MIRABILE-ATLAS Fellowship in Digital Humanities of the Zeno Karl Schindler Foundation

MIRABILE-ATLAS
Fellowship in Digital Humanities
of the Zeno Karl Schindler Foundation

The Zeno Karl Schindler Foundation [www.zenokarlschindler-foundation.ch] has established two fellowships, each of them for six to twelve months, to support the development of the «MIRABILE. Digital Archive for Medieval Latin Culture»  (www.mirabileweb.it) project maintained by SISMEL (International Society for the Study of Medieval Latin, Via Montebello 7, I-50123 Florence;  www.sismelfirenze.it). It is intended that the fellows will carry out their fellowships at the seat of the SISMEL following the development agenda of the MIRABILE project, with a monthly stipend of CHF 2,500 (Swiss francs).

Applications should be submitted (with curriculum vitae and motivation letter) by July 20, 2026 via email to: Prof. Jean-Yves Tilliette here.

The results of the evaluation procedure will be published by August 31, 2026.

Winners will be given the opportunity to start the fellowship from October 1st, 2026.

 

MIRABILE-ATLAS Fellowship core objectives

1. On-line publication, with related search functions for the search and visualization, of data present in MIRABILE on an atlas of medieval culture with related cross-searches and interactive navigation over the data retrieved from the database. All data will be transformed in JSON, which is the standard format for the management of data used by Elastic Search for indexing.


2. Integration with information present in MIRABILE’s databases (such as the author, the texts, the manuscripts and the bibliography). The module will provide the ability to show those database informations on a map and will be improved in data visualization and also in the data representation on the map. The aggregation of all the data in the map needs to be clear to the user in order to reach the information it’s looking for. In order to present information about manuscripts, authors, and works, distinct visual elements will be used to represent each on the map, indicating their presence geographically.


3. Record linkage and human validation across MIRABILE’s databases. We are building an automated pipeline to connect records that describe the same individual across MIRABILE’s authority files (in particular the BISLAM repertory of authors and the NOMI archive of persons), even when names diverge sharply across languages and spellings. Name normalisation, string- similarity scoring, and a context-aware language model propose candidate matches with graduated confidence, but no automatic stage establishes truth on its own: every match enters the scholarly record only once a scholar has reviewed and ratified it. We therefore seek scholars to act as this decisive human stage – examining each proposed match together with its score, the model’s reasoning, and any risk flags, and approving, rejecting, or annotating it by expert judgement. All decisions are recorded in an append-only provenance log, so that every connection is traceable and reversible; only ratified matches are exported and shown on the MIRABILE-ATLAS, where authors, works, and manuscripts appear as distinct visual elements. The machine narrows the search space, the scholar decides.


4. A user interface that will integrate the search engine results, will be implemented in the project, providing a way to view the data on an interactive map.

People with the following characteristics can apply for one of the following profiles or for both, by indicating it explicitely


1. First profile

Ability to understand and process data from database to JSON format and link them to metatextual information present in MIRABILE (with knowledge of cartography)

  • Good knowledge of Latin
  • Knowledge of authors, texts and manuscripts of medieval Latinity
  • Knowledge of the problems concerning the transmission of medieval texts
  • Personality traits: ability to operate in two distinct contexts: that of philology, codicology, and medieval latin literature and that of AI, with aptitude for verification when dealing with results proposed by AI
 

2. Second profile

  • Knowledge of JSON format
  • Knowledge of geo-location service
  • Knowledge of Elastic Search Engine
  • Skills in the use of AI, in particular, writing effective prompts and iterating them, recognising hallucinations, understanding the functionality of an MCP server
  • Good written and oral English
     

    For further information:

    Presidenza Sismel mail here

    http://www.sismelfirenze.it

    https://atlas.mirabileweb.it/schindler/atlas

     

    Applications should be sent to:

    Mail to Jean-Yves Tilliette here

    www.zenokarlschindler-foundation.ch

     

    THE FOUNDATION WILL NOT GIVE ANY INFORMATION, JUSTIFICATION NOR REASON FOR THE DECISIONS CONCERNING THE ACCEPTANCE OR NOT OF AN APPLICATION.

    Please consider that no reply will be given to such requests.

    Affidavit