CAREing for Data in Dialogue (2026)
Perspectives on social, cultural data and beyond — the seventh edition of the “… in Dialogue” series
Research data in the humanities and social sciences is rarely simply “open” or “closed”. Much of it describes people — their words, their images, their suffering — and asking what it means to care for such data is a question of ethics as much as of infrastructure. The CARE principles put collective benefit, authority to control, responsibility and ethics beside the FAIR ones, and this workshop asks what that looks like in practice.
For the seventh edition of “… in Dialogue”, researchers bring concrete datasets to the table and discuss how they handle the ethical questions each one raises. Short pitches are combined with hands-on work and with perspectives from law and information security.
9 December 2026 · Dome Room (Kuppelraum), Main Building (H4), University of Bern · the event is held in English.
Short link to this page: https://bit.ly/imdialog2026
Programme
| Time | Session | Speakers |
|---|---|---|
| 09:00–09:15 | Welcome | |
| 09:15–10:05 | Keynote — Beyond FAIR: CARE, Community Responsibility, and the Challenges of (Historical) Sociolinguistic Data | Anita Auer |
| 10:05–11:05 | Dataset pitches | |
| The Case of Digital Scholarly Editing: How to Mediate Discrimination in Annemarie Schwarzenbach’s Texts and Photographs | Elias Zimmermann & Levyn Bürki | |
| Patient texts from German psychiatric hospitals of the 19th- and early 20th centuries | Markus Schiegg | |
| Who CAREs about Witches? | Christa Schneider | |
| Ethnographic Data: Between In-Depth Description and the Need to Protect Interlocutors | Stefan Leins | |
| Data Steward’s perspective | Data Stewards, University of Bern | |
| 11:05–11:35 | Coffee break | |
| 11:35–12:30 | Whose Data, Whose Benefit? The values behind CARE — and what they demand of the scientific community and of infrastructures like CLARIN | Cristina Grisot |
| 12:30–14:00 | Lunch break | |
| 14:00–15:00 | Dataset pitches | |
| Securing Data Vulnerabilities | Elisa Fornalé | |
| A Safe Space for Testimony: CARE and the Research Infrastructure for Studying Sexual Abuse in the Environment of the Catholic Church in Switzerland | Moritz Mähr & Magda Kaspar | |
| “There was a time when demons just did the trick.” — CAREing for cultural property in the Information Age | Sebastian Borkowski | |
| When Every Minute Identifies: Emergency Medicine Data at the Boundary of Openness | Martin Müller | |
| 15:00–16:10 | Hands-on (parallel, coffee served) | |
| Learning to See the Problem: Discrimination-Sensitive Metadata in Practice | Moritz Mähr | |
| Bring your own data | Christa Schneider | |
| 16:10 | Open, but Not at All Costs: Balancing Openness, Data Protection, and Ethical Responsibility in Research Data | Gianfranco Valsecchi |
| Afterwards | Closing & apéro |
Abstracts
Beyond FAIR: CARE, Community Responsibility, and the Challenges of (Historical) Sociolinguistic DataAnita Auer · keynote
(Historical) Sociolinguistic research presents a range of challenges concerning the collection, representation, and ethical handling of linguistic data. While the FAIR principles (Findable, Accessible, Interoperable, and Reusable) provide guidelines for research data management, the CARE principles (Collective Benefit, Authority to Control, Responsibility, and Ethics) foreground questions of representation, ownership, and community responsibility.
This talk explores the opportunities and tensions between these principles through two sociolinguistic case studies. The first – historical sociolinguistic – case study examines pauper letters from 1795-1834, highlighting challenges of archival regulations, questions of copyright, and the creation of reusable historical corpora. The second focuses on Swiss Heritage Tamil, where the question of responsible (meta) data sharing raises ethical considerations. Drawing on practical experiences from both projects, I discuss data collection, curation, as well as the ethical considerations that place limits on data sharing. In particular, I highlight how commitments to CARE may conflict with expectations of openness and reuse under FAIR, making certain data unsuitable for public access. Ultimately, the talk asks how we can navigate these tensions while creating sociolinguistic datasets that are both scientifically valuable and ethically responsible.
The Case of Digital Scholarly Editing: How to Mediate Discrimination in Annemarie Schwarzenbach’s Texts and PhotographsElias Zimmermann & Levyn Bürki
Once a corpus has been defined on the basis of scientific and legal considerations, scholarly editions face less of a challenge in selecting which sensitive data to collect or display. Their task might be to edit data that is ethically problematic but — sometimes precisely for that reason — historically and philologically important. Our focus is therefore on the question of how to mediate such material from the data level up to its display in the frontend.
Since 2025, the Schwarzenbach Edition (in particular Levyn Bürki, Daniela Kohler and Elias Zimmermann) has developed its own discrimination-sensitive editorial guidelines in consultation with postcolonial researchers and GLAM institutions. The guidelines set out fundamental considerations regarding the presentation of content and translate them into specific policies for tagging, annotation and presentation. For example, discriminatory expressions are marked up in TEI/XML to enable the automated and targeted display of content notes with references to explanatory editorial comments in the frontend. We argue that discrimination-sensitive editing should not be viewed merely as a technical, presentational or philological challenge, but that all three aspects should be considered in close conjunction with one another.
Patient texts from German psychiatric hospitals of the 19th- and early 20th centuriesMarkus Schiegg
The Corpus of Historical Patient Documents (CoPaDocs) is based on texts written by patients from 19th- and early 20th-century psychiatric hospitals in Germany. Despite being over 100 years old, consulting these sources means entering the realm of medical confidentiality. Their high scientific relevance, however, justifies their archiving and use (Fichtner 1997: 116). This presentation shows how CoPaDocs deals with this tension in practice. The sociolinguistic focus of my research (see Schiegg 2022) contradicted with a full pseudonymisation, meaning that only a patient’s surnames was abbreviated to the first letter, and the first name, together with a large amount contextual information relating to the individuals, has been published. However, access to certain contextual data, in particular a patient’s medical diagnosis, was made more difficult by limiting the search options. They were designed to prevent quick searches within the illness metadata and any resulting ahistorical or voyeuristic misuse.
References
- CoPaDocs = Schiegg, Markus (ed.) (2023): Corpus of Patient Documents. https://deutschestextarchiv.github.io/copadocs/; https://www.dwds.de/d/korpora/copadocs
- Fichtner, Gerhard (1997): Das Problem der Schweigepflicht und der Historiker. In: Dietrich Meyer & Bernd Hey (eds.): Akten betreuter Personen als archivische Aufgabe. Beratungs- und Patientenakten im Spannungsfeld von Persönlichkeitsschutz und historischer Forschung. Neustadt/Aisch: Degener, 111–122.
- Schiegg, Markus (2022): Flexible Schreiber in der Sprachgeschichte. Intraindividuelle Variation in Patientenbriefen (1850–1936). Heidelberg: Winter.
Who CAREs about Witches?Christa Schneider
How much ethical responsibility do researchers have towards individuals who died centuries ago, particularly when their personal histories are preserved in records of criminal prosecution? Drawing on research with early modern witch-trial records from Switzerland and beyond, this talk explores how the ethical treatment of historical data varies across national and institutional contexts. While Swiss archival practices impose very little restrictions on working with these records, access to e.g. comparable Scottish material has required ethical procedures similar to those applied to contemporary court records. This contrast raises questions about the relevance of anonymisation, the identifiability of historical individuals, and the extent to which contextual information should be preserved or disclosed. Using concrete examples, the talk considers how the CARE Principles might inform our responsibilities towards individuals represented in historical records, even centuries after their deaths.
Ethnographic Data: Between In-Depth Description and the Need to Protect InterlocutorsStefan Leins
As social anthropologists, we often generate ethnographic data through participation in other people’s everyday lives, work routines, and social situations. We also generate data through interviews, as well as through informal conversations that are often not standardized and not even recorded. The idea is that the closer we get to our interlocutors, the better we can portray their lives and practices through what we call thick description. However, these data are often highly personal and intimate. How can we navigate the discrepancy between our desire to provide detailed, personal accounts and our responsibility to protect our interlocutors? This tension will be at the center of my data pitch.
Whose Data, Whose Benefit? The values behind CARE — and what they demand of the scientific community and of infrastructures like CLARINCristina Grisot
Language resources shared through infrastructures like CLARIN are usually judged by FAIR criteria — Findable, Accessible, Interoperable, Reusable — but FAIR never asks who the data serves, controls, or benefits. This talk turns to the CARE Principles for Indigenous Data Governance and asks: what do CARE’s underlying values — collective benefit, authority to control, responsibility, ethics — demand of the scientific community, and of infrastructures like CLARIN, when the community isn’t Indigenous? Drawing on work done in the CLARIN community and Swiss dialectology research, it proposes practical steps.
Securing Data VulnerabilitiesElisa Fornalé
Building on the fieldwork conducted in Mexico, Italy and Bangladesh, the presentation aims to discuss how ensuring data protection of vulnerable subjects (e.g. irregular migrants/asylum seekers/returnees) in vulnerable contexts. Data protection practices require the adoption of significant steps to ensure that people are not exposed to additional harms. The presentation will describe some practical challenges and ethical concerns faced during the first phase of data collection efforts, storage and transfer of data. Our SNSF Resisting Human Erosion research project aims to collect qualitative data in multiple sites and this presentation will offer the opportunity to get feedback and to exchange good practices for handling data in a safe and reliable manner.
A Safe Space for Testimony: CARE and the Research Infrastructure for Studying Sexual Abuse in the Environment of the Catholic Church in SwitzerlandMoritz Mähr & Magda Kaspar
The research project on the history of sexual abuse in the Catholic Church in Switzerland uses highly sensitive data from church and state archives, as well as many interviews with affected persons. These data are subject to strict data protection and privacy requirements and can therefore be made accessible only with restrictions. At the same time, the research findings must remain transparent and verifiable for other researchers. This contribution addresses this tension and asks how the necessary protection of affected and accused persons can be reconciled with scientific verifiability and long-term archiving.
“There was a time when demons just did the trick.” — CAREing for cultural property in the Information AgeSebastian Borkowski
Capitalism and colonial exploitation converted day-to-day objects of past societies into a monetisable resource with unregulated profit margins and heralded nationalist and fascist coping strategies for postcolonial identity crisis. By destroying the archaeological context, illicit excavations of antiquities destroy the primary source of information about the ancient social and cultural customs fossilised in archaeological objects. To prevent looting and the illicit trade in antiquities, the location and dating of archaeological sites are treated as sensitive data by governments and archaeologists. This secrecy conflicts with open research data policies and ethical responsibilities for knowledge dissemination and public education — a debate that does not necessarily involve the voices of descendant communities.
In the input presentation, I provide a short outline of the general problematic of illicit excavations and current (best practice) solutions, before I turn the attention toward the current state of negotiations about the reuse and open access publication of unpublished and published archaeological data from Southern Iraq between the SNSF Sinergia project HyMes and the Iraqi State Board of Antiquities and Heritage.
When Every Minute Identifies: Emergency Medicine Data at the Boundary of OpennessMartin Müller
Emergency medicine depends on detailed information: exact times and locations, clinical narratives, images, treatment sequences and information about relatives or witnesses. These details are essential for accurate diagnosis and meaningful research, but they can also make patients identifiable even after names have been removed. Using fictional cases and examples from emergency medicine registries, this short talk examines which data are particularly sensitive, who may be affected by their disclosure, and how public availability could lead to individual and systemic harm — from violence, discrimination and coercion to loss of trust, delayed care and biased research. It argues that valuable knowledge, metadata, code and aggregate findings should be shared openly, while person-level emergency records require purpose-limited, secure and accountable access.
Learning to See the Problem: Discrimination-Sensitive Metadata in PracticeMoritz Mähr · hands-on session
Metadata are not neutral: the ways in which historical sources and research data are described shape what can be found, how people and events are represented, and which historical power relations are reproduced. This hands-on session introduces the Handbook for Discrimination-Sensitive Metadata Practice, which combines critical reflection on discrimination, bias, representation, and power with practical guidance and checklists for working with metadata. Participants will use these checklists to examine examples from their own institution, or an institution of their choice, and identify potentially problematic descriptions, classifications, omissions, and contextualisations. The aim is not to provide a simple checklist of “right” and “wrong” terms, but to develop awareness of how discrimination can become embedded in metadata practices and to strengthen participants’ ability to recognise and critically discuss problematic cases.
Open, but Not at All Costs: Balancing Openness, Data Protection, and Ethical Responsibility in Research DataGianfranco Valsecchi
What does it mean to share research data fairly and responsibly? It is now widely recognised that research data should be as open as possible and as restricted as necessary, while legal and ethical considerations may require additional safeguards and restrictions. Different degrees of openness can help balance these two principles by allowing data to be shared and reused under varying conditions, depending on the level of access provided and the types of reuse permitted, while taking into account the legal, ethical, and social considerations at stake.
Speakers
Anita Auer
Professor of English Linguistics at the University of Lausanne; her work in historical sociolinguistics includes the language of the labouring poor in Late Modern England.
Sebastian Borkowski
doctoral researcher at the Institute of Archaeological Sciences, University of Bern, and a member of the HyMes project.
Levyn Bürki
Digital Humanities Center, University of Bern.
Elisa Fornalé
Professor at the University of Bern and gender coordinator at the World Trade Institute; works on international law, human rights, migration and the environment.
Cristina Grisot
Linguistic Research Infrastructure (LiRI), University of Zurich; national coordinator of CLARIN-CH and scientific coordinator of DARIAH-CH at DaSCH.
Magda Kaspar
historian at the Department of History, University of Zurich, researching sexual abuse in the context of the Catholic Church in Switzerland.
Stefan Leins
Professor of Social Anthropology at the University of Bern, working on the cultural practices of financial markets and commodity trading.
Moritz Mähr
Institute of History & Digital Humanities, University of Bern, and Department of History, University of Zurich; works on research infrastructures for sensitive historical data.
Martin Müller
Department of Emergency Medicine, Inselspital, University Hospital Bern, where he leads the Resource & Risk Management group.
Markus Schiegg
Professor of German Linguistics at the University of Fribourg; his research on patient letters from psychiatric institutions works closely with highly sensitive personal records.
Christa Schneider
Institute of German Studies, University of Bern; her research includes the Basel witch trials.
Gianfranco Valsecchi
lawyer with the Competence Center in Digital Law (CCdigitallaw), Università della Svizzera italiana; advises and trains Swiss higher education institutions on data protection, copyright and open science.
Elias Zimmermann
editor of the digital edition of Annemarie Schwarzenbach’s short forms and letters, University of Geneva and University of Zurich, and the Swiss Literary Archives; the project has developed its own discrimination-sensitive editorial guidelines.
Registration
Registration is free and runs through the University Library Bern:
Contact & Organisers
If you have questions, please contact imdialog@ub.unibe.ch. Further details are announced here as the programme is settled.
Organisers:
- Dr. des. Christa Schneider (Institute of German Studies)
- M.A. Ursula Loosli; Dr. Gero Schreier (Open Science, University Library Bern)
- Prof. Dr. Tobias Hodel (Digital Humanities)
Header image by Simon Lee on Unsplash, used under the Unsplash licence. Cropped to a banner format.