Artistic Data and Network Analysis Figure 1. Aby Warburg’s Panel 45 with the color version of the images mapped over the black-and-white photographic reproductions. Peer-Reviewed Images as Data: Cultural Analytics and Aby Warburg’s Mnemosyne

Stefka Hristova

Abstract: In this paper, by extending the methodology of media archaeology to the praxis of Cultural Analytics/Media Visualization I ask how have we compared multitude of diverse images and what can we learn about the narratives that these comparisons allow? I turn to the work of Aby Warburg who attempted to organize close to two thousand images in his Mnemosyne Atlas. In comparing contemporary methods of image data visualization through cultural analytics method of remapping and the turn of the century methodology developed by Warburg under the working title of the “iconology of intervals,” I examine the shifts and continuities that have shaped informational aesthetics as well as data-driven narratives. Furthermore, in drawing parallels between contemporary Cultural Analytics/Media Visualization techniques, and Aby Warburg’s Atlas, I argue that contextual and image color data knowledge should continue to be important for digital art history. More specifically, I take the case study of Warburg’s Panel 45 in order to explore what we can learn through different visualization techniques about the role of color in the representation of violence and the promise of prosperous civil society. Keywords: images as data, Aby Warburg, Cultural Analytics, color, visualization, violence, reconciliation

In their current state, the methods information about the artifact. In this of Cultural Analytics and Digital project, I take on a hybrid Digital Art Humanities provide two radically Historical methodology that combines different ways of interpreting images. Cultural Analytics as articulated by Cultural Analytics calls upon the Lev Manovich with Digital Humanities understanding of images as objects paradigms. I apply this method to with features, and more specifically as the case study of Aby Warburg’s image-data in and of themselves. The Mnemosyne Atlas from 1924, and potential of this computational method to Panel 45 more specifically. The to digital art history was presented twenty-four images in Panel 45, many in detail in Manovich’s essay “Data of which were created by the Italian Science and Digital Art History.”1 fresco painter Domenico Ghirlandaio, Digital Humanities methodologies on comprise the objects for this study. I the other hand rarely analyze images selected Panel 45 because it grapples per se. Instead, they tend to focus on with the relationship between color and the metadata: historical and cultural violence. Using Cultural Analytics, I Images as Data

render Warburg’s images as color-data. an understanding of not only of I argue that this color-data, situated in the formal elements of the images, the theoretical framework provided but also of their cultural meaning.5 by Christopher Johnson’s 2013 Digital While his panels contained no textual Humanities project about Warburg descriptions, Warburg kept notes that titled Mnemosyne: Meandering through listed provisional titles as well as the Aby Warburg’s Atlas2 and in reference overarching theme of each panel. with traditional art historical texts about Domenico Ghirlandaio, speaks The black-and-white photographs of to the larger tropes of violence and civil artworks were pinned either directly to reconciliation. I demonstrate that color- the black fabric, or framed over a white data becomes art historical knowledge mate. They varied in size, proportion of through an in-depth conversation with the depicted image, as well as distance digital humanities practices as well as from each other. Warburg sometimes with already established disciplinary photographed the work of art in its knowledge. entirety. In other instances, he focused on a detail and included only a close- up. Through the use of the close up, Warburg’s he cropped and framed elements that can be then positioned as points Mnemosyne of emphasis within the panels. The panels themselves have not survived by Warburg worked on the and what we have left instead are black AMnemosyne Atlas project from and white photographs of these panels 1924 until his death in 1929. The Atlas instead. consisted of over two thousand black and white photographs of works of art arranged on “sixty-three wooden Panel 45 boards, measuring approximately 150x200cm, covered with black hristopher Johnson conducted cloth.”3 According to Warburg, these Can extensive study of Warburg’s arrangements of diverse types of work. His research resulted in both images attempted to produce “first a book manuscript titled Memory, of all an inventory of pre-coined Metaphor, and Aby Warburg’s Atlas classical forms that impacted upon of Images and a Digital Humanities the stylistic development of the Project titled Mnemosyne: Meandering representation of life in motion in through Aby Warburg’s Atlas. The the age of the Renaissance.”4 It aimed book illuminates Warburg’s use of the to offer a “comparative analysis” of metaphor as a structuring element. visual and cultural tropes spanning Here Johnson argues that Panel 45 from Classic Antiquity to Renaissance, is a “study in contrasts.6” As cited in and engaged the early 20th century Johnson book, Warburg’s notes on this as well. This analysis was built upon panel read:

118 DAH-Journal Images as Data

Superlatives of the language theme of panel 79), and even of gestures.­ Wantonness of as the serving girl in Birth of self conscious­ ness.­ Individual John the Baptist heralds the “heroes emerging out of the all-important theme of the typological grisaille. Loss of nymph whose ‘life in motion’ the ‘how’ of metaphor. At the [bewegtes Leben] animates center of the panel’s twenty- Warburg’s own thinking, the four images are two frescoes panel heightens the lethal by Ghirlandaio from the threats (e.g., plague, tyranny, Torenabouni Chapel, Mas­ war) against the possibility sacre­ of the Innocents and of achieving psychological Apparition of the Angel to ‘balance’ [Ausgleich].7 Zechariah.7 “ According to Johnson, “this The Digital Humanities project panel explores the tension between Mnemosyne offers further insight into violence and passion on one hand the specific images that constitute the and the aesthetic formal properties of panels.8 Here the twenty-four black- grisaille – a technique that relies on and-white photographs of artworks varied in brightness gray palettes and that constitute Panel 45 are named desaturated colors - on the other.9 Here and numbered. They are also linked Warburg explored the role of grisaille to the color version of each artwork. technique as well as color in conveying Johnson’s interpretation of the panel violence. reads: Building upon Johnson’s work, I A study in extremes, panel argue that Panel 45 is indeed a com­ 45, builds on the se­quence of parative­ study of contrasts – contrasts panels 41, 41a, 42, 43, & 44. But that engage both visual and cultural “it does so to signal a perilous tropes. Warburg explores violence and ‘loss’ of metaphoric distance. promise of civil reconciliation rather Here frescos by Ghirlandaio than passion; grisaille and brightly emblem­ ize­ the ‘afterlife’ of saturated color, Classic Antiquity and classi­ ­cal ‘expressive values’ Renaissance. Warburg articulates these in the Renaissance. Yet dyads by positioning images of rebirth Ghirlandaio’s­ use of the on the left side of the panel – such as the grisaille technique­ is not able Birth of John the Baptist by Domenico to fully moder­ ate­ or mediate Ghirlandaio, and images of violence the inten­ ­sity of the passions.­ on the right – such as Massacre of the Thus even as the Massacre of Innocents by Matteo di Giovanni. As the metonymically the Inno­ ­cents two frescoes represented in the center yields to The Blood of of Panel 45 show, the possibility of the Redeemer (while also the social is rooted in a long-standing anticipating the Eucharistic tradition of violence.

DAH-Journal 119 Images as Data

between color and gray scale in relation Panel 45 in Color to violence and possibility of civil reconciliation that Warburg positioned he extensive documentation pro­ in first place as his motivation behind Tvided by the Digital Humanities the composition of Panel 45. Mnemosyne Project (2013) allows for the reconstitution of this panel in color (Figure 1). Cultural Analytics further Cultural Analytics con­firms the importance of the question of color in Warburg’s work. I argue that as a Methodology color is an important agent of meaning in Panel 45 not only because of its aving translated Panel 45 into primacy as a feature for data extraction color, I turn to the methodology in Cultural Analytics, but also because H of Cultural Analytics in order to it was the primary diving principle convert its images into color-data. in the organization of Warburg’s Cultural Analytics is a methodology thought. Warburg questions the spearheaded by Lev Manovich and transience of violence and the promise the Software Studies Initiative, which of social rebirth in the Classic as well “allows the users to work with different as Renaissance periods through an kids of data and media all shown exploration of both color and . grisaille together.”10 It adapted Information Although panel 45 consists of black Visualization techniques used in STEM and white photographs, I argue that he to the fields of Art History, Film and was working with and conceptualizing Media Studies, and Popular Culture, to through works of art in color. The name a few. Information visualization, black-and-white effect is due to the as defined by Lev Manovich, is “a widely photographic reproduction technology used as a tool for understanding data – deployed in the constitution of the i.e. discovering patterns, connections, panels. The s images for Warburg Atla and structure” and as such delivers were pictures of intense color. “new knowledge about the world Digital Humanities projects, such through systematic methods – such as experimentation, mathematical as Johnsons’s Mnemosyne, engage with modeling, simulation.” Furthermore, the context in which Panel 45 with this new method is invested in design, its images exists. They offer relevant as “it involves the visual presentation information on the level of metadata of data in a way that facilitates the about images such as: author, period, perception of patterns.”11 title, etc. Remapping visualization techniques allow for a closer look at Deploying visual databases and the data that constitutes the images: graphic plotting softwareImageJ ( , namely brightness and saturation. which originally was developed by Image visualization thus allows us the National Institute for Health), the to further analyze the relationship Cultural Analytics approach allows

120 DAH-Journal Images as Data

for the comparison of large sets Col­lection Montage – or the creation of visual data through techniques of image-grids where images are ar­ previously deployed by STEM: namely ranged in neat rows and columns with “visualization, visual analytics, data uni­form size and spacing through mining, and information visu­al­­ the Montage function of the ImageJ ization.”11 A prerequisite for the meth­ software;­ (2) Temporal and Spatial od­ology of Cultural Analytics is an Sampling in which the archive of understanding of “culture as data” images is sampled either based on that can be “mined and visualized” – a the date of the image (temporally) or move that distills visual culture into a based on the selection of portions of set of data patterns.12 In their seminal images (spatially); (3) Remapping – the essay “How to Compare One Million plotting of images in order to identify Images?” Manovich, Douglass, and patters across temporal media artifacts Zepel write that “having at our dis­ or series of artifacts. Thus, Cultural posal very large cultural data sets Analytics, with its methodology for which can be analyzed automatically media visualization, “is based on three with computers and explored using operations: zooming out to see the interactive interfaces and visualization” whole collection (image montage), opens up new ways of understanding temporal and spatial sampling, and re­ culture.13 Tracing the grayscale, bright­ mapping (re-arranging the samples of ness, hue, and saturation of images on media in new configurations).”14 In this macro as well as micro scale allows project, I am particularly interested in for the articulation of trends across the possibilities for shaping digital art con­temporary­ visual landscapes as historical methodologies centered on well as across historical periods. In color that Remapping offers, as it is short, Cultural Analytics provides a allows for a wide range of visual land­ computational algorithmic answer scapes to be drafted. to the question: “How to Compare One Million [Diverse] Images?” It is important for new paradigm of ex­ Remapping ploring humanistic big data however to continue to emphasize the cultural Panel 45 context of this new visual landscape in order to offer a critical interpretation n the context of Cultural Analytics, of the significance of this algorithmic Icolor becomes one of the key feature knowledge. components through which analysis is conducted and data attained. Coupled In comparing images, whose origin with art historical knowledge, it pro­ lie in user-generated media, television, vides a crucial look at the role of color film, as well as photography, and in the articulation of violence as well as art broadly defined, Lev Manovich the possibility of civility (Figure 2). In proposes three major strategies: (1) re­mapping Panel 45, I worked with the

DAH-Journal 121 Images as Data

Figure 2. Remapping the images of Panel 45, using the ImageMeasure and ImagePlot modules of the ImageJ Software.

color versions of the images. Using the 14, which included too much black or ImageJ’s module ImagePlot, developed white negative space around the actual by the Software Studies Institute at round-shaped image and thus escaped UCSD and CUNY Graduate School, accurate numerical representation. I first ran analytics on the following This data was then mapped onto a features: Grayscale, Brightness, Satu­ ­ black canvas with the same size as ration, and Hue. Exceptions were made the photographs of Warburg’s panel in the case of image 17, which I was – namely 650x847 pixels (the canvas un­able to locate in color, and image size is a combination of the plotting

122 DAH-Journal Images as Data

area of 550x747 pixels and 50px border of the works. The image number and space). Next, whereas Warburg had the metadata­ correlation can be found in abili­­ty to choose images with different the Appendix to this article. size and scale, ImagePlot’s insists on a uniform image size. I calculated the The remapping plane became a size of the images in the photograph landscape of color in which the images of Warburg’s panel and then reduced coming closest to a grisaille technique that size in half for legibility purposes. lie in its lower right corner, while I arrived at the average width of 65 the images with highest brightness pixels for the size of all images in the and saturation lie in its upper left re­mapping. The X-axis was used for corner. Thus grisaille images occupied the distribution of the brightness via the space of the lower Y-axis, while brightness_median­ (average of grey images with bright colors clustered scale values for the pixels in an image) around the upper Y-axis. As described while the Y-axis was used to display earlier in the article, the grisaille the saturation via saturation_median technique attempted to subdue color (average of saturation, i.e. purity of by decreasing saturation to a point of color for each pixel in an image). gray monochrome composition and Even though brightness_median and by working within varied attributes saturation_median range between of brightness within this single shade the values of 0 and 255 according to in order to express meaning. In other the standard RGB schema, where as words grisaille artwork can exhibit 255 designates highest brightness or black, white, as well as shades of white and purest saturation, I chose a gray as colors and thus tends to stay smaller range that reflected the subset altogether in the low spectrum of of color available in Panel 45. I limited the saturation range –hence on the the X-axis brightness range to 100-200 lower Y-axis. Grisaille images at the and the Y-axis saturation range to 20- same time vary in brightness and thus 150. This restriction indicates that all span along the horizontal X-axis. The images included fall in the mid range in terms of brightness and in the lower images coming closest to a grisaille to mid-range of saturation, given the technique on the remapped canvas overall­ digital RGB color spectrum. In lie in the lower level of the Y-axis. In prefacing my argument about the need this case they are 45-7a – color image of art historical context for unders­ of a bronze fresco, 45-6 – color image tanding image-data, I included to ad­ of sculpture, 45-18 – color image of an di­tional variables. First, I added a pa­ engraving, and 45-19 – color image of rame­ ­ter indicating whether the title of copperplate engraving. These images the artwork signals associations with lie on the outer rim, while 45-4 and 45- vi­o­lence or civility and reconciliation. 10 take the central stage. Second, I included the image numbers in the plotting of the data in order to Surprisingly, the two images with keep reference to the title and author highest saturation and brightness

DAH-Journal 123 Images as Data

levels are positioned adjacent to each helpers in 1485. As Art Historian Steffi other both at Warburg’s Panel 45 Roettgen writes, as well as in the Remapping of the panel. Images 45-10 (Massacre of the Iconographically, the Mas­ Innocents, Domenico Ghirlandaio, sacre­ of the Innocents forms fresco, 1485-90; Florence, Tornabuoni an antithetical counter­ Chapel Santa Maria Novella) and 45-4 “part of the Adoration of the – same as 4a and 4b – (Annunciation Kings, the one con­sti­tuting to Zacharias, Domenico Ghirlandaio, one of the seven joys of the fresco, 1485-90; Florence, Tornabuoni Virgin, the other one of her Chapel, Santa Maria Novella) lie at seven sorrows.…In the Bible the heart of the Warburg’s Panel 45. narrative of the massa­ cre­ of They are positioned in the vertical and the children in Bethle­ hem…­ it took place a year after the horizontal center of the panel. These magi had returned home and two images are of the same size. And Joseph and Mary had fled to both are larger, thus more prominent Egypt. Herod, having heard than the rest of the images included in that a king of the Jews had Warburg’s panel (See Figure 1). In the been born in Bethlehem, remapped canvas, they are positioned had all of the year-old male together at the top left corner. This infants of the city killed – positioning indicates that they carry among them, unwittingly, the highest brightness and saturation his own son, who had been values of all images included in entrusted to a nurse.15 Warburg’s original canvas. One image represents violence, the other – civil This dramatic fresco portrays“ reconciliation. the massa­ cre­ both in the foreground through the main scene as well as on the background as part of the architectural Art Historical relief included in the scene. Taking a closer look at this artwork dem­ Account of on­strates that Ghirlandaio deploys grisaille technique in both planes. In the Ghirlandaio’s background plane grisaille is associated with architecture. In the foreground frescoes plane grisaille comes to represent the living dead: both people and animals Massacre of the Innocents and An­ appear through desaturation pale and nunciation­ to Zacharias are among the bloodless. In the foreground plane, in six frescoes coming out of Tornabuoni contrast to the grisaille of flesh, dress Chapel, Santa Maria Novella in appears to be in bold colors such as Florence, Italy. Both were created gold, blue, purple, and crimson flow by Domenico Ghirlandaio and his across the image.

124 DAH-Journal Images as Data

Figure 3. Visualization the hue and saturation in images 45-4 (Annunciation) on the left and 45-10 (Massacre) on the right

Where as the Massacre is a fresco Florence as well as the assurance that representing totalizing violence, with God’s will salvation will come.16 Annunciation to Zacharias signals the According to the story of Zacharias, beginning of salvation and the pos­ here, the promise of salvation lies in si­bility of civil reconciliation. This prophesy­ of his unborn son. Warburg fresco, featuring the portraits of in­cluded three versions of this image – “several generations of the Tornabuoni one of which presents a tightly cropped family” as representatives of Florence, version of the fresco to exclude the ar­ cele­­brates the prosperity of 1490’s chitectural­ framework of the chapel,

DAH-Journal 125 Images as Data

and another, which under the name panel as well, with fragments of the Angel Appearing to Zacharias presents An­nunciation repeated and scaled a drawing which was presented to the three times. The centrality of these two patron by the painter for approval, works in Warburg’s panel as well as before beginning to paint.17 their positioning in relation color, thus grisaille in the Remapping, articulate Yet the prosperity of the city and vi­o­lence and civility/prosperity, not vi­ the promise of salvation is founded o­lence and passion as Johnson argues upon death and violence. This triumph as the major tropes of this panel. Where is evident in the inscription on the as Johnson argues that “Ghirlandaio’s background­ architectural detail of the use of the grisaille technique is not city wall, which reads: “In the year able to fully moderate or mediate 1490, when the most beautiful cities, the intensity of the passions,” I argue owing its wealth, its conquests, its based on the data provided by the undertakings­ and buildings, enjoyed Remapping, that Ghirlandaio, de­ pros­ ­­perity and peace”.18 Florence per­ ployed similar techniques in the por­ ceived itself as the new Rome and in trayal­ of both sorrow and joy, death order to assert its prominence and and rebirth. While these observations legit­ ­imacy, sought to situate itself are meaningful, taking a closer look at within the Classical tradition. Warburg the art historical background provided hints at this tie as well by positioning by Warburg of both of Ghirlandaio’s Ghirlandaio’s fresco of Brutus, Scaevola, frescoes: Massacre of the Innocents Camillus, which signals precisely the and Annunciation to Zacharias, re­ estab­­­ lish­ ment­ of Rome, immediately to veals a powerful critique of both the the right of the Annunciation in Panel in­stitution and visual representation 45. The Classical tradition, portrayed of civil society founded upon violence, in the background architectural details waged by Aby Warburg. through grisaille, includes violent scenes of triumph in keeping up with Situated in the context of art the legends of the establishment of history, the image-data produced Rome. The civility and prosperity of through the remapping of Panel 45 Florence as the new Rome is thus demonstrates­ that images representing secured first through the legitimacy violence as well as civility/prosperity of the violet triumph in Classical used color in similar fashion. The Tradition, thus contemporaneously, proximity of Massacre of the Innocents as well as simultaneously through the and Annunciation to Zacharias con­ destruction of the enemy as seen in the firms Warburg’s inquiry into the value Massacre of the Innocents fresco. of color in representing violence. Reading these two images as image- Ghirlandaio’s frescoes Massacre data prompts a budding observation: of the Innocents and Annunciation to in the age of Renaissance, death and Zacharias lie at the heart of Warburg’s civil reconciliation do appear to be

126 DAH-Journal Images as Data

portrayed­ with similar color footprint. In translating Ghirlandaio’s Massa­­ Were we to expect grisaille being cre of the Innocents and Annunciation more present in images of death to Zacharias into color-data in and or destruction, reserving color for of themselves, patters of similarities the heroes that emerge out of total between the two images are revealed. violence? The remapping of Panel 45 I created a second remapping that il­lustrates one of Warburg’s central focused on the color visualization concerns: “Individual heroes emerging of these two images, based on the out of the typological grisaille.” And features of hue and saturation. I first further, imbricates this concern in the resized them to the same width of 800 aesthetics of color and the tropes of pixels. I then divided each image into violence and civil reconciliation. 1024 pieces (each piece containing 25 pixels of color data, both images amounting to 2048 pieces. These 2048 Color-Data in Art segments became the objects of the second remapping that I conducted. History The color-data in each segment was measured and mapped accordingly: raditional art historical accounts Y-axis represented hue (hue_median) Tallows us to understand the on the scale of 0 to 255, while the X-axis techniques and material basis of the showed the saturation (saturation_ colors, while Cultural Analytics pro­ median). vides an indispensible overview of the quantity or amassment of hues used. In Both Ghirlandaio’s Massacre of painting the chapel, Ghirlandaio used the Innocents (representing total all of the colors available: “di pingere violence) and Annunciation to facere et exornare cum omnibus Zacharias (depicting the possibility coloribus ut vulgariter­ dicitur posti in of civil reconciliation) exhibit similar frescho.”19 As Julia DeLancey argues, the hue/saturation distributions (Figure artist used “a palette of fully saturated, pure hues, employing few pastel or 3). Their color blueprints signal dark colour and generally only white in dominance of red, orange, and green order to create relief, lending the cycle highly saturated hues, as well as green- a vibrant and legible appearance.”20 It is blue highly desaturated elements. This in­ter­esting that Ghirlandaio’s work in color-data gains significance when Tornabuoni Chapel is notable in terms interpreted through the framework of its pure hues and bright saturations. fo art history. The color footprints Yet two of Ghirlando’s brightly colored demonstrate not only articulations of frescos were chosen by Warburg in the tropes of violence and the promise a discussion about grisaille – hence of civil rebirth in terms of foreground grayness­ and desaturation. Cultural and background planes, but also An­a­lytics sheds additional light into in terms of temporal dissonance: this paradox. Renaissance and Classic Antiquity in

DAH-Journal 127 Images as Data

Renaissance thought respectively. Our Cultural Analytics further extends this movement across the horizontal X-axis argument to mirroring by providing a thus becomes movement through time. comparative framework of the images Desaturated grisaille, almost white as color-data. chips come to signal the perception of the Classic Antiquity in the late 1400s and the deeper reds and oranges Color, Violence, represent the full saturation hues of the Renaissance. This visualization and the Promise illuminates comparatively the way color has come to structure violence of Rebirth of Civil and the possibility of reconstruction across two different temporal periods Society (Figure 3). oving both literary as well as The two frescoes indeed feature Msymbolically back and forth grisaille architectural elements that between grisaille representations of operate as backdrop to the scenes violence in the Classical tradition that mirror both in content as well as and pure saturated colors in the technique sculptures or engravings Renaissance, Warburg’s Panel 45 poses of violence. The backdrop of both important question about the role frescoes projects Classic Antiquity into of color in articulating memory and the age of Renaissance. In representing history as well as distance from the “classical motifs and inscriptions, past in which contemplation about the battles, cavalcades and warriors,” these present and belief in the future can take frescoes allow for an investigation of place. As Giorgio Agamben has argued, the contemporaneity of grisaille.21 This for Warburg “just as the creation and dualism is theorized by Christopher enjoyment of art require the fusion of Johnson. He articulates tension two psychic attitudes that exclude each between form and content is manifested other (‘a passionate surrender to the in the image itself where “the slaughter self leading to a complete identification colorfully depicted in the foreground with the present – and a cool and is repeated monochromatically in the detached serenity which belongs to scenes portrayed on the triumphal the categorizing contemplation of 24 arch in the background.”22 Through things]).” The examination of grisaille repetition of visual motifs both within in relation to color through projects the image as well as within the panel informed by Digital Humanities and Cultural Analytics techniques prompts - a close up fragment of this same us to further ask questions: what is image was pinned on the panel as the role of color in allowing for or well, Warburg illustrates how “the precluding contemplation; how are Renaissance language of gestures we to understand contemporary acts is prefigured in classic antiquity.”23 of totalizing violence; how do we

128 DAH-Journal Images as Data

represent hope for the possibility of plotted data) that connect the nodes social and well as spiritual humanistic (the dots that plot data) in practices of reconstruction? information/media/visualization, but within the images that constitute these The oscillation between tiiden­ ­fi­ca­ nodes in and of themselves. tion and contemplation that Warburg proposes structures the comparative methodology that accounts for both Contextual Visual form and content: hence both data and metadata associated are imbricated Knowledge in a framework of duality. It is in this movement within images as well as n articulating Digital Art History in-between images, in other words in Ithrough methodologies that analyzes the intervals within and in-between data and metadata, I attempted to the visual, that critical thinking integrate cultural and historical can be provoked. In the context of knowledge with data science. While Cultural Analytics, this space of adapting Big Data quantitative contemplation too is visible in the paradigms to the study of art, media, and movement in-between bits of visual popular culture allows for important data. But as much as we try to break new ways of seeing the still and down images into unique molecules, moving image, I argue that this mode into the smallest units of information of visualization is always and already possible, even those miniscule units are imbricated in a complex network: one complex, multilayered, and dynamic, that is not purely algorithmic, but also thus not stable and self-evident. This socio-economic, political, and last is particularly relevant to the study of but not least historical. I argue that visual media, as even a single pixel is Digital Art History should continue invested in color and every color carries its investment in contextual visual within itself complex historical and knowledge by combining quantitative theoretical implications.25 Relational image-data paradigms with traditional art history in order to foster critical visual knowledge is thus manifested interpretations of visual culture. not only in the edges (lines in-between

Appendix

Panel 45. Image 1. Mary enters the Temple, Domenico Ghirlandaio, fresco, 1485-90; Florence, Tornabuoni Chapel, Santa Maria Novella. Panel 45. Image 2. Herod’s Banquet, Domenico Ghirlandaio, fresco, 1485-90; Florence, Tornabuoni Chapel, Santa Maria Novella. Panel 45. Image 3. Birth of John the Baptist, Domenico Ghirlandaio, fresco, 1485-90;

DAH-Journal 129 Images as Data

Florence, Tornabuoni Chapel, Santa Maria Novella. Panel 45. Image 4a&b. An Angel appears to Zacharias, [Annunciation to Zacharias] Domenico Ghirlandaio, fresco, 1485-90; Florence, Tornabuoni Chapel, Santa Maria Novella. Panel 45. Image 5. Angel appears to Zacharias, Domenico Ghirlandaio, preparatory sketch, ca.1485; Vienna, Albertina Museum. Panel 45. Image 6. Offering in the Temple with a Hercules Relief (?) on the Altar, Jacopo and Tommaso Rodari, sculpture, 1491-1509; Como, Como Cathedral. Panel 45. Image 7.a. Battle (with Hercules), Bertoldo di Giovanni, bronze sculpture, ca. 1478; Florence, Museo Nazionale del Bargello. Panel 45. Image 7.b. Battle (with Hercules), Bertoldo di Giovanni, bronze sculpture, ca. 1478; Florence, Museo Nazionale del Bargello. Panel 45. Image 8. Reconciliation of the Romans and Sabines, Bartolomeo di Giovanni, spalliera painting, late 15th cent.; Rome, Galleria Colonna. Panel 45. Image 9. Death of Peter Martyr (Saint Peter of Verona), Domenico Ghirlandaio, fresco, 1485-90; Florence, Tornabuoni Chapel, Santa Maria Novella Panel 45. Image 10. Massacre of the Innocents, Domenico Ghirlandaio, fresco, 1485-90; Florence, Tornabuoni Chapel Santa Maria Novella. Panel 45. Image 11. The Blood of the Redeemer, Giovanni Bellini, oil on wood, ca. 1460- 65; London, National Gallery. Panel 45. Image 12. Brutus, Scaevola, Camillus, Domenico Ghirlandaio, fresco, 1481-85; Florence, Sala dei Gigli in the Palazzo Vecchio. Panel 45. Image 13. Massacre of the Innocents, Matteo di Giovanni, tempera on panel, 1482; Naples, Museo Nazionale di Capodimonte. Panel 45. Image 14.1. Mehmed II, Bertoldo di Giovanni, bronze medal, 1482. Helsinki, National Museum of Finland. Panel 45. Image 14.2. Mehmed II, Bertoldo di Giovanni, bronze medal, 1482. Helsinki, National Museum of Finland. Panel 45. Image 15.a. Resurrection of Christ, Domenico Ghirlandaio (design), with Davide and Benedetto Ghirlandaio Gemäldegalerie, tempera on wood, ca.1494; Berlin, Gemäldegalerie. Panel 45. Image 15.b. Resurrection of Christ, Domenico Ghirlandaio (design), with Davide and Benedetto Ghirlandaio Gemäldegalerie, tempera on wood, ca.1494; Berlin, Gemäldegalerie. Panel 45. Image 16.a. Drawings of scenes from a relief on Trajan’s Column, Workshop

130 DAH-Journal Images as Data

of Domenico Ghirlandaio [both scenes depict the flight of the Dacians], ca. 1490; from a sketchbook, Codice Escurialensis, 28-II-12, fol. 63. Panel 45. Image 17. A prisoner is led before the emperor, Maso di Bartolomeo, relief sculpture, 1452; Florence, Palazzo Medici-Riccardi. Panel 45. Image 18. The Battle at Anghiari, Lorenzo Zacchia, (after Leonardo), engraving, 1558. Panel 45. Image 19. The Phrygian Plague (Il Morbetto), Marcantonio Raimondi (after Raphael), copperplate engraving, ca. 1520.

index.php/projects/how-to-compare (accessed Notes September 1, 2014). 14 Lev Manovich. “Media Visualization: Visual 1 Lev Manovich “Data Science and Digital Art Techniques for Exploring Large Media History” International Journal for Digital Art Collections“Media Studies Futures, ed. Kelly History. Vol 1. No. 1. 2015. p. 12-35. Gates. (Blackwell, 2012). 2 Christopher Johnson Mnemosyne. 2013. http:// 15 Steffi Roettgen.Italian Frescoes: The Flowering warburg.library.cornell.edu/ of the Renaissance, 1470-1510. 3 Christopher Johnson. Memory. P.9 (New York, London, Paris: Abbeville Press 4 Aby Warburg. “The Absorption of the Publishers, 1997), p. 174. Expressive Values of the Past.” Trans. 16 Ibid. Matthew Rampley. Art In Translation 1.2 (July 17 Jean K Cadogan. Domenico Ghirlandaio: 2009). P. 227-8. Artist and Artisan (New York and London: Yale 5 Ibid. University Press, 2000), p. 46. 6 Johnson, p. 98 18 Roettgen 1997, 174. 7 Ibid. 19 Roettgen 1997, 166. 8 Christopher Johnson Mnemosyne. 2013. http:// 20 Julia DeLancey. “Before Michelangelo: warburg.library.cornell.edu/http://warburg. Colour usage in Domenico Ghirlandaio and library.cornell.edu Filippino Lippi” Apollo 145 (1997), p. 18. 9 Grisaille, is a technique of monochrome 21 Micheletti 1990, 48 painting in grey or greyish color that 22 Emma Micheletti,. Domenico Ghirlandaio Renaissance painters used for effects of Florence. Italy: SCALA, 1990). P. 48. sculpture. See Irene Earls. Renaissance Art: A 23 Ibid. Topical Dictionary. Greenwood Press, New York 24 Giorgio Agamben. “Aby Warburg and the (1987) p. 130 Nameless Science.” Potentialities: Collected 10 Lev Manovich,. “Cultural Analytics: Analysis Essays in Philosophy. Trans. Daniel Heller- and Visualization of Large Cultural Data Sets” Roazen. (Stanford: Stanford University Press, [2008]. www.manovich.net/cultural_analytics. 1999), P. 94. pdf (accessed September 1, 2014). 25 The complex history of color is evident in 11 Manovich “Cultural Analytics” 2008. the works of Manlio Brusatin, Philip Ball, 12 Ibid. and Micheal Pastureau to name a few color 13 Manovich et al. “How to Compare One historians. Million Images?” [2011] http://manovich.net/

DAH-Journal 131 Images as Data

Bibliography

Agamben, Giorgio. “Aby Warburg and the Nameless Science.” Potentialities: Collected Essays in Philosophy. Trans. Daniel Heller-Roazen. Stanford: Stanford University Press, 1999: 89-103. Cadogan, Jean K. Domenico Ghirlandaio: Artist and Artisan Yale. University Press, New York and London (2000). DeLancey, Julia. “Before Michelangelo: Colour usage in Domenico Ghirlandaio and Filippino Lippi” Apollo 145 (1997): 18. Huhtamo, Erkki and Parikka, Jussi. “Introduction: An Archaeology of Media Archaeology” Media Archaeology: Approaches, Applications, and Implications. University of California Press, Berkeley and Los Angeles (2011). Johnson, Christopher. Memory, Metaphor, and Aby Warburg’s Atlas of Images. Cornell University Press, Cornell (2012). Manovich, Lev. “Cultural Analytics: Analysis and Visualization of Large Cultural Data Sets” [2008]. http://www.manovich.net/cultural_analytics.pdf (accessed September 1, 2014). Manovich, Lev. “Foreword” in Miguel Lima’s Visual Complexity Princeton Architectural Press, Princeton (2011). Manovich, Lev. “Media Visualization: Visual Techniques for Exploring Large Media Collections“ Media Studies Futures, ed. Kelly Gates. Blackwell, 2012. Manovich, Lev. “Museums Without Walls, Art History Without Names: Visualization Methods for Humanities and Media Studies” Oxford Handbook of Sound and Image in Digital Media, ed. Carol Vernallis. Oxford University Press, Oxford (2012). Manovich, Lev and Douglass, Jeremy and Zepel, Tara “How to Compare One Million Images?” [2011] http://manovich.net/index.php/projects/how-to-compare (accessed September 1, 2014). Manovich Lev. “Data Science and Digital Art History” International Journal for Digital Art History. Vol 1. No. 1. (2015): 12-35. Michaud Philippe-Alain. Aby Warburg and the Image in Motion. Zone Books, New York [2004]. Micheletti, Emma.Domenico Ghirlandaio Florence. SCALA, Italy (1990). Roettgen, Steffi.Italian Frescoes: The Flowering of the Renaissance, 1470-1510. Abbeville Press Publishers, New York, London, Paris (1997). Warburg, Aby M. “The Absorption of the Expressive Values of the.” Past Trans. Matthew Rampley.Art In Translation 1.2 (July 2009): 273-283.

132 DAH-Journal Images as Data

Stefka Hristova, Ph.D., is an Assistant Professor of Digital Media at Michigan Technological University. She holds a PhD in Visual Studies from the University of California Irvine. Her research examines the digital visual cultures of war and displacement.Stefka teaches in the Communication, Culture, & Media Undergraduate Program and the Rhetoric, Theory, and Culture Graduate Program. Correspondence e-mail: [email protected]

DAH-Journal 133