From Iconology to Digital Image Studies
2022-2023 Artl@s/Visual Contagions Research Seminar.
Organizers: Béatrice Joyeux-Prunel (UNIGE), Catherine Dossin (Purdue University), and Nicola Carboni (UNIGE)
In the framework of the project Visual Contagions on the global circulation of images, the 2022-2023 Artl@s international research seminar continues its conversation on the notion of “style”. To better understand this ubiquitous notion, the seminar will further the dialogue begun in 2021, between art history and recent computational approaches, between aesthetics and cognitive approaches, between research, analytical distance, and common-sense.
This year's contributions pertain to various domains, from art history to digital visual studies, from literary studies to data science, from computer vision to cognitive sciences, from traditional approaches to computational ones. Participants at all stages of their careers are welcome.
Questions we aim to address and explore include, but are not limited to, the following:
- What do we mean by “style”? What constitutes a style? Is it a transhistorical and transcultural concept? How have different periods, traditions, cultures, etc. used this notion – if they have used it at all? Are there alternative, understudied historiographies of the notion of style? How have other disciplines, such as literature, architecture, cinema, or dance, conceptualized style? Such alternative approaches of style might help us think globally in order to revisit, enrich and renew our existing concepts – but to which extent?
How should we approach “styles?: When talking about style, are we making a de re or de dicto statement? Are there tangible qualities that we use to recognize a specific style and distinguish it from other styles? Current trends in computational sciences formalize these qualities. What do they teach and not teach us on styles? These last questions prompt another one that has been too little considered in computational studies of style.
What is a style for an algorithm? Which are the best existing algorithms that could help us to classify images according to styles? which dimensions and features do they take into account? Could the results obtained from one style be reproduced forothers? Could computer vision and machine learning techniques help us come up with completely a new way of thinking about the history of art through styles?
Finally, does circulation affect style? For example, does a style become recognized as such only through circulation? How do cultural transfers affect styles? Do they strengthen styles, or rather dilute them? It seems to us that the study of stylistic circulations could escape the center/periphery model and its implicit hierarchies – but how it should do it, remains to be discussed in seminars...
Artl@s’s annual research seminar is organized collaboration with the SNFS project VISUAL CONTAGIONS at the university of Geneva (Switzerland) and with Purdue University (USA).
The seminar will take place online once a month, on Mondays at 2pm (GMT+1). The dates a priori considered are: October 17, November 14, 2022 and January 23, February 6, March 6, April 17, May 15, June 5, 2023.
Charles van den Heuvel, Huygens Institute for History and Culture of the Netherlands
Veruska Carretta Zamborlini, Technological Center of the Federal University of Espírito Santo
Giuditta Cirnigliaro - Digital Diagrammatic Models: Formalizing Aesthetic and Scientific Qualities of Leonardo’s Oeuvre
Eva Cetinić - Deep Learning of Style - From Stylistic Concepts to Computational Metrics, and Back?
- Canhui Liu - University of Cambridge
- Yuchen Yang - EPFL
- Linzhi Zhang - Courtauld Institute of Art
- Digital Visual Studies project Team - University of Zurich
Michela Passini - Université Paris 1 Panthéon-Sorbonne
Rémi Mermet - École normale supérieure Paris
Tristan Dot - University of Cambridge
Andres Karjus - Tallinn University