A Digital Analysis of Common Law: Social Dynamics in Early English Justice

Essoins – Medieval English Court Rolls

The team at Bielefeld University is particularly interested in analysing the court rolls of the early English Common Law. There are two main aspects to the sub-project: 1. Digitization of different types of court rolls of the 13th and 14th centuries using HTR and OCR models; 2. Analysis of the specific role of essoins within the Common Law focusing on the impact of social categorization on legal procedure using NLP tools.

As part of the FLOW project, Bielefeld will contribute ground truth data to improve the language models used in HTR procedures for the Anglicana script used in the court rolls. We evaluate existing tools and standards for medieval HTR to make it easier for historians to integrate digital methods into their workflow.

On the basis of this, the Bielefeld team will try to enhance the output of NLP tools for the detection of named entities and employ this technology in the analysis of the social relations and structures imminent at court. A special focus will lie on questions of intersectionality and group formation as it is present in the court proceedings.

Navigating the Hanse: A Long-Term Analysis of Hanse Diets and Policy Making

Hanserecesse – Protocolls and Resolutions of Urban Meetings (1358-1669)

The project carried out at the Research Centre for Hanse and Baltic History (FGHO) in Lübeck delves into the extensive sources on the inner workings of the German Hanse, a voluntary association of merchants and towns shaping economic and political landscapes for centuries. The hanserecesse, spanning from 1358 to 1669, document the common voluntary decision making of towns located in the vast area from today‘s Netherlands to Estonia. These records covering about three centuries form the central source corpora on the history and development of the German Hanse.

Within The Flow, we examine the recesse with the help of digital applications and thus present a long-term study of the development of the German Hanse for the first time. Next to this study, we are working on the creation of suitable HTR and NER models for Low German manuscripts in order for other research to have easier to access the extensive source material of Hanse towns and their collaboration.

The Construction of the Sulṭān's Authority in the Arab Provinces: Sharia, Law Practices and Institutionalization in Jerusalem

sijillāt maḥkama sharʿīyya – Sijills. Shari‘a Court Records (16th and 17th centuries)

This sub-project focuses on the records (sijills) of the Shari‘a courts of Ottoman Jerusalem. These records, spanning about three centuries, constitute a major corpus of sources for the legal history of the Ottoman periphery. The sub-project analyses criminal cases, particularly those based on customary law, contributing to the debate on how law was understood and applied by various actors (judges, litigants, court officials) and how the Ottoman Sultan intervened in the lawmaking process. It also explores what law can reveal about social practices in contexts considered marginal to the central domains of the Ottoman Empire, especially the Arab province of Palestine, and contributes to the development of machine-learning-based workflows for sources in Arabic and Ottoman Turkish.

The project will construct a data set for the judicial corpora of Ottoman Jerusalem. Based on this data set, it will explore whether the frequency of particular kinds of cases or features of cases can be contextualized, providing insights into the underlying dynamics of Jerusalem’s pluralistic legal system. After an initial phase, in which text recognition using HTR models for the Arabic and Ottoman Turkish script used in the sijills is developed, the project will focus on the large-scale comparison of single criminal cases using NLP tools. This will advance our knowledge of a lesser-known chapter of Jerusalem’s legal history, offering us a glimpse of the transformation of judicial practices, and their complex interactions with social reality, over an extended period.

Analyzing Delinquency Cases in the Tower of Bern

Tower Books – Bernese Interrogation Protocols (1545-1745)

The sub-project focuses on questions of categorization of different cases brought to trial (and, in most cases, involving “peinliche Befragung,” that is, torture) in the tower of Bern. The project inquires if, from a textual perspective, frequency and type of delinquency can be contextualized, leading to insight into the inner workings of the legal system of one of the major city-states north of the Alps. The corpus-level perspective is deliberately chosen in order to avoid modern categorizations and to bring into focus instead the commonalities of single cases recorded in the tower books. This perspective takes discussions of alterity seriously that ask for an understanding of pre-modern pasts that is not biased by modern language. At the same time, it is coupled with recent approaches to the linguistic analysis of written documents.

Technically, we will, after an initial phase involving text recognition, focus on large-scale comparison of single cases. The tower books will be made accessible in full-text, and by topic occurrence via adapting existing methods. Topic modeling and comparisons of vectorized text parts (understood as an NLP task), in particular, have so far only infrequently been used for pre-modern textual data that has not been normalized. The labeling of similar cases, yielding categorization, will be based on the close reading of single, “typical” (statistically determined) cases. This will show which cases are highly correlated, and thus could potentially involve similar offenses based on what accusations were brought.