025, the resultant metadata presents a unique opportunity and challenge. In a collecting institution such as the DLA, any archive material needs to be catalogued and made accessible for researchers. This process is often standardised and mirrors (international) norms and guidelines. DLA metadata thus permits synchronicity with systems such as the GND (the Integrated Authority File managed by the German National Library). When the material entering the archive's collection is a work of fanfiction, the fan constructed metadata system and the system implemented by an institution such as the DLA collide. This collision raises questions: Do the needs of a fanfiction platform like the AO3, which its metadata is uniquely designed to fulfil, overlap with the needs of a library collection? And if so, can this overlap be utilised productively in order to facilitate and enrich the archiving process or even to generate knowledge for future strategies of cataloguing? It is by answering these questions and exploring this overlap that we can discover the value fanfiction collections present for archiving processes and library cataloguing.

• Megan Bushnell

  Researcher in linguistic data, Oxford Text Archive, UK

  Mobilising the Oxford Text Archive in the Development of Computational Linguistic Tools

  The Oxford Text Archive is one of the oldest repositories of digital literary and linguistic data. Founded in 1976 at the dawn of text encoding, the archive boasts over 70,000 deposits, featuring digital editions of texts, linguistic corpora, lexicons, dictionaries, semantic networks, and language analysis tools, from a variety of time periods and languages. With the advent of AI, this massive dataset has new potential uses for the development of large language models and AI tools – especially regarding responsibly sourced, quality data. However, work needs to be done to investigate how to effectively harness this data, given its previously stated diversity, and how this diversity can be a strength, especially for generating tools to benefit lesser-resourced languages and historical varieties. This presentation outlines initiatives the Oxford Text Archive has led to expand its holdings and to explore how its data might contribute to new research. In collaboration with its CLARIN partners, the Oxford Text Archive has been working to strengthen the volume and quality of data made available for lesser-resourced languages of the UK, leading to the creation of a new CLARIN centre for Digital Resources for the Languages in Ireland and Britain (DR-LIB). In collaboration with the DR-LIB Centre, the Oxford Text Archive has recently carried out a pilot project, 'Unlocking AI for the Languages in Britain and Ireland', which aims to improve the available digital resources for lesser-resourced languages in the UK and adapt training resources using new and enhanced language models. This presentation will discuss lessons learned from both the DR-LIB centre and 'Unlocking AI' project, and will outline future plans for supporting the development of computational tools for lesser-resourced and historical languages.

Preserving Integrity: Collective Actions for Trust and Impact

• Convener

  
  Michael Levine-Clark

  Dean of Libraries, University of Denver Libraries, Colorado, USA

• Emilie Hardman

  Curatorial and Archival Practice Director, JSTOR Digital Stewardship Services, ITHAKA, USA

  From Distribution to Stewardship

• Cristina Blanco-Sancho

  Senior Director of Product Management, Academic AI, Clarivate, USA

  How Infrastructure, Metadata, and Analytics Shape Trust

• Stephen Rhind-Tutt

  President, Coherent Digital, USA

  Using AI for Increased Efficiency and Functionality

• Teri Oaks Gallaway

  Executive Director, SCELC, USA

  The Consortium as a Trust-Building Organism

Collection development under constraint: coordination, collaboration, and partnership in a «coopetitive » world

• Convener

  
  Julien Roche

  University Librarian and Director of the Libraries, University Lille, France; LIBER President

• Stephanie Chancy

  Caribbean Partnerships Librarian, Director of Operations Digital Library of the Caribbean (dLOC), George A. Smathers Libraries, University of Florida, Gainesville, USA

  When Cooperation is the Key, an Introduction to the Digital Library of the Caribbean

  For more than two decades the Digital Library of the Caribbean has endeavored to build partnerships, advance capacity building among partners, and encourage the preservation of Caribbean and Caribbean diaspora research resources. Over the years these efforts, which take many forms, have established dLOC as a non-extractive organization that fosters mutually beneficial relationships, while serving as a centralized access point for Caribbean research material. In the session, dLOC's Director of Operations introduces the organization, its operating model, and post-custodial approach to collection development. The presentation highlights the cooperative nature of the partnerships where all partners have a voice, all contribute to its sustainability, and all come together to give an international user base free access to collections that focus on the Caribbean and its diaspora.

• Grégory Miura

  Executive Director, CollEx-Persée consortium / Campus Condorcet, Paris, France

  CollEx-Persée in a Networked World: Can Federated Infrastructures Balance Access, Sovereignty, and Permanence? A Case Study in Coopetitive Collection Development Under Constraint

  This intervention critically examines CollEx-Persée as a federated yet flawed experiment in balancing access, sovereignty, and permanence within today's constrained academic ecosystem. While it breaks from legacy consortium models through distributed governance and collective innovation, it remains caught between transformative potential and structural constraints—fragmentation risks, ephemeral resource dependencies, and uneven standard adoption. By incubating shared practices such as collaborative cataloging and open archiving, CollEx-Persée redefines collection development, yet its success hinges on addressing funding instability, institutional disparities, and the digital divide. Far from an idealized vision, this case study grounds its analysis in the complexities of coopetitive dynamics, offering a sober yet forward-looking perspective on the possibilities and limits of federated infrastructures in an era of scarcity.

• Brian Hole

  CEO of Dromologics, London, UK, and founder of Ubiquity Press.

  Opportunities and principles for Libraries to collaborate for more impact in publishing

  Academic publishing doesn't change quickly, yet it is probably entering its greatest period of change since the advent of the internet. Developments in open science, publishing and funding models, and especially AI all provide opportunities for a stronger and more innovative role for libraries in publishing, both independently and in collaboration with other partners. The speaker will discuss these topics and suggest principles on which such collaboration can be centred, based on his extensive experience.

• Jérémie Roche

  Director of institutional relations at Cairn.info, France

  Under pressure - Collaboration and innovation in francophone scholarly publishing

  Just over twenty years ago, several independent (and competing) francophone publishers made a bold and unusual decision. Rather than going at it alone, they decided to join forces and launch a common online platform to offer digital versions of their journals: Cairn.info. The platform has since grown significantly, now gathering publications by close to 400 scholarly publishers, from all over the French-speaking world. Since its inception, Cairn.info's faced challenges - some unique, others similar to those faced by publishers of all sizes, all over the world. How can one achieve a representative collection from a truly equal and collaborative perspective? How can one invest, innovate and make a difference while under significant pressure, in a small linguistic area and target market? What can small(er) francophone publishers learn from their much bigger, international counterparts? The presentation will answer these questions, and more, through several milestones in Cairn.info's history, from initial catalog development to a current foray into generative AI, from innovative commercial models to international partnerships.

• James O'Donnell

  University Librarian, Arizona State University, USA