Missions

As part of the PRAIRIE cluster, the History of Astronomy team at the LTE laboratory of the Paris Observatory is offering a two-year postdoctoral position dedicated to interdisciplinary research at the interface of artificial intelligence and digital humanities.

Specifically, 80% of the postdoctoral researcher's time will be devoted to research conducted within the framework of the EIDA project, while 20% will focus on promoting, at both the national and international levels, initiatives aimed at structuring and expanding key research communities working at the intersection of AI and the humanities.

EIDA (ANR-22-CE38-0014) seeks to develop a new critical approach to astronomical diagrams produced between the 8th and 18th centuries, originating from Chinese, Sanskrit, Arabic/Persian, Greek, Hebrew, and Latin sources. This interdisciplinary project brings together historians of science and computer vision researchers to design algorithms specifically tailored for the extraction and analysis of these historical diagrams.

EIDA considers these diagrams both as visual heritage and as tools for astronomical or mathematical reasoning within the astral sciences. Large-scale study of these materials requires fine-grained automatic analysis of their content, which is often disconnected from their visual appearance and for which no satisfactory AI-based solution currently exists. The project's ambition is therefore to propose and implement a radically new approach to the study of premodern and early modern astronomical diagrams by developing innovative, domain-specific computer vision tools.

The two main objectives of EIDA are:
1. To develop a specific critical framework for astronomical diagrams and define standards for their native digital critical editions, taking into account the diversity of their functions and modes of circulation;
2. To extend this innovative analysis to an unprecedented Afro-Eurasian scale by developing new computer vision approaches capable of decomposing an astronomical diagram into semantically meaningful elements for analysis and editing, without relying on human-annotated datasets.

EIDA combines recent advances in “analysis-by-synthesis” approaches in computer vision with the “visual” and “practical” turns in the history of astronomy, in a deeply interdisciplinary framework. This collaboration benefits from the growing availability of data, enabling large-scale, refined analyses that open access to historical insights previously out of reach.

In line with the principles of open science, EIDA will also develop open-source tools for the collection, study, critical edition, and visualization of diagrams through a web-based interface inspired by the DISHAS platform dedicated to astronomical tables. This new platform will provide the scholarly community with access to the tools used to produce EIDA's results, and the release of the source code will facilitate the further development of these methods.

Aligned with EIDA's goals, the research conducted by the selected candidate will contribute to deepening our understanding of the documentary, epistemological, and broader intellectual and cultural roles of astronomical diagrams. This research will focus on a well-defined corpus of manuscripts and/or printed sources chosen by the postdoctoral researcher. The sources may come from — but are not limited to — one or more of the project's central areas: Chinese, Sanskrit, Byzantine, Arabic/Persian, Hebrew, and Latin sources. Proposals focusing on materials in other languages, especially from Central Asia, the Caucasus, or Africa, are also welcome.

Activités

Building on their own research project, the selected candidate will take an active part in the collective work of the EIDA project by contributing to the development and dissemination of new artificial intelligence tools applied to the history of astronomy. This collaboration will take place within the project team, which brings together historians of astronomy, computer vision specialists, and digital humanities engineers, both at the Paris Observatory and at other partner institutions.

One of EIDA's core research areas concerns the plural definition of image similarity. This exploration goes beyond representations generated by computer vision and investigates broader methods for establishing dynamic and meaningful groupings of images. Such groupings may be influenced by documentary context, astronomical content, or visual styles in which these images were produced.

A second key research focus of EIDA concerns the potential features of a natively digital critical edition of astronomical diagrams. In practice, the selected candidate will contribute actively to these inquiries within the team. This may include leading dedicated seminars or coordinating specific collaborative tasks within the EIDA framework.

The postdoctoral researcher will work within the project team at the Paris Observatory and devote 80% of their research time to their own project within this collective, international program. They will take part in joint EIDA publications and will be encouraged to organize scientific events related to their research (workshops and seminars). The postdoctoral researcher will also have access to dedicated research funds, including support for travel to specialized libraries or relevant conferences.

In addition to the research activities described above, around 20% of the postdoctoral researcher's time will be devoted to research coordination, through the supervision of two key initiatives:
1. Co-direction of the DHAI Research Initiative (Digital Humanities / Artificial Intelligence) and the DHAI Seminar
• Organization of a monthly seminar dedicated to the intersection of AI and the humanities.
• Co-organization of two intensive doctoral training weeks on AI and the humanities.
• Encouragement and support of DHAI-related initiatives in the Paris region.
2. Coordination of the ELLIS Paris Unit
• Participation in the monthly meetings of ELLIS coordinators to ensure alignment with the broader ELLIS network.
• Assistance in coordinating the activities of the Paris unit, including facilitating and promoting ELLIS-related events in the region.
• Exploration of the creation of a specific program on Digital Humanities and Artificial Intelligence (DHAI) within ELLIS, with the aim of strengthening the unit's engagement in this interdisciplinary area.

These responsibilities will provide the selected candidate with the opportunity to play an active role in structuring and expanding research communities situated at the interface between artificial intelligence and the humanities.

Compétences

Candidates must hold a PhD in the history of science, history, or art history. A solid background in the analysis of primary sources, particularly those related to the history of astronomy, mathematics, or art, is highly valued.

In addition, a demonstrated interest in and willingness to engage with digital humanities, computer vision, and artificial intelligence tools is essential.

Excellent written and oral communication skills in English are required to ensure effective collaboration within the project's international team.

Contexte de travail

The position will be based within the History of Astronomy team of the Laboratoire Temps-Espace (LTE) at the Paris Observatory.
Further information can be found at the following address: https://syrte.obspm.fr/spip/science/histoire/.

This position falls within a sector covered by the Protection of Scientific and Technical Potential (PPST) regulations. Consequently, your appointment will require authorization from the competent authority of the French Ministry of Higher Education and Research (MESR) before your arrival.

Contraintes et risques

Nothing.