[Report] Seals and Linked Open Data, December 2nd to 8th 2020

From 2 to 8 December, the symposium Seals and Linked Open Data was held as part of the online Linked Pasts 6 conference. The symposium was organised by John McEwan, Associate Director of the Center for Digital Humanities at Saint Louis University. Three webinar panels during the event brought together several experts on sigillography from research institutions, museums and archives who presented various projects, databases and ontologies on sigillographic data.

Starting from the basic idea of considering seals as an important form of pre-modern communication, they can also be regarded as Big Data – an assumption that is confirmed in view of the quantity and heterogeneity of preserved seals and (printed) seal catalogues. The primary goal of most of the projects presented at the event was therefore to homogenise the sigillographic data and publish it in a machine-readable form. To achieve this, it was considered important to include older, non-digital projects in the form of published seal descriptions, unpublished card catalogues and pre-digital taxonomies. Homogenising these various earlier approaches to seal description would allow seal data to be made publicly available in a standardised way, which in turn opens up promising possibilities for applying data-driven research methods to this particular historical source.

This trend in seal studies is also of great importance for the study of heraldry in general and for our own project Coats of Arms in Practice in particular. Since seals are one of the most important sources of heraldic evidence, standardised access to seal descriptions would be beneficial for finding coats of arms on seals and researching them in the context of their historical use. This need is recognised by pre-digital sigillographers and thus by most of the projects presented at the symposium.

One example is the pre-digital taxonomy Vocabulaire Internationale de la Sigillographie, which defines standardised multilingual terms for describing seals and already includes a concept for seals showing a coat of arms. This taxonomy has been transformed to RDF/SKOS by Georg Vogeler. His work is already used in a database on the seals of the bishops of Salzburg. The structure of the mentioned vocabulary allows to easily filter out seals with coats of arms – of which more than 400 are included in the database.

Another important resource for the study of heraldry is provided by the Sigilla database (presented by Laurent Hablot and Philippe Jacquet), which aims to bring together seal collections from all local archives in France. Both persons and coats of arms are modelled as singular entities, allowing the linking of multiple seals representing one coat of arms. The coats of arms are even blazoned, although not in a structured, machine-understandable form. This data will also be made available via an application programming interface (API) in the near future.

The Digisig database (presented by John McEwan) draws from several analogue resources, including seal catalogues and seal and matrix collections, mainly focused on medieval Britain. The aim was to provide a system that allows searching across multiple seal collections with different reference systems. These underlying descriptions were made machine-readable and homogenised with a new singular data model. The database contains over 42,000 seals, more than 12,000 of which have coat of arms motifs.

These few examples illustrate a great awareness of the importance of standardised access to seal data. Further linking and joint integration of data is intended by the sigillographic community. Since heraldry also relies heavily on large amounts of source material, it will benefit greatly from this development.

In case you missed the event, all lectures were recorded and made available on YouTube.