ResearchJanuary 15, 2016 version

“... art history could profit from today’s data-centric methodologies and reveal insights that were impossible until now [...] See the pioneering works by Maximilian Schich ...” – Martin Warnke, Leuphana

Background ^

My research work adresses questions and challenges of art, architectural, and cultural history, using a multidisciplinary approach that integrates qualitative inquiry and observation, with methods of computation, natural science, and information design. The resulting research processes are characterized by international collaboration and co-authorship, while also nurturing single-authored student work, such as monographic PhDs. The research processes are expressed in a lab-style environment inspired by architectural think tanks, corporate design studios, and labs in physics or systems biology. Products aim at hi-impact journals and proceedings, catering to a broad audience, while striving to deal with images and figures in the manner of high-quality artist publications. Even though inspired by science, my approach is not without precedence in art and architectural history, standing on the shoulders of practitioners such as Geymüller, Lanciani, Barr, Malraux, Kubler, the Eames, Venturi & Scott-Brown, Doxiadis, Koolhaas, and many more. The following paragraphs lay out the roots of my approach, current and future work, the role of computation, the relation of quantification and qualitative inquiry, the function of information visualization, and a general grand challenge.

Early career ^

Expertise acquired during my early career informs current and future work in an essential way. I have spent substantial amounts of time at Zentralinstitut für Kunstgeschichte, the Glyptothek in Munich, and Bibliotheca Hertziana in Rome, working on exhibitions and database projects. For over a decade, I dealt with programmers, curators, customers, and users on both sides of the Atlantic as a consultant for large-scale projects in art research. In a function best described as “database pathologist”, I got deep insights into issues of technology and content. Technical expertise, acquired during this phase, covers data modelling, user interaction design, workflow analysis and planning. Content expertise covers large subject classifications, ancient sculpture, antique reception, artist monographs, and documentation of modern architecture. Individually, I have also done substantial research on imperial Roman baths, Roman topography, and architectural representation in visual documents. Working with, then rare, now abundant, graph databases since 1996, I was able to indentify emerging complex network structure as a property and source of many issues in art research. This crucial discovery around 2002 eventually gave direction to my PhD on Antique Reception and Visual Citation as Complex Networks [1]. Subsequent work in the area of Complex Networks in Art Research [2] was funded by the President of Max-Planck, Albert-László Barabási at Northeastern University (a key leader in network science), German Research Foundation, and Dirk Helbing at ETH Zürich (a key leader in computational social science). The domain of Arts, Humanities, and Complex Networks is now an established part of both network science and digital humanities.

Current and future work ^

My current and future work essentially deals with large numbers of observations, complex emerging structure, and evolutionary dynamics in art and culture. My research makes use of large and very large data sources, from thousands to millions of records, both proprietary and open accesss, covering visual, descriptive, and numerical aspects, collected over hundreds of years and recently assembled by crowd sourcing or automatic sensing. Coming out of the Munich school, using the trained eye of an art historian is part of my approach. This practice is augmented by data science techniques, visualization, high performance computing, and quantitative modeling in collaboration with physicists. Currently, I work with several large image datasets of paintings and tourist photos, Google Ngrams, Wikidata, Freebase, Allgemeines Künstlerlexikon Lexikon, the Getty Provenance Index and Vocabularies, as well as data on the brain connectome. Major topics of investigation include hidden structures and dynamics in multidimensional knowledge graphs, non-intuitive dynamics of ranking and canon, evolutionary processes in the continuity and variation of images, and the dynamics of cognition and attention related to artistic and architectural production and consumption. Regarding order of magnitude, I am currently working with millions of records in a typical project, which is large but not big data. The goal is to grow at least by one order of magnitude with every major project – a trajectory that holds for my work since 1996. In terms of high performance computing a recent project used 4.5 million system units on the ETH-BRUTUS compute cluster to extract regional cohesion from historic data on cultural spreading (to appear). I also do qualitative work, including virtual “excursions” in urban ecology, harnessing high-bandwidth real-time browsing capabilities in Google Earth, based on a fast network connection in the lab. Upcoming publications include a featured perspective article laying out a systematic science of art and culture in the Journal of Digital Art History (LMU-Munich), peer-reviewed journal papers on the Dynamics of the Europan Art Market and Person Networks in Wikipedia, as well as an innovative iteration of our Charting Culture video for the Lincoln Center Global Exchange in New York City.

Computation ^

Regarding the role of computation in my research, I argue in a recent perspective, laid out in response to John Brockman’s Annual Edge Question[3], that what is now called “cognitive computing” or “deep learning” may not surmount to “machines that think”, but powerful enough machines to enter a new era of exploration. Such machines allow for a systematic cartography of thoughts and observations, both implicit in large amounts of data, and potential within the constraints of what we know. This emerging situation can be harnessed to facilitate the established practice of art and architectureal history, similar to innovations in the animation industry, where computation is increasingly driven by mathematicians and hidden in cloud processing to allow for artists to concentrate on the act of drawing, in a very classical sense. In this vision, the human researcher is not threatened to be replaced by algorithms, but takes the active role of the violinist who is enhanced by the mastery of her instrument.

Natural science methods ^

Nurturing natural science methods within art and architectural history promises to overcome the long-standing separation of “nomothetic” law disciplines such as physics, and “ideographic” event disciplines such as history, as postulated by Windelband in 1894 [4] and famously lamented by C.P. Snow in 1959 [5]. Weaver 1948 [6] and implicitly Jacobs 1961 [7] have already argued that such an integration is possible and indeed necessary to address abundant problems of “organized complexity”. A Network Framework of Cultural History, published in Science Magazine [8], provides a mutual justification for quantitative and qualitative research and lays out several aspects of mutual benefit. In terms of justification, the article shows that quantification in the humanities does indeed work by bringing evidence for the physical “laws of migration” spanning over 800 years, based on simple birth and death records of large numbers of artists and other individuals; the article also shows that quantification cannot replace qualitative inquiry, as the system of cultural history is characterized by massive fluctuations on a local level. In sum, the article provides evidence that science can indeed contribute towards a quantitative understanding of cultural processes (in the sense of Feynman); it also shows that sense-making cultural meta-narratives emerge from large amounts of granular information; and despite local fluctuations quantitative maps can help to cross-fertilize qualitative domain expertise within the context of a big picture. The rigorous quantfication of bias on mesoscopic and global levels adds to the general usefulness of this combined approch.

Information design and visualization ^

Information design and visualization are a subject of study and practice in my work. Figures are at the base of multidisciplinary research, functioning as a lingua franca of scholarship that allows for efficient exploration of data and communication with practitioners from highly diverse backgrounds. Figures enable us to overcome difficulties of mutual illiteracy, regarding difficult visual comparison, results of differential equations, statistical tables, or differences in terminology. Art and architectural history are uniquely positioned to further integrate and advance such visual practice, as the disciplines practically own visual literacy and knowledge about the evolution of visual representation. Beyond using advanced figure conventions for exploration, choosing the right visual paradigm carries the potential to reach very broad audiences, as in case of the very-easy-to-understand “airline map” of cultural migration in our recent Nature video, which got over one million views on Youtube so far.

Grand challenge

Human population and their artifacts are growing faster than exponential, while cultural interaction approaches the fidelity of a global nervous system. Every day more than 350 million images are loaded into the social network of Facebook alone. As this myriad of new artifacts threatens to veil our view into the past, like city lights covering the night sky, it is easy to forget that there is more than one Starry Night, the painting by Van Gogh. Like in ecology, where saving rare species may help us in treating disease, art and architectural history can reveal insights into the past, which may hold keys to our own future. In a world of humanism under threat, facing the challenge of understanding the nature of culture is more crucial now than it ever was.

References ^

[1] M. Schich: Rezeption und Tradierung als Komplexes Netzwerk. (Munich: Verlag Biering & Brinkmann, 2009).

[2] M. Schich: Complex Networks in Art Research – Exemplary Proofs of Concept. (Project report, Bonn: Deutsche Forschungsgemeinschaft, 2012). URL:

[3] M. Schich: Machines mostly steal thoughts but open a new era of exploration. in: J. Brockmann (ed.): The Edge Question 2015 – What do you think about machines that think? (Online/print to appear, New York: Edge Foundation, 2015). URL:

[4] W. Windelband: Geschichte und Naturwissenschaft. (Straßburg: Heitz, 1904 [1894]).

[5] C.P. Snow: The two cultures. (London: Cambridge University Press, 1959).

[6] W. Weaver: Science and Complexity. American Scientist 36 (1948) 536-544 URL:

[7] J. Jacobs: The Death and Life of Great American Cities (New York: Random House, 1961)

[8] M. Schich, C. Song, Y.-Y. Ahn, A. Mirsky, M. Martino, A.-L. Barabási, D. Helbing: A Network Framework of Cultural History.
Science 345,6196 (2014) 558-562. URL: