The Social Life of Early Medieval Normative Texts

Project Information

SOLEMNE

Grant agreement ID: 101087979

DOI

10.3030/101087979

EC signature date 31 March 2023

Start date 1 September 2023

End date 31 August 2028

Funded under

• European Research Council (ERC)

Total cost

€ 1 955 000,00

EU contribution

€ 1 955 000,00

Coordinated by

STICHTING RADBOUD UNIVERSITEIT
Netherlands

Periodic Reporting for period 1 - SOLEMNE (The Social Life of Early Medieval Normative Texts)

Reporting period: 2023-09-01 to 2026-02-28

Summary of the context and overall objectives of the project

Early medieval Western Europe has left us a rich and varied corpus of medieval primary sources, with thousands of extant manuscripts containing works of biblical exegesis, historiography, poetry, educational texts and (late-)antique literature. These texts allow us a glimpse of the dynamic processes of European identity formation, the formation of cultural memory, religious conversion, cultural exchange, and the literary uses of the past. Canon law stands at the core of many of these social developments as medieval people tried to make sense of and structure a post-Roman, ‘European’ society.
The project ‘The Social Life of Early Medieval Normative Texts’ (SOLEMNE) studies the canonical collections (big and small) as sources for medieval society. It subscribes to the statement that canon law is a genre that intersects ‘with every aspect of medieval life and society’ [RENNIE, 2008] and recognises that the dynamic process of selecting and organising authoritative statements into certain combinations reflects the medieval compilers’ ambition to formulate their thoughts on social norms for a specific audience and a particular context.
The general aim of SOLEMNE is to chart and contextualise the spread of social norms as articulated in specific combinations of canons in a bottom-up approach starting from the vast corpus of manuscript witnesses of canonical compendia in every shape and form. The project has identified five goals:
(1) The creation of a sophisticated digital dataset of early and high medieval canonical compendia, charting their structure and content within their manuscript context.
(2) To provide a sophisticated digital environment visualising the essential elements of the canonical tradition up until Gratian (approx. 500-1200 CE) within their textual and material context.
(3) To ensure easy, lasting access to the essential elements of the canonical tradition
(4) To study not only the ‘grand’ collections, but to be exhaustive in its range of extant collections (large and small, sophisticated and ‘unstructured’, widely-spread and unique collections).
(5) To break new ground by addressing a set of broader questions on narrative strategies to multiple canonical sources.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The SOLEMNE project has a dedicated website (canones.org) on which it publishes its news. To date, the project has achieved a number of important results:

a. In pursuit of goal 1, the SOLEMNE team has compiled a sophisticated dataset of canonical works and collections, drawing on earlier research outputs (transcriptions and editions), online resources provided by international colleagues, older printed publications, and new manuscript transcriptions produced by team members. Formatted as document-based XML files in accordance with industry standards, the data is highly accessible for current and future research. Although the dataset continues to grow, its current scope — 415 works and 2,251,142 words — already constitutes the largest known digital corpus of early medieval canonical material. Work on cleaning the current dataset to a gold standard in anticipation of integration into the web application is now nearly complete.
b. Serving goal 2, the IT-support of the Radboud Institute of Culture and History (RICH) is developing a Django web application for easy access to the SOLEMNE dataset. Leveraging the OpenAI API, the application integrates an advanced text-overlap detection module. The tool has been tested on small data subsets and awaits the full dataset for large-scale evaluation.
c. To ensure durable and open access to the core elements of the canonical tradition (goal 3), SOLEMNE fosters collaboration with scholars of medieval canon law and disseminates project outcomes. SOLEMNE also contributes data to external repositories, where it is more appropriately integrated within established, domain-specific online databases. This includes over 38 detailed manuscript descriptions and 5 descriptions of canonical collections in the MGH's Clavis Canonum database (see: https://data.mgh.de/databases/clavis/wiki/index.php/Main_Page).
Following an inaugural workshop in January 2024 at Heyendael House (Radboud University), the SOLEMNE project continues to convene specialists in early medieval normative sources and digital datasets to exchange insights and define research directions. In March 2026, SOLEMNE teamed up with the IMAFO project at the Österreichische Akademie der Wissenschaften and organised a two-day workshop on detecting text-similarity with AI in historical sources (see: https://www.oeaw.ac.at/imafo/veranstaltungen/detail/ruse-of-reuse). In addition, project members frequently contribute to (competitive) conferences on digital approaches to historical texts and/or medieval canon law.
d. The SOLEMNE project is dedicated not only to the study of the major canonical collections, but also to smaller and less structured compilations. In this context, three members of the team have undertaken the first in-depth analysis of a brief collection on marriage, known as the 'De ratione matrimonii'. A publication presenting an analysis of its form and content, together with its first critical edition and an English translation, has been accepted by a leading international journal. The project’s open-access budget will ensure that this work is widely accessible. Further studies of related collections are currently in progress.
e. In pursuit of Goal 5—breaking new ground by addressing broader questions of narrative strategy across multiple canonical sources—a pilot study employed a computational method (“perplexity”) to assess the extent to which biblical language and motifs are reflected in canonical texts. Testing a hypothesis formulated by Jean Gaudemet, the study found that the "biblicality" of canonical works indeed increased during the ninth century, suggesting that—at that time—biblical language served a marker of canonical authority. The methodology was presented at two international conferences and has resulted in an open-access publication.
In parallel, the team is exploring the application of topic modelling to the corpus, with a view to identifying thematic structures within canon law over time. As part of a research internship, the project has also developed a prototype chatbot for medieval canon law (“Canonist”), which will be integrated into the project’s web application. Building on this line of work, the team has been accepted to present its research at two competitive international conferences in Spring 2026.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The SOLEMNE project is grounded in the application of advanced Digital Humanities techniques to trace the dissemination of "clusters" of canons—combinations of texts transmitted together—rather than focusing on individual canons in isolation. The systematic compilation of a substantial corpus of canonical material now enables the use of increasingly sophisticated computational approaches, designed to illuminate patterns and relationships that would remain invisible through traditional philological methods.
Applying distant-reading methodologies to canon law allows for the detection of linguistic tendencies, thematic structures, and intertextual networks across the entire corpus. The interpretive value of these methods emerges when critically combined with contextual expertise in the historical, social, and intellectual dimensions of the canonical tradition—something the SOLEMNE project is exceptionally well equipped to do.
This methodological potential is exemplified by the pilot study on "biblicality". In this experiment, a language model was trained exclusively on scriptural corpora and tested on canonical texts to measure its predictive performance using perplexity, a quantitative metric of uncertainty indicating textual affinity with biblical language and register. The study demonstrates how computational modelling can contribute to long-standing questions regarding intertextual influence and the formation of canonical authority. The work was presented successfully at a conference for historical linguistics, benefiting from and contributing to interdisciplinary knowledge exchange.
A SOLEMNE supported internship has produced a legal "expert system" ("Canonist") thanks to the interdisciplinary collaboration with colleagues from Computer Sciences. Building on this success, the project is aiming at constructing a vector-based knowledge graph. Related to this, the SOLEMNE project explored methods for multi-author authorship detection, which is particularly relevant to the studies of canonical "forgeries", such as the ninth-century so-called False Decretals. Both initiatives contribute directly to the project’s overarching objective of understanding why certain canons achieved greater authority and circulation than others—beyond the sometimes misleading attributions of authorship traditionally invoked to explain such patterns.