Ecole normale supérieure,
Paris, 45 rue d’Ulm
Thursday, June 13
Tea and Coffee – Amphithéâtre Rataud
10:00 – 10:15
10:15 – 11:00
Béatrice Joyeux-Prunel, École normale supérieure, Paris – Université Paris Sciences Lettres
ImageContagion is a new sub-project of the Artl@s research group (www.artlas.ens.fr), which focuses on the global circulation of images from the 1890s to today. Nobody knows how images have fueled historical upheavals and modernization in the 20th c., which images have fostered eventual global convergences of taste, and according to which geography. Visual circulations pose also challenges to the cognitive sciences and to visual studies. Our objectives are to identify the most recurrent images of the past century, their cognitive features, the characteristics of their authors, and to study their impact in copies, imitations, visual quotations, their channels of circulation and their contribution to a global cultural homogenization, and to question the old geopolitical model of prescriptive centres and imitative peripheries. We will thus contribute to an epidemiology of images –an innovative model for the analysis of iconographic and stylistic diffusions that can be applied across periods, languages, and disciplines. To do so, we start from the unprecedented availability of massive digitized corpora, whose images can be extracted, described and studied with deep learning applied on pattern recognition, text mining, historical geomapping and the methodologies of global history, art history, visual studies, and cognitive sciences. These digitized images will be crossed with press reviews and archival material, to explain how artistic and media images circulated and how they have helped to disseminate styles and visual patterns, as well as new ideas about politics, the body, and nature. The Project ImageContagion is made possible thanks to the collaboration with the EnHerit project led by Mathieu Aubry at École des Ponts ParisTech (http://enherit.enpc.fr/).
Béatrice Joyeux-Prunel is Maître de conférences HDR at the Ecole normale supérieure, Université Paris Sciences Lettres (PSL). At ENS she is in charge of the teaching of modern and contemporary art history, as well as in the curriculum in digital humanities. She has published on the transnational history of modern and avant-garde art (Éditions Gallimard, 2016 and 2017, and forthcoming). She leads the Artl@s Project which she founded it in 2009 (www.artlas.huma-num.ens.fr). She also leads the Postdigital research group on contemporary art and digital cultures (www.postdigital.ens.fr).
11:00 – 11:15
11:15 – 11:45
Active Diagrams: Reconsidering Visual Representations of Networks
in Avant-Garde Magazines of the 1920s
Gábor Dobó and Merse Pál Szeredi, Petőfi Literary Museum–Kassák Museum
Our presentation analyses ‘network diagrams’ illustrating the interrelationships between avant-garde magazines of the 1920s, based on recent research conducted for an exhibition of the Petőfi Literary Museum–Kassák Museum. Our primary approach is historical instead of digital, yet, we are convinced that a critical re-reading of the function, structure and political as well as financial embeddedness of such networks can contribute to the theoretical considerations related to (digital) network analysis. Focusing on early visual representations of avant-garde networks, we analyse the politics of networking and different strategies employed by editors of magazines.
During the early 1920s, communication between avant-garde groups was elevated to a supranational level. They were involved in a process of intense exchange, debate and competition, and viewed each other as important points of reference. The ‘network diagram,’ represents this dynamic, ever-changing relationship, and was given a different meaning and function in each magazine. ‘Network diagrams’ may be based on starkly different concepts: they can represent both the hierarchy and the equality between magazines, as well as be consistent with or go against the program of the given magazine.
A reconsideration of ‘network diagrams’ published during the 1920s also supplements the discourse about what kind of networks do we talk about either in a historical perspective or in the field of (digital) network analysis. In our presentation, we demonstrate the potential of historical investigation through our interdisciplinary research project and related exhibition, focusing on the international network of Lajos Kassák’s magazine Ma (Today) between 1920 and 1925.
Gábor Dobó is PhD literary historian. He studied at universities in Budapest, Florence and Angers. He has been a researcher at the Petőfi Literary Museum–Kassák Museum since 2014. He is researching periodicals of the avant-garde. He is a member of the European Network for Avant-Garde and Modernism Studies (EAM) and committee member of the European Society for Periodical Research (ESPRit).
Merse Pál Szeredi is MA art historian working at the Petőfi Literary Museum–Kassák Museum, and PhD candidate at the Eötvös Loránd University. His research focuses on interwar avant-gardes in East-Central Europe, especially Lajos Kassák’s Vienna exile and his magazine Ma during the 1920s.
11:45 – 12: 15
Regional Cohesion and the Centrality of the Arts
Maximilian Schich, UT Dallas
In previous work we have quantified the cultural migration of 153.000 artists and 120.000 noted individuals, spanning a time-frame of more than 2000 years (Schich et al. Science 2014). Since then, others have picked up on the paradigm, including aspects of gender, and datasets with more individuals on a global scale, and more detail in specific regions. Here, we extend over our original work, looking at the attraction of cultural centers differentiated by genre of occupation, including governance, academic, and the arts. We observe both commonalities and striking differences between these genres, as well as between global regions and countries. These differences, it turns out, are not rooted in East versus West, but in the more centralized or federal nature of the respective region. Indeed, more federal areas in Europe appear to resonate well with China and India, while more centralized regimes can be found all over the world. In a second aspect of analysis, we used high-performance computation to visualize the effective regional overlap in Europe over several centuries.
This analysis strikingly recapitulates the history of the sub-continent in a single static picture, while putting into question the ongoing rediscretization of the sub-continent. Together, both aspects indicate a differentiated situation, where both centralization and broader regional exchange are evident, defying centralist and anti-central theories at the same time.
Maximilian Schich is an associate professor for arts and technology at the University of Texas at Dallas and a founding member of the Edith O’Donnell Institute of Art History. His work converges hermeneutics, information visualization, computer science, and physics to understand art, history, and culture. He is the first author of A Network Framework of Cultural History (Science Magazine, 2014) and a lead co-author of the animation Charting Culture (Nature video, 2014). He is an editorial advisor at Leonardo Journal, an editorial board member at Palgrave Communications (NPG), and the Journal for Digital Art History.
12:30 – 13:30
Lunch for participants: Restaurant of the ENS
Coffee in Conference Room
14:00 – 14:30
Patterns of Transregional and Transnational Circulations
in American Women Artists’ Professional Networks
Catherine Dossin, Purdue University
The Women Artists Shows.Salons.Societies project was launched in 2017 as a collaboration between Artl@s and the Archives of Women Artists, Research and Exhibitions. Combining AWARE’s ambitions to restore the presence of 20th-century women artists in the history of art and Artl@s’s desire to provide scholars with the data and tools necessary to question the canonical art historical narratives through quantitative and cartographic analyses, we decided to focus on group exhibitions of women artists and collectively write a history of all-women exhibitions from the 1860s to the 1960s.
As part of the project, I worked on women artists in the United States and identified more than a thousand exhibitions. While at first the figure may seem rather high, it assuredly does not account for all the women group exhibitions prior to the 1970s. Although the research is greatly enabled by the historical sources available online, such as library repositories, historical newspapers and women’s magazines, the value-system used to establish these sources is indeed inherently limiting it.
After entering these exhibitions in the Artl@s database, it became possible to map their development over time and uncover geographic and chronologic patterns that a close reading of archival material helps explain or at least interpret. It allows us to reconstruct the networks that connected American women not only at a regional, but also national and even international level. Replacing these networks in the larger context of the Women’s Club movement and the Women’s International Movement further highlights the transregional and transnational circulations of women artists and the connectedness of the early woman’s artworlds.
Catherine Dossin is Associate Professor of Art History at Purdue University and serves as Editor of the Artl@s Bulletin. She is the author of The Rise and Fall of American Art, 1940s-1980s: A Geopolitics of Western Art Worlds (Routledge, 2015), the co-editor with Thomas DaCosta Kaufmann and Béatrice Joyeux-Prunel of Circulations in the Global History of Art (Routledge, 2015), and the editor of France and the Visual Arts since 1945: Remapping European Postwar and Contemporary Art (Bloomsbury, 2018).
14:30 – 15:00
Paths of (French) Glory. The Visual and Physical Circulation
of Matsukata’s Confiscated Collection (1944-1959)
Léa Saint-Raymond, Collège de France),
and Maxime Georges Métraux, Sorbonne University / Galerie Hubert Duchemin
In December 1944, the French State Property Authority decided to sequestrate Kojiro Matsukata’s collection of paintings, drawings and sculptures, considering them as enemy goods. In the interwar period, this Japanese industrialist, who knew Claude Monet and Léonce Bénédite very well, had gathered together hundreds of works by the impressionists, Gauguin, Cézanne, Van Gogh, Rodin, Courbet, and by many contemporary artists. When the French State confiscated them, in 1944, the issue of its becoming arose very quickly. Bernard Dorival, curator of the Musée national d’art moderne, considered this collection as “one of the most important that exists not only in France, but in the whole universe” and, thus, as a “unique opportunity, the only way to fill irreparable gaps” of the French heritage. While the museum directors wanted to keep all or part of the collection, the Ministry of Foreign Affairs and the AFAA – Association Française d’Action Artistique – also considered that the circulation of the works towards foreign exhibitions would ensure an even greater influence in the post-war geopolitical game.
Built on a digital database of exhibition catalogues and reproductions of Matsukata’s confiscated collection, this paper aims at identifying the political and artistic networks of its circulation, and seeks to understand the opposing reasons why some works travelled all around the world, some finally stayed in the French museums – like Van Gogh’s Bedroom in Arles, curated at the Musée d’Orsay – and the bulk of them were “given” back to Japan in 1959.
Léa Saint-Raymond defended her PhD in art history in 2018, focusing on the Parisian auction market from the 1830s through 1939. Member of Artlas’s core team, she has been contributing to the global database of exhibition catalogues and she co-funded GeoMAP, Geography of the Parisian Art Market 1815-1955.
Maxime Georges Métraux is a PhD candidate in art history at Sorbonne University. He is also an expert for the Galerie Hubert Duchemin and a founding member of GRHAM (Groupe de Recherche en Histoire de l’Art Moderne). He teaches at the Paris-Est-Marne-la-Vallée University.
15:00 – 15:15
15:15 – 15:45
Raphael All Over: Mapping and Qualifying Originals and Copies
Marco Jalla, Université de Genève
In order to realize the first catalogue raisonné of the works of Raphael, Johann David Passavant had to deal with paintings scattered all over Europe. He spent about ten years of his life locating the old master’s works, identifying them and distinguishing the originals from their many copies. On the basis of this tremendously detailed documentation, I would like to answer a simple question: Where could Raphael’s artworks be seen around 1850? Originals and copies do not reside in a particular place by pure chance: local, regional, national, and/or global factors determine their distribution. In addition, the copies offer a peculiar ubiquity to the original works painted by Raphael’s own hand. I will focus on these two points for analyzing the dissemination of Raphael’s paintings within and outside Italy. We will show how, for a long time, the greed of the European monarchies for Raphael’s painting has effectively dislocated the old master’s oeuvre. However, our results will also demonstrate a new driving force at work in the 19th century, which reassembled and definitely fixed the paintings in new places called “museums”. These new compendia offered for the first time comprehensive images of Raphael’s artistic styles and evolution, which were also going to challenge the perception each nation had of his work.
Marco Jalla is PhD candidate in Art History at the University of Geneva. He is writing a dissertation under the supervision of Professor Dario Gamboni about the functions of copy painting in the 19th century. Previously, he studied at the University of Neuchâtel, at the Ecole Normale Supérieure in Paris, and the Ecole des Hautes Etudes en Sciences Sociales in Paris, and at the Technische Universität in Berlin. He also was a fellow at the Global and European Studies Institute in Leipzig and at the Zentralinstitut für Kunstgeschichte in Munich.
15:45 – 16:15
New York, Latin America, Barcelona. On Geolocating
a Portapak as Means to Historicize Relational Networks
Pablo Santa Olalla, Universitat de Barcelona, MoDe(s)
In the early seventies, a Portapak (first portable video tape recording system) was acquired in New York by Muntadas. It traveled with him through Argentina, Brazil, Peru, Bolivia, Caracas and Mexico. The camcorder landed in Barcelona, where it was used to organize a workshop that started Video-Nou, a project of alternative television. This Portapack took part in various networks, and helped to create other in distant geographies. Does geolocating its movements just illustrate what art history already knows? Or is there a knowledge supplement that can be achieved only through map visualizations? This paper addresses the historiographical problem of transnational approaches, where art historians have to confront themselves with the swamp terrains of interdisciplinarity and the center/perifery dichotomy. But what does happen when you place the coordinate axis not in the ground but in the moving things?
Pablo Santa Olalla is an APIF-UB (Grant for Research Personnel in Training) scholarship holder at the Department of History of Art of the University of Barcelona, where he teaches “Artistic Languages”. He is currently working on the relational networks between Latin American and Spanish artistic conceptualisms within the framework of the research group Art Globalization Interculturality (AGI) and the research project MoDe(s).
16:30 – 17:00
Images of Political Leaders in Circulation in Africa.
A Digital Cognitive Approach
Sophie Bodénès-Cohen, CogMaster, ENS Paris
The reappropriation of the propagandist representation of revolutionary leaders of the 1980s in Sub-Saharan Africa, in the revolutionary and popular art of the 2000- 2018 period, constitutes a micro-process of cultural evolution that can be modelled and quantified thanks to artificial intelligence. This phenomenon tends to converge the artistic portraits of these idealized leaders into popular icons with recurring facial features and characteristics. The traits that recur are cognitive attractors that have spread virally on social networks since the 2000s, constituting a phenomenon of “cultural epidemiology” (Dan Sperber). Through the development of computer vision and deep learning algorithms, it is possible to visualize the emergence of recurrent patterns and to understand their attraction factors. From this hypothesis, we applied computer vision algorithms on a database of more than 1000 portraits of Thomas Sankara, Haile Selassie and Nelson Mandela. Thus, we will be able to understand how cognitive, visual biases and environmental attraction factors (economical and historical) shape the evolution of the representation of these portraits. In this presentation, we will first present a pattern recognition algorithm that enables to gather symbolic attributes of leaders’images that resemble the most. A study on color features will reveal the visual biases that shape the evolution of revival portraits into attractors with color histogram and color segmentation. Average face (mixing the pixels of all images) and Eigenfaces (the selective extraction of the main visual components of the images) with Principal Component Analysis machine learning algorithm will show that the portraits are stabilized according to more frontal visual components, edges accentuated on specific facial features (especially smile and eyes) and brighter colours. Then, once georeferenced and visualized through chronologically, the clusters of portraits can help trace the diffusion of specific political visual semantics, and highlight cultural transmission phenomenon.
Sophie Bodénès-Cohen is a Master student in cognitive science (Cogmaster-ENS) and an Artlas intern. She is interested in applying computer vision and deep learning to model artistic and cultural evolution phenomena.
A Formal Ontology for the Description and Contextualization
of Iconographical Representations
Nicola Carboni, University of Zurich
The perception of our visual heritage is based on sign-functions which relate representations to types and results in perceptual judgements over physical objects. The recording of these type of assertions and their differentiation in relation to the level of interpretation is paramount for the comprehension and analysis of our heritage. The contribution proposes a functional theory for the organisation of the units of information related to the visual works using the identity of their single constituent elements as identifying criteria. The framework developed is, then, used for the formalisation of an ontology which is constructed as an extension of CIDOC-CRM and present the possibility to record statements about the diverse visual elements present in a representation. The result is tested with artworks coming from the Byzantine and Renaissance tradition, showing how we can assign meaning to iconographical objects, and distinguish or aggregating similar compositions. The use of the ontology for the annotation of visual representation would help to relate each element to its referenced entity, as well as help navigate the diversity in the symbolic assignment. Moreover, thanks to already established classes and properties from CIDOC-CRM we can also ground the information about a representation in time and space contextualising the meaning-assignment to a specific cultural period.
Nicola Carboni is a Research Fellow for the University of Zurich. Previously appointed Marie Curie Fellow and Digital Humanities Fellow at Harvard University – Villa I Tatti, his interest lies in the conceptualisation and formal description of visual content from the tangible/intangible domain.
Side by Side: Al Freeman’s Art History
Taylor Walsh, Harvard University/Museum of Modern Art, New York
This paper is the first in-depth case study of the Comparisons series (2017/18) by the Canadian artist Al Freeman (b. 1981). The work consists of 51 cut-paper collages divided down the middle, with a reproduction of a painting or sculpture on one side and a found digital photograph on the other. Masterpieces of twentieth-century art are brought low by Freeman’s unexpected pairings, which often hinge on certain felicities of shape and color: Matisse’s Bonheur de Vivre sits alongside a Bento box of rainbow sushi, and the flesh tones of a Josef Albers Square are picked up in the interlocked arms of grappling wrestlers. This is search-engine art, born of and reliant on the internet’s glut of images. But Freeman’s dyads also have a mock-pedagogical feel, recalling the slide comparisons used to teach the history of art.
Dating back nearly a century to Heinrich Wölfflin’s masterful deployment of twin slide projectors to illustrate his lectures, this method of compare-and-contrast was born of populist intentions, placing faith in the untrained eye to make distinctions and judge superior quality without the need for historical knowledge. By culling her comparative material from online message boards and niche social media sites—the most democratic of forums—Freeman turns the tables on that time-honored mode of visual analysis, asking how it might fare under pressure from the search algorithms and recognition software that have revolutionized how we find and use images. Ultimately, Freeman’s Comparisons urge us to look askance—both at canonical works of art and the ways in which we learned to see them.
Taylor Walsh is a PhD candidate at Harvard University and a Curatorial Assistant at the Museum of Modern Art, New York, where she co-organized the retrospective Bruce Nauman: Disappearing Acts (2018-19) co-edited the accompanying catalog. Her academic research has been supported by fellowships from the Terra Foundation for American Art, the Smithsonian American Art Museum, the Getty Research Institute, and the Whitney Independent Study Program, and her writing has been published in October and The Burlington Magazine, as well as in exhibition catalogs for MoMA, the Harvard Art Museums and the ICA Boston.
Dinner for participants
Friday, June 14
Tea and Coffee – Amphithéâtre Rataud
10:00 – 11:00 Keynote Address
Time Machine Europe
Frédéric Kaplan, École polytechnique fédérale de Lausanne
Replica: Looking for Patterns in Big Digital Artwork Databases
Isabella di Lenardo & Benoît Seguin, École polytechnique fédérale de Lausanne
The digitization of millions of artworks by museums, libraries and cultural institution brings unprecedented access and opportunity to engage with collections. It also poses new challenges in the search, retrieval, and analysis of visual material. Currently, online databases of art can only be searched through keywords, or tags, namely the textual metadata of the original object. Textual queries, however, are not designed to search visual information which has not been indexed – they cannot for instance find similar shapes, forms, or motifs. This is why a new tool for searching art through its visual attributes is needed.
Started in 2015 in partnership with the Giorgio Cini Foundation in Venice and Factum Arte in Madrid, the Replica project aimed at leveraging modern Computer Vision to tackle this challenge. By combining novel scanning hardware and recent developments in Deep Learning, it was possible to analyze and index a significant portion of the fototeca of the Cini Foundation. Being fed on this large amound of data, a visual search engine, tailored for the search and exploration of artistic collections, was developed. It is opening new avenues for exploring and engaging with the data by assisting art historians in the study of iconography, the transmission of forms, styles and patterns, while engaging larger audiences in a process of learning and discovery.
Prof. Frédéric Kaplan holds the Digital Humanities Chair at the École Polytechnique Fédérale de Lausanne and heads the Digital Humanities Lab (DHLAB). In this context, he leads projects combining archival digitization, modelling and museographic design. He is currently working on the “Venice Time Machine”, an international project in collaboration with Ca’Foscari University to model the evolution and history of Venice over a period of 1000 years.
Isabella di Lenardo is Ph.D in Theories and Art History, Post Doc researcher in the DHLAB at the EPFL. Her research interests are focused on the production and circulation of artistic and architectural knowledge in Europe XVIth – XVIIIth centuries with a particular stress on North-South relationships and influences. She is Project Head of the Urban Reconstruction for the Venice Time Machine project, held by Frédéric Kaplan, and Project Head of “Replica” project, the digitisation of 1 million of photos of works or Art in the Fondazione Cini (Venice) and their Patterns extraction through a search engine for visual similarities.
Benoît Seguin is an independent Engineer specialized in the applications of Machine Learning, especially to the domain of Cultural Heritage. He received a Diplôme d’Ingénieur from École Polytechnique (Paris), a Master of Science and a PhD from Ecole polytechnique de Lausanne (EPFL). His thesis, at the Digital Humanities Laboratory Lab at EPFL, was based on the Replica Project where he used machine learning and modern image processing to help Art Historians navigate very large iconographic collections.
11:15 – 11:45
Tracking Venus Through the Ages. Deep Learning and the Study of Long-Term Iconological Circulations
Mathieu Aubry, LIGM lab, École des Ponts ParisTech, K. Bender, Independent Researcher, Oumayma Bounou, École nationale des Chartes, and Béatrice Joyeux-Prunel, ENS, and Xi Shen, École des Ponts ParisTech
This paper questions the modalities of the temporal and spatial dissemination of an iconographic theme – Venus – through reproductions, engravings and illustrations, making the most of a computational analysis of the sources. The study is based on a digital collection of more than 20,000 works of art representing Venus (sculptures, drawings, paintings, frescoes, engravings and illustrations) distributed from the Middle Ages to Modern Times, assembled by K. Bender since 2004 and preserved in the repository heiDATA. The names of artists are identified, the dates of the works are specified, and the places of production, conservation and creation have often been identified. What are the most reproduced works of art in history? Where were they reproduced and where and how did specific patterns circulate over time and space? This paper presents the results of a statistical, geographical and algorithmic approach applied to image descriptors (distant reading), as well as to the images themselves (distant viewing) – with the application of the deep learning algorithms of the EnHerit project led by Mathieu Aubry (http://enherit.enpc.fr). Identifying which motifs have circulated the most – copied, imitated, caricatured – and analyzing the geographical trajectory of specific patterns, we will propose results on the European diffusion of iconographic motifs and the fashion of certain styles, as well as new questions for art history.
Mathieu Aubry is a Computer Vision researcher at the LIGM lab at École des Ponts ParisTech.
- Bender is an independent scholar working since 2004on the world diffusion and evolution of pictorial representation of Venus, with statistic analysis.
Oumayma Bounou is an engineer at Ecole nationale des Chartes, working on the Filigranes pour Tous project with the algorithms developed in the EnHerit project.
- Joyeux-Prunel (ENS – PSL) and Mathieu Aubry have started working together in 2017, to apply deep learning algorithms to the artistic and visual databases in order to trace the global diffusion of images, patterns, and styles over centuries.
Xi Shen is a PhD candidate in Computer Vision at École des Ponts ParisTech.
11:45 – 12:15
Reddit “Place”. A Warburgian Case Study in Image Circulation
and Affective Afterlife in Collective Online Art
(Philipp Wüschner, Freie Universität Berlin)
On April 1st 2017, the online network reddit offered its users a blank space in form of a pixelated grid, called ‘Place’, with a simple set of rules: 1. You can fill any pixel in any of 16 colors. 2. You can paint over any already colored pixel. 3. You can fill only one pixel every 5 minutes. The result was a collectively created digital tapestry of globally circulating art-works, pop-cultural memes, and online imagery, creating a snap shot of our current cultural memory. Today, certain techniques of animated data-visualization and re-mediatisation offer a unique look onto the artistic and affective dynamics behind this snap shot. Building on these techniques, I will try to give a brief iconological analysis of the ‘art-work’ as a whole, touching the pop-cultural history of some of the images at hand, focusing especially on displays of (national) identity, its re-appropriation, and humorous subversion. My analysis will highlight the migrating nature of online images in the sense of Warburg’s ‘image vehicles’, as well as their affective afterlife not only on the internet but also materialized in the real-life-word. I will claim that by fusing the acts of production and distribution, the circulation of images itself and the (affective) afterlife it creates become of main aesthetic value. My reading will attempt a critical cartography of these circulating images with particular regard to ways of aesthetic
Friday, June 14
identity forming: Can the iconological analysis of the image with all its national flags, brands, and pop-cultural symbolism validate the alleged utopian nature of online networks?
Philipp Wüschner is a philosopher, associated at the SFB Affective Societies at Free University Berlin. There, he received his PhD 2014 with a thesis on Aristotle’s concept of habit. His research interests are affect-theory (with a focus on art and digital media), aesthetics, and habits as methods in the humanities.
12:30 – 13:30
Lunch for participants: Restaurant of the ENS
Coffee in Conference Room
14:00 – 14:30
Beyond Text. Retracing Artistic Shapes Through Computer Vision
(Tino Mager, Delft University of Technology)
Major digitisation projects currently provide access to a wide range of digitised cultural assets and improve the visibility and social relevance of these objects. However, the entire search and linking in the data continues to be based on metatags and annotations attached to the visual material. Distortions based on incorrect and inadequate textual descriptions will continue to skew our understanding of the large-scale context of the objects concerned. Visual phenomena, which elude our ability to verbalize, evade a comprehensive and comparative access, as well as the novel potentials of the technology of the digital century.
ArchiMediaL, an interdisciplinary research project in cooperation between Delft University of Technology and Vrije Universiteit Amsterdam, deals with methods of artificial intelligence for the recognition of visual characteristics of image content. This is not only about the recognition of visually similar representations, but rather about the recognition of the represented – in our case architectural – object. The pursuit of forms on the basis of object characteristics avoids linguistic inadequacies and helps to explore completely new contexts of form and meaning between art objects, as well as the global spread of specific formal phenomena. In our research project, art historians, computer scientists and architects work on the development of deep learning algorithms for the recognition of image content and the associated methodological challenges. My presentation will introduce you to our research and give you an outlook on the possibilities that will arise in the near future.
Tino Mager studied media technology and art history in Berlin, Leipzig, Barcelona and Tokyo. Currently postdoc at the Faculty of Architecture of the TU Delft, he is working on the development of methods for the use of artificial intelligence in architectural history research.
14:30 – 15:00
Visual Style in Two Network Era Sitcoms
(Taylor Arnold and Lauren Tilton, University of Richmond)
In this presentation, we show how face detection and recognition algorithms, applied to frames extracted from a corpus of moving images, are able to capture many formal elements present in moving images. Locating and identifying faces makes it possible to algorithmically extract time-coded labels that directly correspond to concepts and taxonomies established within film theory. Knowing the size of detected faces, for example, provides a direct link to the concept of shot framing. The blocking of a scene can similarly be deduced knowing the relative positions of identified characters within a specific cut. Once produced on a large scale, extracted formal elements can be aggregated to explore visual style across a collection of materials. It is then possible to understand how visual style is used within the internal construction of narrative and as a way to engage broadly with external cultural forces. The talk will focus on the application of our techniques for extracting formal elements to a corpus of moving images containing every broadcast episode of two Network Era sitcoms: Bewitched (1964-1972) and I Dream of Jeannie (1965-1970).
Taylor Arnold is an assistant professor at the University of Richmond (Virginia, U.S.A.). Arnold studies massive cultural datasets in order to address new and existing research questions in the humanities and social sciences. He specializes in the application of statistical computing to large text and image corpora. Arnold is a 2019-2020 fellow of the Collegium – Lyon Institute for Advanced Study.
Lauren Tilton is an Assistant Professor of Digital Humanities in the Department of Rhetoric & Communication Studies and Research Fellow in the Digital Scholarship Lab (DSL) at the University of Richmond (Virginia, U.S.A.). She also serves on the Association for Computing in the Humanities (ACH) Executive Council.
15:00 – 15:15
15:15 – 15:45
EUROPEANA: Looking at a screen on cultural heritage, art, and remembrance
(Idalina Conde, ISCTE-IUL University Institute of Lisbon)
Operating since 2008, Europeana is the EU digital platform for cultural heritage as umbrella that includes items of art, cultural history and remembrance. Being the largest aggregator (of aggregators of) of contents in Europe, the platform gathers over 58,000,00 items (artworks, artefacts, books, films, music) from about 3,700 institutions: museums, galleries, libraries, archives. So, Europeana constitues one more case for global availability and circulation of images/forms. The purpose of this paper it to look at such a screen on Europe, however with an unbalanced portrait to explain by complementary perspectives, and to see through the lens of visual semantics as it is concerned in this conference. Almost unavoidable in digital plaftforms with wide scope, that portrait depends on the content providers (their profile, span, and engagement across European countries) plus the own visual curatorship to organise series in digital collections, galleries and exhibitions. Nonetheless, in spite of temporal, spatial, or thematic gaps under the visual opulence, alternative choices to display images open to the exploration of aspects or artifacts less known. To illustrate the interplay of gaps and serendipity, recognition and discovery, the paper shows how the category “art” appears in several files in Europeana (visual arts, mixed types, and artists included), parallel to a focus on Faces of Europe. A collection online with 83 artworks from the 14th century to the 1970s, itself example of hypertrophy and absence of references that reveals artists and artworks also by geographies beyond borders of the main history of art in Europe.
Idalina Conde, PhD, is lecturer and researcher in ISCTE-IUL University Institute of Lisbon, and member of the Europeana Network Association. Author of numerous research, presentations and publications on art, culture, memory and visual studies, parallel to a strand on contemporary Europe from which Setting the Soul. Europe and Culture in Dialogue with Images is one among other forthcoming books.
15:45 – 16:15
Warburg’s Mnemosyne Atlas as Art Historical Method
(Allison Leigh, University of Louisiana, Lafayette)
Artworks hold a special place in the historical order because of their unique ability to exist both in and out of the time of their making. In recent years, museums have begun to capitalize on this facet of art’s distinctive relationship to time by displaying works of art from disparate periods side by side, in contrast to traditional exhibition practices centered on firm cultural and chronological boundaries. Building on Aby Warburg’s profound belief that images of great power consistently reappear throughout history, this talk investigates the growing trend in art history to push against historical continuities and create what Warburg called “thought space” [Denkraum]. Drawing together recent exhibitions, plans for dramatic reinstallations of permanent collections, and cutting-edge digital modes for viewing art objects, this talk will explore how these projects are changing the discipline of art history. All seek the common goal of reanimating objects from the past, but those emerging in museums might be seen as a new methodological approach – one which produces union even among disparate kinds of knowledge and materials. By examining contemporary artworks in relation to those from distant earlier periods, museums and cutting-edge digital platforms provide a means of exploring the theoretical arena that lies beyond the bounds of linear historical narratives.
Releasing art from its history in this way continues the work of Warburg but draws his original project into the twenty-first century by envisioning how digital and computational mapping might allow for radical new understandings of both artworks and their beholders.
Allison Leigh is an Assistant Professor of Art History at the University of Louisiana at Lafayette. She is a specialist in European and Russian art of the eighteenth and nineteenth centuries. Her first book, Masculinity and Modernity in 19th-Century Russian Painting, will be published by Bloomsbury in 2020. She works on topics ranging from the cultural bilingualism of elite Russian portrait sitters in the eighteenth century to the prescriptions for masculine conduct among military officers in the 1840s.
16:15 – 16:45
Dangerous Dematerialization: Countering Techno-Utopianism
with Material Specificity
(Alexander M. Strecker, Duke University, Durham, NC)
Our understanding of photography has been bedeviled since its digital transition in the early 1990s. As the medium celebrated its 150th birthday, anxiety around the dematerialization of the image abounded and the ubiquity of digital technologies quickly obscured a more nuanced understanding of what had truly changed about the photograph. In the 2000s, the meteoric rise of internet-connected mobile cameras recapitulated this dynamic in more condensed form. Today, the process is playing out, yet again, with the adoption of machine learning-fueled computer vision algorithms.
Drawing on the infrastructural analyses carried out by media scholars such as Lisa Parks, Jane Bennett, and Lisa Gitelman, my paper challenges this tendency. I focus on the underlying material operations of contemporary image interpretation systems in order to better understand the creation and circulation of today’s photographs. In particular, by opening up the “black box” of machine learning, I hope to show how these systems “see,” allowing us to understand their biases and blindspots when they are tasked with interpreting the world.
Ultimately, my approach reflects my belief that “digital” does not (and must not) mean “disembodied.” Many tenets of the 1990s’ “digital revolution” have been uncritically accepted and repeated in the rhetoric surrounding the networked image and computer vision. These collected revolutions peddle a fantasy of a dematerialized digitality that renders every image into a bit, byte, and indistinguishable part of an infinite flow of data. Digital objects are different than their analog counterparts, yet they maintain a physical trace in the world—one that is being dangerously overlooked.
Alexander M. Strecker is pursuing a PhD in Art, Art History and Visual Studies at Duke University, focusing on contemporary photography and the impact of technological changes on the medium’s functioning in the past, present, and future. He received a BA in English Literature from Amherst College, graduating magna cum laude in 2013. His writing has appeared at Art Basel, Paris Photo, and Les Rencontres d’Arles as well as in The Architectural Review, The Sunday Times, LensCulture, Creative Insights and Aldebaran.
Moderation: Paula Barreiro-Lopez (Université Grenoble-Alpes, Laboratoire LARHRA)
18:00 – 19:30
Artl@s’ 10th Anniversary Celebration
This Conference is organized by Artl@s (www.artlas.ens.fr),
with the support of the Laboratoire d’excellence TransferS (www.transfers.ens.fr)
(programme Investissements d’avenir ANR-10-IDEX-0001-02 PSL and ANR-10-LABX-0099)
the Département d’histoire et Théorie des Arts, ENS (www.dhta.ens.fr)
the Institut d’Histoire moderne et contemporaine (ENS/CNRS: www.ihmc.ens.fr),
and the Laboratoire de Recherche Historique Rhône-Alpes (LARHRA, http://larhra.ish-lyon.cnrs.fr/),
in collaboration with the project MoDe(s)
– Modernidad(es) Descentralizada(s):
arte, política y contracultura en el eje transatlántico durante la Guerra Fría
(Universitat de Barcelona, ref. HAR2017-82755-P).