Signed, sealed, delivered – digital approaches to Byzantine sigillography

2.1 Why seals?

Seals are coin-like objects mostly made of lead, whose two sides display iconographic depictions, inscriptions, and/or monograms. They were used to secure and validate the private and official correspondence of the Byzantines.

Seals are the only remnants of the written documents used for daily administration and correspondence in the Byzantine Empire, whose public and private archives are almost entirely lost. As such, seals do not supplement archival material but have to substitute for it, which explains the unique importance of sigillography for Byzantine Studies. Each seal is an object of study in its own right, but only serial analysis allows the artefact’s potential as a historical source to be fully exploited.

Figure 1:

Seal of Michael, vestarches and oikonomos of the Nea, mid-11th century (Sopracasa Collection, 53).

According to current calculations, approximately 100,000 lead seals have survived, most of them preserved in the Dumbarton Oaks Research Library and Collection (Trustees for Harvard University) in Washington DC (17,000), the State Hermitage Museum in St. Petersburg (12,000), and the Département des monnaies, médailles et antiques of the Bibliothèque nationale de France in Paris (7,000). As such, seals provide a unique corpus of textual evidence which is steadily increasing in volume, but not yet in accessibility.

2.2 Impact on Byzantine Studies

Sigillography contributes to the advancement of various research fields within the broader domain of Byzantine Studies. It stands as the cornerstone of the prosopography and social history of the Byzantine empire: each individual seal gives us access to one of its citizens, and a significant number of individuals are only known from the seals they left [1]. The study of individuals, families, and networks has greatly benefitted from the testimony of seals, as exemplified by their role in the main Byzantine prosopographical projects, the Prosopography of the Byzantine Empire (http://www.pbe.kcl.ac.uk), the Prosopographie der mittelbyzantinischen Zeit Online (https://doi.org/10.1515/pmbz), and the Prosopography of the Byzantine World (https://pbw2016.kdl.kcl.ac.uk).

Seals give insights into various aspects of the economic history of Byzantium, such as public land management, economic and fiscal administration of civil and ecclesiastical institutions, commercial routes and the circulation of goods. Given that seals mention honorary titles and offices of the civil, military, and ecclesiastical administration, they offer the opportunity to reconstruct individual careers and, collectively, provide the clearest picture of the civil, military, and ecclesiastical administrative apparatus, both central and provincial. Their testimony is invaluable for filling the gaps in written sources and reconstructing administrative geography and political history. Moreover, the potential of the iconography depicted on seals for the development of Byzantine art history has been scarcely tapped, even though Byzantine seals offer the largest corpus of religious imagery. Finally, the legends of seals are of utmost interest to Byzantine philology, not only because they document linguistic practices not attested in literature, but also, conversely, because thousands of them convey versified legends illustrating the development of medieval Greek poetry. Sigillography is therefore a seminal discipline within the broader field of Byzantine Studies.

3.1 Shortcomings in Byzantine sigillography

Due to the scattered scholarly landscape, different traditions and editorial practices in Byzantine sigillography have evolved, thus leading to several structural problems within the field.

First of all, there is only limited access to the research material. Unpublished material greatly exceeds the published one and remains hard to access, since it is spread widely across various public and private collections. Moreover, existing paper publications are not readily available, are not cost-effective, do not allow for updating, amending, and improving, and in most cases offer only low-quality images which do not reflect the material properties of the objects and in many cases provide only limited readability.

Besides, edition criteria are often inconsistent: standards for the scholarly publication of Byzantine seals were established during the 20th century by Vitalien Laurent [2], [3], George Zacos [4], [5], Nicolas Oikonomides [6], [7], [8], [9], [10], Jean-Claude Cheynet [11], [12], and Werner Seibt [13], [14]. There is, however, no agreement regarding epigraphical rendering, the normalisation of language variants of Byzantine Greek, consistent use of the Leiden conventions, iconographic analysis, and physical description. This lack of consistency hinders the cross-referencing of similar materials across different corpora.

There is also at present little to no training in Byzantine sigillography within an established academic framework, which limits the creation of a new generation of sigillographers, a key factor in the survival of the discipline. One of the very few exceptions in this respect is the curricular, practice-oriented research seminar in Byzantine sigillography held since 2015 at the Department of Byzantine and Modern Greek Studies of the University of Cologne.

3.2 Lack of digital solutions

In recent years, the Humanities have increasingly benefited from specialised digital solutions to address many of the aforementioned (and comparable) problems. During the last two decades, disciplines such as numismatics, epigraphy, and papyrology have developed unique digital approaches to their material, and their online corpora have grown rapidly; a comprehensive and up-to-date list of projects in Digital Classics is given by the Digital Classicist Wiki (https://wiki.digitalclassicist.org/Category:Projects).

Sigillography, too, has recently received increasing attention from experts in the Digital Humanities, particularly Western medieval seals. Projects such as Sigilla (http://www.sigilla.org) and DigiSig (https://www.digisig.org) are meant to collect and unify data coming from various collections based in France and the UK respectively, delivering online meta-catalogues. For Byzantine sigillography the only consistent online presence is the Catalogue of Byzantine Seals at Dumbarton Oaks (https://www.doaks.org/resources/seals).

Praiseworthy though these initiatives may be, the solutions they offer – both on the sigillographic and the digital level – do not yet match the potential of the subject. A fully digital approach to sigillography, i.e., a way to think of the seal as a digital object, is yet to come. What is needed now is not another isolated online catalogue or database, but something further: a common standard and a shared sustainable way to work digitally on various collections to ensure the interoperability of data amongst and outside the encoded corpora. In short: a digital turn of the field as a whole is required.

4.1 DigiByzSeal

The DFG-ANR project DigiByzSeal focuses on the scholarly edition and publication of four major but as yet unpublished collections of seals: the collections of the Institut Français d’Études Byzantines, the former George Zacos collection of the Bibliothèque nationale de France, the private collection of Robert Feind (the largest collection of Byzantine seals in Germany), and the collection of the Carthage National Museum in Tunisia, as well as the digital enhancement of several paper-published corpora (ca. 4,000 seals).

The digital objectives of the project include the further development of SigiDoc as a standard for encoding seals and the creation of a digital ecosystem for Byzantine sigillography, using the encoded corpora as a critical test mass to build a centralised sigillographic portal allowing for global cross-corpus search, and, eventually, a centralised hub for Byzantine sigillography.

Furthermore, the reading and presentation of individual seals is enhanced through Reflectance Transformation Imaging (RTI). In addition, we will explore a range of resources to draw on external information, such as geodata and prosopographical data, with the aim to make each item available as Linked Open Data.

4.2 DiBS

While the DFG-ANR project DigiByzSeal focuses on the more technical aspects of sigillography and its enhancement through digital methods, DiBS adds another important component in strengthening this rare subject, namely training and teaching. Responding to the call World Knowledge – Structural Support for ‘Rare Subjects’ of the Volkswagen Foundation, the project is geared towards the implementation of a strategic concept for ensuring the sustainability and accessibility of Byzantine Studies at the University of Cologne through the integration of digital tools and practices. Although DiBS takes a broader perspective on Byzantine Studies as a whole, most of the aims of DiBS also hold true for Byzantine sigillography.

Building on an excellent working synergy between the Department of Byzantine and Modern Greek Studies and the CCeH, the strategic measures envisaged in this project aim to create a sustainable interdisciplinary bridge between the historical-philological discipline of Byzantine Studies and the Digital Humanities through the creation of a digital infrastructure for research-based teaching in Byzantine Studies.

The strategy is based on a number of measures which consist of a combination of research, the crafting of a digital infrastructure, the strengthening of human and material resources, the introduction of new, pioneering teaching formats, internationalisation, and transfer of knowledge to non-academic institutions. It will allow us to promote awareness of the rare, multicultural, and trans-national discipline of Byzantine Studies for students at both the undergraduate and graduate level, but also among scholars in other historical and cultural disciplines, as well as to promote its advancement in Cologne and on an international level, within and outside the academic sector, while fostering methodologies and working practices from the Digital Humanities.

5.1 Reflectance transformation imaging (RTI)

RTI is an advanced but affordable imaging technology widely used in the field of cultural heritage. With RTI, an object is photographed multiple times with lighting from different angles. These images are then computationally processed and presented in a way that allows the viewer to virtually move the light source on-screen, thus revealing the finest and most subtle structure of the object’s surface. Legibility and analysis of visual features can thus be significantly improved.

Compared with other advanced imaging techniques such as photogrammetry or 3D scanning, RTI is especially well suited to seals, due to its superior resolution of surface texture while at the same time providing the highest cost/benefit ratio [15].

Within DigiByzSeal we use Dome RTI, a variant of RTI that uses a fixed array of LEDs instead of a moveable external flash operated manually, allowing for significant savings in time and workload (Figure 2).

Figure 2:

The CCeH RTI dome.

In an earlier cooperation, the CCeH built a custom RTI dome, which had already obtained highly promising results during earlier test runs on a number of Byzantine seals [16], [17], [18]. RTI has proved to be an ideal imaging tool for damaged, corroded, and flattened seals, as long as the relief-like structure of the inscription or a faint outline of the depicted image is still discernible on the surface. During the analysis of the seal with the RTI viewer, each sign or letter can be focused by dynamically varying light angles; at the same time, the images can be enhanced through various display modes (e.g., diffuse gain, specular enhancement, normals visualisation). This allows for the observation of specific epigraphic or iconographic details which would be impossible to see with the naked eye or with a high-resolution image. Even if this does not lead to a complete interpretation of the inscription, it prompts further research on the object of study. In this respect, RTI applied to sigillography opens interesting prospects for the analysis of specimens hitherto neglected because of their poor state of preservation [16]. Moreover, from a curatorial point of view, RTI images enable scholars to study the seal in greater detail without excessive handling of the object, thereby extending its lifespan (Figures 3 –6).

Figure 3:

Image of SB-66 (Robert Feind Collection) captured with a Canon EOS 300D.

Figure 4:

RTI image of SB-66 (Robert Feind Collection) displayed with the rendering mode “diffuse gain”.

Figure 5:

RTI image of SB-66 (Robert Feind Collection) displayed with the rendering mode “specular enhancement”.

Figure 6:

RTI image of SB-66 (Robert Feind Collection) displayed with the rendering mode “normals visualisation”.

5.2 Workflow

Within DigiByzSeal we are currently implementing a workflow for RTI imaging, establishing the first systematic application of RTI technology to a collection of Byzantine seals and setting the pace for a systematic use of this technology to unlock previously lost information. However, since we want to provide a homogeneous corpus of digital objects, we will apply RTI not only to seals with a high degree of damage or corrosion, estimated at 20 % of the sample, but to all seals of the four collections that are part of the DigiByzSeal project (ca. 4,000 objects).

The workflow comprises the preparation of the seals such as positioning them under the dome, taking one or more test exposures to determine suitable capture settings (such as aperture and exposure time), capturing the RTI series consisting of 60 images each from a different lighting angle, and processing the images into relightable RTI images suitable for viewing using the RTIViewer software (and, prospectively, for exporting them into interactive web versions that can be integrated into the collection website).

For capturing the images, a custom capturing software has been developed by the project team that communicates with both the camera and the controller of the RTI Dome, providing a user-friendly interface for researchers to allow for a fully automated, efficient tethered capture process specific to the project’s needs. After capturing, the images are processed using the Relight software suite developed by the Visual Computing Lab of the Istituto di Scienza e Tecnologie dell’Informazione (CNR-ISTI), which is in active development and provides a more modern, user-friendly, alternative to the prior de facto standard RTIBuilder (https://github.com/cnr-isti-vclab/relight). Since the beginning of the project, the workflow has been continuously refined and optimized in order to allow each seal to be captured and fully processed in about 2–3 minutes.

The resulting RTI files can then be taken as a starting point for the editorial process within the project. Besides, the RTI images can also be made available to the public through the presentation interfaces of the respective collections (see Section 7), allowing other scholars to re-examine the seals under the conditions used by the editors.

Furthermore, we currently evaluate the requirements for providing all image data via an API. The current state of the art for image provisioning is IIIF, which could also help in tackling size and processing overheads. But since the integration of relightable imaging techniques like RTI is not yet fully supported by IIIF, this needs further exploration [19].

6.1 SigiDoc

SigiDoc represents the extension of EpiDoc, the digital approach already applied to inscriptions, coins, and papyri, to Byzantine seals. SigiDoc was initially envisioned, about two decades ago, by Charlotte Roueché, but digital tools were then lacking; in 2015, Alessio Sopracasa resumed the work on a new basis within a Marie Curie Fellowship at King’s College London under the supervision of Roueché and Gabriel Bodard, one of the main architects of EpiDoc.

Since 2017 the development of SigiDoc has been carried out in cooperation with the University of Cologne. Work has focused on the core elements of SigiDoc, such as the conversion of the sigillographic standards into a digital form, and on the customisation of EFES (EpiDoc Front-End Services), a platform for the online publication of ancient texts in EpiDoc [22] (https://github.com/EpiDoc/EFES), and a fork of the open source XML publishing platform Kiln of the Department of Digital Humanities at King’s College London (https://kiln.readthedocs.io).

6.2 Technical features of SigiDoc

SigiDoc is a clean subset of EpiDoc and provides XML-based and TEI-compliant encoding standards for the edition of Byzantine seals: as such, it aims to establish a set of guidelines for publishing Byzantine seals in a digital form. SigiDoc is intended both for the creation of digital-born editions and the digital enhancement of paper publications.

SigiDoc is largely based on the ongoing experience of EpiDoc, the international, collaborative effort using a subset of TEI, which has established itself as the most robust and widely supported format for encoding editions of ancient texts on a wide range of text-bearing objects [23] (https://github.com/EpiDoc). As good scientific practice and as a guarantee of success, we have opted for adapting and developing an already existent and reliable solution, allowing us to connect the sigillographic data to those coming from other disciplines.

6.3 Current achievements

SigiDoc aims to become the standard tool for the digital encoding of Byzantine seals. The developers designed a set of standards to encode and digitally publish Byzantine seals, and customised EFES for web visualisation and data valorisation. In 2018 SigiDoc became one of the pilot projects using and testing EFES, actively contributing to its development [24], [25], [26].

At the time of writing, SigiDoc consists of a set of encoding guidelines (http://sigidoc.huma-num.fr), a schema, a template, a stylesheet, and a highly customized EFES instance (the source code is available here: https://github.com/SigiDoc/SigiDoc). In late 2021 a first attempt at an online corpus encoded in SigiDoc was published: https://sigidoc.raketadesign.com (Figure 7).

Figure 7:

Homepage of SigiDoc’s test corpus.

During the first year of DigiByzSeal, SigiDoc has been continuously refined thanks to the preliminary encoding of a critical mass of seals (an estimate of ca. 1,000 specimens between the German and French teams). This has led to the further finetuning of the EFES instance, with the creation of new indices and of new search filters, in order to perform fine-grained queries which were hitherto not possible. The new and improved search filters are (in order of appearance): institution and collection (search terms related to the seal’s location), persons, personal names, family names, place names, dignities, civil offices, ecclesiastical offices, military offices, and titles (search terms related to the issuer of the seal), Marian terms, Christ-related terms, saints-related terms, and the grammatical case (search terms related to the legend of the seal), and finally iconography (Figure 8).

Figure 8:

The EFES-based search portal with string search, lemmatised search, date slider, and search filters.

Thanks to the invaluable feedback from the project members who are actively carrying out the encoding, we have been able to adjust the structure of the edition fields and refactor the encoding to make it less cumbersome. In addition, thanks to the contribution of our partners, we have laid the groundwork for the creation of default multilingual corpora: in addition to English, German and French, the core languages of the project (Italian, Modern Greek, and Bulgarian) have now been added. Controlled vocabularies as well as shared authority files for the creation of dynamic indices and a faceted search will also be added, thus triggering a new release of the source code (v1.1).

7.1 Database backend and authoring environment

In previous work we have established a set of data shared among all sigillographic projects that use SigiDoc for encoding their seals, consisting of a set of XML schema files, XSL stylesheets, TEI-encoded authority files, several controlled vocabularies, and unique IDs.

For supporting the consistent encoding of seals in SigiDoc within and beyond the DigiByzSeal project, we are currently implementing an authoring environment that provides a set of entry forms based on the current SigiDoc schema that can be continuously adapted to the further evolving encoding standard, as we expect further specifics and peculiarities to emerge during the sigillographic investigation carried out within DigiByzSeal, which are fed back into SigiDoc to continuously refine the standard.

On the one hand, the authoring environment is a means to ensure the homogeneity and consistency of the produced data. While experienced sigillographers can easily encode seals in XML directly, the process of revision, correction and/or extension of already encoded datasets can still be very cumbersome and time-consuming. In a collaborative setting, an authoring environment facilitates mutual corrections as well as subsequent additions by allowing individual input fields to be specifically checked.

On the other hand, the authoring environment creates simplified access for data entry without the need for prior training in SigiDoc, such as for curators in museums or private collectors who enter data only selectively and in a decentralised manner, like in the case of individual items or small sets of specimens. At the same time, the use of such an editing environment forms a very good starting point for training and teaching scenarios, since the entry forms help to illustrate the underlying XML structure and provide guidance and support for data entry.

Due to the multifaceted, multilayered, and multidisciplinary forms of representation of the seals and texts in the XML data (transcription, edition, translation; text, apparatus, commentary; physical, iconographic, epigraphic, and textual descriptions), we decided to create a custom solution based on a BaseX database providing APIs in combination with a Vue.js Javascript frontend to create an environment that efficiently supports the research staff (Figure 9).

Figure 9:

An early version of the SigiDoc editor with metadata of the seal of Michael, vestarches and oikonomos of the Nea, mid-11th century, depicted in Figure 1 (Sopracasa Collection, 53); the entry forms are derived from the current SigiDoc schema.

The authoring environment consists of two components: first, an editor that maps the XML fields of the current SigiDoc schema onto a Javascript-based entry form that allows users to enrich metadata without knowledge of XML; second, a live preview that provides a real-time rendering of the data resembling the way it will be displayed in the EFES-based presentation layer, thus giving the user immediate visual feedback on the editorial process.

Besides, it is possible to connect the XML editor Oxygen (https://www.oxygenxml.com) directly to the BaseX database via WebDav, providing an additional input method for experienced users. In this way, trusted editors can access and edit SigiDoc files stored in the underlying BaseX database or upload new files via Oxygen’s database connector. This allows the sigillographers to choose between two approaches for editing: using the tagless editor to edit and correct existing data or to work on the raw XML data.

In a further step, the infrastructure will prospectively be extended to include a module for entity reconciliation, for example based on the lobid-gnd (http://lobid.org/gnd) service provided by the Hochschulbibliothekszentrum des Landes NRW (HBZ), that interfaces with existing APIs wherever possible to facilitate the matching of entities with existing authority data. By these means, it will be possible to enrich the datasets by integrating information from external sources, such as geographical or prosopographical information.

7.2 Presentation layer

As a further building block for a comprehensive ecosystem for Byzantine sigillography, the DigiByzSeal project aims to provide an adequate presentation interface for hosting individual collections. Regarding the FAIR data principles, we decided to build upon an already existing solution, namely the EFES platform (see also Section 6.1) originally created for the online publication of ancient texts in EpiDoc, and to adapt it to the specific needs of SigiDoc.

EFES is already a central and useful software driving the process of SigiDoc metadata creation while simultaneously serving as an online presentation layer to the public. However, certain central shortcomings have to be addressed in order to make it suitable for everyday research, including an adjustment of search indexes and facets, the integration of bibliographical references and footnotes, and a public interface implementing the OAI-PMH protocol (http://openarchives.org/pmh) for enabling cross-corpus searches and allowing external harvesting and re-use of data.

To support the understanding and sensemaking of researchers in face of heterogeneous and ambiguous data, we will add several advanced visualisations that make use of the data produced and enriched in the editorial process: (i) an RTI viewer based on the open-source software suite Relight developed by the CNR-ISTI Visual Computing Lab (see also Section 5.2) to improve the legibility and analysis of visual features of the seals, e.g. specific epigraphic details; (ii) a date visualisation and filter, building on the already existing ‘date slider’ in EFES, inspired by partner projects such as Pelagios and GODOT; (iii) geographic visualisation features based on the interlinking of standard place identifiers to their coordinates, allowing for the display of place references from a specific set of search results in a map; (iv) provided that successful entity linking can be done, prosopographical information, particularly family relations and dignities, can be visualised as graphs, comparable for example to Palladio (https://hdlab.stanford.edu/palladio-app).

7.3 Federated search portal

As mentioned above, one of the main shortcomings in Byzantine sigillography is the limited accessibility of resources due to their dispersal in scattered collections. The presentation of individual collections via EFES can thus only be a first step, one which must be complemented by a central portal website hosting a sigillographic search engine leveraging the OAI-PMH-interface that will be added to the EFES-based presentation environment utilised by the collections to enable a federated cross-corpus search, thus addressing the current problem of the findability and accessibility of published Byzantine seals as research data.

As a key infrastructural effort, we will implement a central hub for Byzantine sigillography with faceted browsing and full text search over distributed collections (comparable to the CLARIN-VLO, Virtual Language Observatory). Building upon the metadata harvesting interface (OAI-PMH) of the individual EFES instances, the metadata from the respective corpora and collections can be made available to both the federated search platform and potentially other metadata aggregation services relevant to the field. Furthermore, in order to meet the specific needs of Byzantine seals, we will add support for the AthenaRuby font (https://www.doaks.org/resources/athena-ruby), allowing for search functionalities on character level, which will be based on the markup and epigraphical indexing of the diplomatic transcriptions in SigiDoc.

Prospectively, the portal will provide centralised access and search functionality not only for the collections edited during the projects, but over time it will allow other external collections successively to be added to the portal, e.g. those resulting from the training activities described in the next section.

8.1 University courses

We propose a series of strategic measures in close coordination between the Department of Byzantine and Modern Greek Studies and the CCeH to enhance the course offer in Byzantine Studies, aiming to promote collaborative digital approaches to Byzantine text-bearing objects (in particular to seals, coins, inscriptions, and manuscripts), providing students in Byzantine Studies with transferable digital skills, fostering interchange with the Digital Humanities, implementing internationally shared concepts of training in the field of Byzantine Studies, and establishing good practice in sustainable data management and reuse.

These measures include the implementation of innovative teaching and learning methods, the design of a new curricular module for BA and MA study programmes at the University of Cologne, and the internationalisation of the study offer.

Intending to be sustainable and accessible to the greatest possible extent, all training outcomes will be thoroughly documented. Consequently, we will implement a set of teaching skills by providing a pool of structured reusable materials for teaching and self-training.

At the time of writing, a study module called “Digital Edition of Ancient and Byzantine Text-Bearing Objects” has been implemented and will start from the Winter Term 2023–24. The module will be compulsory for students in the MA program in Byzantine Studies and it will be among the elective courses for students at BA and MA level within the degree program “Ancient Languages and Cultures” (encompassing a wide array of disciplines, from Classical Philology to Jewish Studies, from Egyptology to Ancient History), for students at BA and MA level in Information Processing, for students of the MA program in Medieval Studies as well as other neighbouring disciplines.

At its core, the module will include three courses with different formats and complementary objectives. A lecture series will deal with broad methodological questions regarding the application of the Digital Humanities to Byzantine Studies. This represents our best possibility of reaching a broad audience, as each lecture will be live-streamed and later released as a podcast. There will be an intensive one-week course which will start with a day-long kick-off meeting early in the semester where the teachers will lay the foundation for the work, both in terms of the subject and of the technologies which will be deployed. After this meeting, the students will enter the phase of asynchronous learning. This part is crucial, as it will enable the students to gain awareness of the subject and to be more actively involved in the face-to-face training, which will take place over one week after the end of the term and will mainly involve hands-on practice. The third and last format will be a digital crowd-sourcing encoding sprint, addressed not only to students, but also to cultural heritage specialists and the general public alike.

In an effort to establish an international teaching network in the field of digital Byzantine Studies, all academic courses offered within the scope of DiBS will be carried out as co-teaching with our international cooperation partners, depending on the academic profile of each partner and institution. For this purpose, an integrated system of in-presence and distance teaching is being established.

Starting in the Winter Term 2023–24, these courses will be accessible from a customized instance of EduLabs, a platform supporting innovative teaching formats that promote digital literacy and are freely reusable (https://edulabs.de). All teaching materials, consisting of slideshows and short video tutorials, will be designed as Open Educational Resources and will be distributed under a Creative Commons-BY license. The training materials will be stored in ORCA.nrw, an online portal for digitally enhanced teaching and learning at universities (https://www.orca.nrw).

8.2 Training activities

With a keen eye towards the future of the discipline and its sustainability, as part of scholarly networking in the field of Byzantine Studies and, more specifically, Byzantine sigillography, it is of paramount importance to engage museums and institutions in the field of cultural heritage, given that seals are often preserved in such institutions.

In the framework of both DigiByzSeal and DiBS we strive to ensure extensive knowledge exchange by working with curators in order to enable them to manage and enrich their holdings and to reach a wider public. At the same time, we have already started training scholars, curators, and students in the use of SigiDoc and EFES for their own research projects.

Several museums holding some of the finest collections of Byzantine seals worldwide (Dumbarton Oaks, Numismatic Museum of Athens, Museum of Art and History of Geneva, Regional Historical Museum of Shumen, Bulgaria) have agreed to participate in these projects as external collaborators. Curators and members of these institutions will attend SigiDoc training weeks in order to learn how to create a definite and stable digital record of their Byzantine seals, be it a digital-born publication of the collection or a digital enhanced version of the paper catalogues. Thanks to the skills acquired during the training weeks, the museums will have permanent access to the digitised collections and will enhance their curatorial catalogues while providing an appealing and accessible presentation of their collections to physical and virtual visitors alike.

The first three-day hybrid training of curators from three European cultural heritage institutions was held in March 2023: teaching subjects included semantic markup, SigiDoc encoding for metadata and text, and a basic introduction to data visualisation using EFES. A follow-up training session was held in October 2023 with the same curators, who in the meantime will have begun encoding their seals (estimated total: ca. 500), where the subject of study will be the use of TEI-encoded authority files for the creation of dynamic indices and an interoperable faceted search portal.

8.3 Dissemination

One of the key objectives of the DiBS project is to create a constantly growing knowledge base for Byzantine Studies, comprising relevant digital resources, information on events, teaching material and so on. For this purpose, we have set up a website based on WordPress (https://dibs.uni-koeln.de), which allows the project team and its collaboration partners continually to add and curate content relevant to the field. At the time of writing, the DiBS website has just recently been released and initially only provides some basic information about the project and the network of cooperation partners. On the home page, the Twitter account @digibyzantine is integrated as well as the option to subscribe to the DiBS newsletter (Figure 10).

Figure 10:

The DiBS project homepage.

For the future, it is planned to continuously publish the project news or articles related to the project in the form of blog posts. In addition, further menu categories will be introduced, e.g., to link learning and teaching resources and to list other relevant digital resources, as for instance the individual collections produced in the scope of the DigiByzSeal project. In principle, it is possible to integrate any other contents or functionalities, as the page is easily customisable.