Submission deadline: April 15, 2026
In recent years, the field of manuscript studies has been profoundly transformed by the success of deep learning. Tools such as Transkribus, Arkindex, eScriptorium, and various other OCR/OCR engines have reached such a level of maturity that "near-perfect" transcriptions of complex historical scripts are no longer a distant dream but an everyday reality at all scales, from the transcription of isolated documents to the creation of massive corpora. After the heroic phase of automatic recognition of handwriting, a paradigm shift has taken place and the question remains: what next?
It is no longer just a question of transcribing better, but of understanding how ATR transforms the regimes of evidence, authority, editorialization, and interpretation in the historical sciences. The widespread use of ATR on a large scale opens up a The widespread use of ATR on a large scale opens up a new field of research: that of the secondary, critical, and epistemological uses of data produced by AI models.
Once the digital ink has dried and millions of searchable text tokens are available, how does this transform the way we interact with the past? This conference proposes to explore the "life after" ATR (Automatic Text Recognition) data. We want to overcome the technical obstacles associated with character error rates and enter the fields of big data analysis, automated cataloging, digital publishing, archival discovery, and new research methodologies.
Biblissima+ Cluster 3 invites researchers, librarians, computer scientists, and digital humanities specialists to submit proposals focusing on the following topics (non-exhaustive list):
(1) New Uses of Text: From Text to Knowledge (structuring, modelling, interoperability); post-processing & LLMs; distant reading and automated cataloguing; scholarly editing and textual criticism.
(2) Materiality and the History of Written Objects: new computational paleography; text, materiality, and image.
(3) Heritage, Epistemological, and Ethical Issues: epistemology and ethics; sustainable heritage policies for ATR.
In particular, we propose to explore these themes from the following perspectives:
New Uses of Text
- From text to knowledge: structuring, modeling, interoperability
How can we transform a mass of raw text into an ecosystem of usable knowledge?
Associated concepts
Named entity extraction (NER), disambiguation, and alignment with existing repositories ; Linked Open Data (LOD) on ATR outputs ; Construction of knowledge graphs from ATR corpora ; Interoperability with heritage protocols and ontologies (IIIF, TEI, CIDOC-CRM) ; Transition from linear text to relational data
- Post-processing & LLMs
How can large language models be used for automatic correction, translation, or summarization of historical manuscripts and where is scholarly authority situated in the span between human expertise and computational power?
Associated concepts
Automatic correction guided by language models; Specialized machine translation for ancient languages; Summarization, segmentation, structure detection; Semi-automatic annotation and collaborative validation (human-in-the-loop); Advanced post-processing and human-machine hybridization
- Distant reading and automated cataloging
What becomes of cultural history when it is based on millions of interrogable witnesses and how to integrate quantitative analyses and macro-historical trends revealed by large-scale ATR?
Associated concepts
Distant Reading and macrohistory; Serialization of manuscript data; Studies on the dissemination of texts and manuscript traditions; Intellectual networks and the circulation of knowledge; Mapping of scriptural practices; Statistical approaches to linguistic transformations; Automated cataloging and expertise; Automatic generation of descriptive records; Fragment detection and virtual recomposition
- Scholarly publishing and textual criticism: rethinking critical editing in the age of abundance
Is critical editing destined to become an infrastructure rather than a finished product?
Associated concepts
From single critical editions to dynamic editions; Automatically generated variants; Large-scale multi-witness alignment; Exploratory rather than linear interfaces; New forms of scientific publication
Materiality and the History of Written Objects
- New computational paleography
How do massive corpora of transcribed texts transform our understanding of writings and scribes, and their evolution?
Associated concepts
Quantitative studies of scribal hands; Graphic stylometry and the evolution of handwriting; Attribution and identification of scribes; Interregional and interlinguistic comparative study; Recognition of workshop practices
- Text, materiality, and image
Beyond text! how to reintegrate materiality into an environment dominated by textualization?
Associated concepts
Joint text-image analysis; ATR correlation and media analysis (paper, parchment, seals, coins); Detection of layout, marginalia, diagrams; Linking material studies (XRF, multispectral, damaged manuscripts, etc.) with ATR
Heritage, Epistemological, and Ethical Issues
- Epistemology and ethics
How can we conceive of a scholarly critique of automation?
Associated concepts
Linguistic, geographic, and scriptural biases; Invisibility of minority scripts; Authority effects of automatic systems; Transparency and explainability of models; Redefining scientific practice
- Curatorial challenges
How can ATR be incorporated into a sustainable heritage policy?
Associated concepts
Infrastructure for the management, archiving, and availability of automated "noisy" or intermediate-quality ("silver standard") transcriptions Versioning of transcriptions; Versioning of transcriptions; Traceability of corrections; Sustainability of models and technological dependence; Ecological and computational cost of models; Prioritization of digitization/preservation campaigns
Submission guidelines
We welcome proposals from the fields of archival science and library science, digital and computational humanities, computer science, philosophy, and philology, for:
- individual presentations (20 minutes presentation / 10 minutes discussion),
- thematic panels (3 presentations),
- lightning talks or posters.
Communications and publication
Authors of accepted papers will be invited and welcomed to Paris by Biblissima+.
For those who want, the articles will be published in a special issue of the Journal of Data Mining and Digital Humanities.
Requirements for abstracts
Individual papers: a 300-word abstract and a short biography (100 words).
Panels: a presentation of the panel (200 words) and individual abstracts for each speaker.
Important dates
Submission deadline: April 15, 2026
Provisional acceptance notification: April 21, 2026
Final paper submission: May 15, 2026
Conference: June 15–17, 2026