scieee AI-readable full text Open interactive document viewer

Making the Digital Humanities More Diverse and Inclusive: Theories and Practices

Lang, Sarah

Publisher
Zenodo
DOI
10.5281/zenodo.17733430
Repository
zenodo.org
Language
en

Abstract

Slides of the presentation Making the Digital Humanities More Diverse and Inclusive: Theories and Practices held by Dr. Sarah Lang in the lecture series Digital Humanities in Focus: Methods, Applications, and Perspectives at the University of Rostock (online) on October 27, 2025. Abstract: This talk revisits the concept of the “dark sides of DH” as a lens for examining persistent structural issues in digital humanities. Originally focused on neoliberalism, these debates revealed discomfort with DH’s institutional positioning and perceived lack of critical engagement. Internally, DH is often seen as a progressive force, yet may be blinded by vocational awe—the uncritical belief in its intrinsic virtue—which can obscure systemic inequalities. These concerns, still as relevant today as when the topic of “dark sides” was first discussed, reflect entrenched disparities inherited from academia and the tech industry. Issues including limited engagement with AI ethics, data gaps, language diversity, and the marginalisation of feminist perspectives, despite their potential to address structural imbalances still persist today. Furthermore, the dominance of AI and other computational (rather than “just digital”) methods narrows the scope of what is considered legitimate research in the field, limiting potential for more inclusive representation. It is a naive assumption that good people will know how to act ethically because this requires skills and tools. This talk highlights practical strategies to address current challenges, drawing on dataset audits and the forthcoming ZfdG Working Paper on Data Feminism from the AG Empowerment group. Grounded in critical archival studies, these initiatives offer accessible approaches for promoting more equitable, inclusive practices in DH. Bio: Sarah Lang is Head of Digital Humanities at the Max Planck Institute for the History of Science (Berlin). Previously, she was a Postdoctoral Fellow at the Centre for Information Modelling at the University of Graz. Trained in History and Classics in Graz and Montpellier, she completed a PhD on early modern alchemical literature in 2021, combining Digital Humanities and the history of science, for which she received the Bader Prize of the Austrian Academy of Sciences.As convenor of the Empowerment Working Group of the German Digital Humanities Association (DHd), where she is also on the board of directors, Sarah Lang is interested in issues like (gender) data gaps, data feminism, diversity in DH, decolonizing data, data ethics and related topics.

Full text

Making the Digital Humanities More Diverse and Inclusive Sarah Lang* DH Rostock Ring-VO *Max Planck Institute for the History of Science (Berlin) 1/17 Current dark sides of DH and how to make DH more diverse 1. ‘dark side of DH’ discourse ca. 10 years ago 2. what are current dark sides of DH? 2/17 Upcoming ZfdG Working Paper (S. Lang & E. Suárez Cronauer 1. Data Feminism isn’t just for women 2. Data Feminism as historical source criticism 3. DH work is more political than it may seem 3/17 Data Feminism 1. Principle – Examine Power (Machtstrukturen untersuchen) 2. Principle – Challenge Power (Machtstrukturen herausfordern) 3. Principle – Elevate emotion and embodiment (Beachten von Gefühlen und Ausdrucksformen) 4. Principle – Rethink binaries and hierarchies (Binaritäten und Hierarchien überdenken) 5. Principle – Embrace pluralism (Diversität fördern) 6. Principle – Consider context (Den Kontext beachten) 7. Principle – Make labor visible (Die Arbeit sichtbar machen) Principles of Data Feminism according to D’Ignazio & Klein (dt. Übersetzungen nach Juen 2021). 4/17 Current ‘Dark Sides of DH’ (Chun et al. 2016; Smithies 2022) The Ethical Gap in Digital Humanities • DH lacks ethics (Berry 2022) • ML lacks accountability (Raji et al. 2020) • Algorithm auditing and dataset probing address these concerns, as seen in dataset revisions (e.g., ImageNet cleanup, cf. Yang et al. 2020; Crawford 2021) Bias and Gaps in Historical Datasets • Colonial and structural bias demand mitigation strategies • Contextualising gaps improves AI explainability • Transparency aids diversity, equity, and rigour by at least being able to state which perspectives are missing (and begin to fill gaps). •Data Feminism, Critical DH, critical data studies, and critical code studies provide frameworks for interrogating data. 5/17 Valorizing data work i FAIR and CARE Principles • Widely recognized in digital humanities (Egan and Murphy 2022) •FAIR: Findable, Accessible, Interoperable, Reusable. •CARE: Collective benefit, Authority to control, Responsibility, Ethics (Global Indigenous Data Alliance 2021). FAIR Data is Not Enough • Needed for explainability, replicability, and critical engagement. • Issues like data gaps, critical DH, and data/code studies require richer metadata. • Current project descriptions often lack the depth needed for true, practical dataset reusability. 6/17 Documentation as Activist Practice Why Document? • Tackles colonial biases, gaps (e.g., gender) • Exposes digitisation politics (Zaagsma 2023) • Supports decolonisation by making structural bias visible Support Needed • Acknowledge documentation as research • Provide funding, platforms, and incentives for probing & documentation 7/17 Documenting datasets Datasheets for Datasets Gebru, Morgenstern, Vecchione, Wortman Vaughan, et al. 2018 / Gebru, Morgenstern, Vecchione, Vaughan, et al. 2021 Model Cards for Models Mitchell et al. 2019 Datasheets for Digital Cultural Heritage Datasets Alkemade et al. 2023 Data-Envelopes for Cultural Heritage: Going beyond Datasheets Luthra and Eskevich 2024 Plus data paper formats, such as: •Reviews in DH, •Zeitschrift für digitale Geisteswissenschaften (ZfdG), •Journal for Open Humanities Data (JOHD) 8/17 Analyzing Historical Biases with Datasheets Datasheets as a Tool for Critical Inquiry • Who is represented in the dataset? Whose voices are missing? • What biases emerge from the collection process? • How can transparency improve dataset reusability? Applying Data Feminism • Recognizing systemic inequalities in datasets. • Understanding the limitations of historical sources. • Making informed decisions on dataset interpretation and usage. 9/17 References iv [13] Inioluwa Deborah Raji et al. “Closing the AI Accountability Gap: Defining an End-to-End Framework for Internal Algorithmic Auditing”. In: Conference on Fairness, Accountability, and Transparency (Fat* ’20). New York, NY, USA: ACM, 2020, p. 12. DOI: 10.1145/3351095.3372873. [14] Donna Riley. “Rigor/Us: Building Boundaries and Disciplining Diversity with Standards of Merit”. In: Engineering Studies 9.3 (2017), pp. 249–265. DOI: 10.1080/19378629.2017.1408631. URL: https://doi.org/10.1080/19378629.2017.1408631. [15] Shawna Ross and Andrew Pilsch. “Labor, Alienation, and the Digital Humanities”. In: The Bloomsbury Handbook to the Digital Humanities. Ed. by James O’Sullivan. First. Bloomsbury Handbooks. London: Bloomsbury Academic, 2022, pp. 335–346. DOI: 10.5040/9781350232143.ch-32. [16] James Smithies. “The Dark Side of DH”. In: The Bloomsbury Handbook to the Digital Humanities. Ed. by James O’Sullivan. Bloomsbury Handbooks. Accessed June 12, 2025. London: Bloomsbury Academic, 2022, pp. 111–122. DOI: 10.5040/9781350232143.ch-11. URL: https://dx.doi.org/10.5040/9781350232143.ch-11. 16/17 References v [17] Kaiyu Yang et al. “Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the Imagenet Hierarchy”. In: Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. 2020, pp. 547–558. [18] Gerben Zaagsma. “Digital History and the Politics of Digitization”. In: Digital Scholarship in the Humanities 38.2 (2023), pp. 830–851. DOI: 10.1093/llc/fqac050. 17/17