408 © 2020 ISAST ︲ Leonardo, Vol. 53, No. 4, pp. 408–414, 2020 doi: 10.1162/LEON_a_01927 Cultural Viz: An Aesthetic Approach to Cultural Analytics Everardo Reyes and Lev Manovich Everardo Reyes Associate Professor Université Paris 8 Information Sciences Department 2, rue de la Liberté 93526 Saint-Denis, France ereyes-garcia@univ-paris8.fr Lev Manovich Professor The Graduate Center City University of New York 365 Fifth Ave New York, NY 10016, U.S.A. lmanovich@gc.cuny.edu See www.mitpressjournals .org/toc/leon/53/4 for supplemental fles associated with this issue. ABSTRACT Cultural Analytics (C.A.) is an approach for analyzing media and digital culture using data methods and visual computing techniques. This article explores the aesthetic value of C.A. by approaching cultural visualizations as digital artworks. The authors present a variety of techniques developed since 2007 by members of the C.A. lab for creating visualizations of media artifacts and collections of images. Through a series of projects conducted by them, the authors discuss the artistic meaning of media visualizations and their experience in art exhibitions, workshops and seminars. Cultural Analytics (C.A.) was introduced by Lev Manovich 15 years ago as an approach to the analysis of culture using data methods and visual computing techniques [1]. C.A. involves designing exploratory methods and visualization models appropriate for diferent kinds of visual cultural data; assembling cultural data sets; applying the methods to these data sets; and describing and interpreting the results. By “cultural data,” we mean creative artifacts produced by both professionals and nonprofessional members of the public (for example, Instagram photos) and also data about cultural events and processes (such as dates and locations of music festivals, art exhibitions and design weeks). Te resulting projects combine new insights about the data and aesthetic value along with innovative organizations of 2D/3D spaces and interactive techniques. While a C.A. project can be considered as an analytical map or as a knowledge representation that helps in questioning a topic and seeing it from diferent angles, in this article we focus on the aesthetic roles of cultural visualizations, as well as their expressive and interpretative value. Trough a series of projects, exhibitions and workshops conducted by us, we discuss relationships between the subjective vision of cultural analysts—who try to express aesthetically a cultural data set—and the possibilities/limits of representation with computing technologies. Overall, we highlight the importance of presenting cultural visualizations projects in art exhibitions in order to give the audience opportunities to see diferently contemporary processes of digital culture. Cultural Analytics: Methods and Tools In 2007, when Cultural Analytics research started, the number of social media posts generated by users already counted in the millions, media software was already popular and computing technologies were already being used in social sciences and humanities. At the same time, visual analysis of cultural data using digital techniques was still in its infancy. First, the available image repositories were constituted of mainly canonical samples, thus leaving out user-generated content. Second, some visual computing techniques were accessible only in specialized software outside the media realm, for example, in scientifc applications. And third, the visualization models that were being used to map collections of images excluded the images themselves in favor of dots and lines symbolization. Since 2007, the C.A. approach has consolidated itself as the research program of a lab at the California Institute for Telecommunications and Information Technology (Calit2). Te lab’s members and collaborators combine skills in art, humanities, sciences and computer science. Together, we have developed more than 50 projects, several of them funded by NSF, NEH, Singapore Ministry of Education and University of Tyumen, among other international institutions. Since its inception, one of the steering objectives of the lab has been to elaborate visual models and to develop easy-to-use open access software and data sets for exploratory media analysis. Below we describe the principles of this software and provide examples of visualizations applied to image collections and media artifacts. All this work has been developed by a number of the lab’s members over the years.