Renaissance 4/2017 - 1

Georg Schelbert

Art History in the World of Digital . Aspects of a Difficult Relationship

Fig. 1: No digital in the digital humanities? (The 3-spheres model of the Digital Humanities by Patrick Sahle 2015).

Is Digital Art History failing at the Digital Hu- currently more or less a German handbook manities? on digital art history.[3] On the one hand, I agree completely with James Cuno’s state- If one looks at compendia for the Digital Hu- ments, but what he is talking about is not in manities, it is striking that art history and re- fact “the internet”, but rather the “Digital lated subjects, even archaeology, are still Humanities”, or whatever you might call his- rather marginal topics (cfr. fig. 1).[1] At Digit- torical and studies using digit- al Humanities meetings, more and more al methods. Moreover, we should take a contributions are being made that include close look at the situation to understand working with pictures, 3D reconstructions or what this relationship really does look like. maps, but it is noticeable that art historians First, Cuno recapitulates the chal- are rarely responsible for this. James Cuno lenges of digital initiatives in a few precise published a frequently quoted article in words: “Keeping up with the pace of 2012 with the title: “How Art History is Fail- change in the digital world is challenging, ing in the Internet”.[2] Hubertus Kohle even and harnessing its potential can be frustrat- makes it into a kind of motto for his 2013 ing. But the biggest mistake many of us in book “Digitale Bildwissenschaft”, which is the arts and humanities academy can make Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 2

is thinking of that potential only in terms of the case: Although she states that an aston- how we can use the new technology to ishing number of inventories of museums, more quickly and broadly disseminate in- libraries, galleries, and collections have formation. The promise of the digital age is been digitized, she claims that this kind of far greater than that. It offers an opportunity effort “is digitized art history, one built on to rethink the way we do, as well as to deliv- the use of online resources. But no particu- er new research in the arts.”[4] lar changes of thought or critical stance But what are exactly, according to come with this convenience”. Cuno, the prospects of the digital? “The The distinction made here between power of our computers to store massive digital and digitized art history is not only amounts of information and then order and crucial but also fatal for the role of art his- reorder it in a near-infinite number of ways tory within Digital Humanities since it weak- should be producing new paradigms in art ens even well-meaning defenses of art his- historical research. Imagine what Erwin tory.[6] Panofsky or Aby Warburg could have done This dichotomy reflects a certain with our technology.” Such a lack of new concept of Digital Humanities that has its epistemic paradigms was also observed by seeds mainly in the philological disciplines, Johanna Drucker in an article provocatively especially linguistics. When I assert a rather entitled “Is There a ‘Digital’ Art History?”, difficult relationship between art history and cited just as frequently as Cuno’s state- the Digital Humanities, I basically wish to in- ment.[5] She states that digital technologies dicate a certain conflict of concepts concer- did not induce any methodological and the- ning the use of the computer in various dis- oretical changes in the discipline: “A useful ciplines of the humanities. contrast might be drawn between the im- We (still) have to assume a wide- pact of and that of digital spread opinion that the Digital Humanities methodologies. In the 1980s, traditional art are primarily about text analysis. This has to history was upended. , structural- do with the fact that the umbrella term Digit- ism, post-, psychoanalysis, al Humanities was coined in linguistics and Marxism, cultural and critical studies, and other text-based disciplines. Here, com- feminist thinking sharply divided art histori- puter-based operations start from a (given) ans. […] Every aspect of art historical know- corpus, i.e. a text in digital format, for ex- ledge was shaken at its foundations.” She ample, the complete works of Thomas Aqui- demands that “we have to see a convincing nas. Their exploration by Padre Busa in the demonstration that digital methods change 1940s and 50s with the support of IBM is the way we understand the objects of our regarded as the birth of Digital Humanities – inquiry. […] what are the ways of thinking at least in accordance with the common about works of art that arise from digital narrative of computer philologists.[7] Though methods and reconfigure our fundamental it was expensive to build up a digital text understanding of what constitutes a work of corpus with these dimensions at that time, it art? What new research questions can be was not too difficult to implement, as char- asked?” Drucker answers her own question acters and text are almost digital per se. by proposing “that we could situate a work Contrary to art history and related subjects within the many networks from which it (we will come back to this later), the exciting gains meaning and value, and then present work being done with the computer in the the results within complex visual argu- text based disciplines only started when the ments.” In her opinion, however, this is not corpus was ready to be analyzed. In the di- Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 3

gital format, text does not have to be read any more laboriously by human beings, but it can be processed and visualized graphic- ally in various ways. Franco Moretti and his “distance readings” of hundreds or thou- sands of books at a time may already be fa- miliar to the audience.[8] Not only do many colleagues from the philological fields understand “Digital Humanities” exclusively in this way, but art historians have begun to adopt this view as well: so-called “digital art history”, on the one hand, involves analytical methods, whereas “digitized art history” is simply the production of digital representations of works of art.

The liaison between art history and computa- tional methods

I would rather see the so-called digitized art history as an essential part of both digital art history and digital humanities. So I ad- Fig. 2: “Taking possession” of the works of art (here still with pre-digital technology): B. von Tieschowitz and R. Hamann- vocate here for a concept of Digital Human- Mac Lean taking pictures in Vezelay, 1927 (Foto Marburg). ities as a discipline, which also includes the specific processes of art history and related subjects. This also intrinsically includes the In the domain of the digital capture of arti- digitization of their objects, which doesn’t facts, art history is operating completely mean just scanning but also modeling and within its own tradition (fig. 2). From cata- virtualization in a broader sense. Finally, a loguing, to describing, to drawing, to photo- clear distinction to so-called digital art his- graphy, the discipline always looked for tory cannot be drawn: Already storing digital ways to “take possession” of its subjects, images captured from works of art even in particularly replicating and emulating them the simplest database, sorting or classifying in order to have them ready for comparison them, means crossing the border to a meth- and other operations. Neither Wölfflin’s odological use of digital possibilities – in formalistic school, nor the Warburgian cul- other words: crossing the border to digital tural history approach, with its interest in art history – even without sophisticated im- iconography and iconology, would have age-technologies like pattern recognition. been possible without prior transformation in portable media formats. Large, systemat- ically structured image archives, as well as classification systems like Iconclass are un- dertakings of the discipline that anticipated the computer as a device for counting, fil- tering and calculation. It is true that com- Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 4

puters had already appeared on the horizon on the other hand are definitively only man- by the mid-20th century, but initiatives like ageable with digital technologies. While Iconclass were definitively inaugurated be- considered by some experts as a core field fore such technology was actually available. of Digital Humanities,[13] data modeling and This is also true of the Census of Antique knowledge management in art history, actu- Works of Art and Architecture Known in the ally is hardly acknowledged as a specific Renaissance which was started as a double field of action. This has certainly to do with file record system on paper before, in the the specific “resistance” of material objects 1980s, it underwent a digital transformation against formalization,[14] but also with a lack with the help of the Getty Institute.[9] of initiative from the discipline itself.[15] Thus For a long time, the field of the ap- the most comprehensive concepts for de- plication of computer technology was seen scribing artifacts, the CidocCRM, even if it mainly in terms of classifying works of art by was created by an ICOM group, largely was form, iconography, and function – these are developed by non-art historians.[16] As the basic cognitive operations an art histori- every formalization of historical events or an starts with. Due to the required memory material objects needs a – conscious or un- and calculation impact, initiatives that conscious – framework of assumptions and tackled the image data itself came rather strategies, the distinction between “digit- late, despite there having been discussions ized” and “digital” art history makes no on image recognition and similar things in sense here. The same is true if we look at the first meetings organized by IBM in the the emerging field of mapping.[17] This is an late 1960s at the National Gallery in Wash- activity which is as well connected to col- ington.[10] Today, images might be the first lection building and data modeling as it is to thing that comes to mind when talking analytical operations and visualization.[18] about digital art history. Within short time, This is even more the case if thinking about valid representations of artifacts might be subsequent steps like formal reasoning and (at least) three-dimensional, as the fourth di- artificial intelligence (AI). mension, time, has to be considered, too. It is certainly true, as Matthew Lin- But it is not just about the digital represent- coln states, that “the sheer numbers of ex- ation of the “measurable” aspects of the tant art objects (particularly multiplicative artwork. Artifacts are also parts of historical works such as prints and photographs) events, as Matthew Lincoln expressed it, present art historians with problems of scale they “are both actors in, and indices of, a that quantitative methods promise to ad- host of historical trade networks patronage, dress.” But again: the “resistance” of mater- gift-giving, commerce, colonization, theft, ial objects, i.e. their history and their refer- and other forms of physical movement and ence to cultural or scholarly concepts, re- exchange”.[11] This requires not only neces- quire attentive data acquisition and man- sarily data concepts which are able to ex- agement of data. Until now one can speak press a certain narrative but also dimensi- only of a few cases of “big data” and one ons, like probability and variants of hypo- should do this with caution in order not to thesis.[12] simplify one’s methodology.[19] It is not sur- At this level of complex knowledge prising that – even before digital art history management, new and exciting fields are could establish itself – the first harsh critics emerging which on the one hand are rooted from a humanist viewpoint, like the article in our disciplinary tradition of cataloguing, “Against Digital Art History” by Claire Bish- description in words and images but which op, have been heard.[20] Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 5

and comparison of images itself is not too difficult for the computer if it uses colour analysis or outline detection. The challenge consists in bringing together the intellectual concepts of culture with the formal aspects of the artifacts. Only if we have enough data about the cultural context and sufficient cal- culation power at our disposal, more com- plex analysis will be possible. Further progress is to be expected, but it should be clear that the application of computational methods requires a certain amount of structured and comparable data. Concerning the idea of Digital Humanities, rather than making a division between digi- Fig. 3: The machine already better than human beings: tal and digitized art history, we should move Ahmed Elgammal and Babak Salehat, Rutgers University towards a circular concept consisting of (MIT Technical Review May 11, 2015). building up structured digital collections on the one side and analytical methods on the other (cfr. fig. 4). So, to a certain extent, it is Notwithstanding contrary announcements up to the discipline itself to claim a much (cfr. fig. 3)[21] and despite the fact that al- more central position in the Digital Humanit- gorithms for pattern recognition are now ies by increasing the emphasis of data cre- quite advanced, computers are still rather ation and curation which are traditionally weak in analyzing works of art. The reading strong fields of art history.

Fig. 4: Digital Humanities as circular model of aggregation and analysis (G. Schelbert). Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 6

Of course, this is also about creating a new graphies and dealer networks in the nine- mindset regarding data and data analysis teenth-century London art market are a rep- within art history itself. Notwithstanding its resentative example for a data-driven meth- technological traditions, there is a strong odology which focuses at the same time on anti-digital attitude in the discipline today spatial and temporal visualization. (as we already heard from James Cuno) which should be taken seriously and ad- dressed accordingly.

Need for opening art history towards a data driven and collaborative methodology

The challenges start with our research and publication practices. If the scholarly narrat- ive, printed as book or article, remains the only target of academic practice, it might be difficult to establish something like digital art history. We should not just see data as primary sources, we should also accept the various stages of data creation as a genuine part of research.[22] But, right now, we are far from such Fig. 5: A data-driven project: Mapping nineteenth-century London’s art market (Pamela Fletcher and Anne Helmreich). a good relationship with data: Driven largely by theoretical paradigms (as also mentioned by Johanna Drucker), the discipline has A data-driven strategy in these fields and shifted more and more to the production of others will not just facilitate interdisciplinary rather detached narratives that tend to fo- and multilingual communication and collab- cus on individual “artistic positions” based oration[26] but promote also new forms of on single aspects, as opposed to clearly publication, such as databases, online cata- defined objects of investigation. It is curious logues, maps, or multidimensional datasets. that particularly in the field of contemporary There is no cause for alarm for the tradition- art – and even computer art – there seems alists, however, as the scholarly narrative to be the greatest distance from the con- will maintain its status – perhaps more and crete methodologies of digital art history.[23] more as a personal view addressed to par- On the other hand, there are a lot of ticular communities on the respective field emerging (or re-emerging) fields that require of study. more fact-based research, like provenance Another crucial point is that data- research, history of collections, art market, driven and technology-enabled research personal network research etc. Also both, cannot be tackled sufficiently by the kinds the so-called spatial and material turns re- of solo practitioners typical in the humanit- quire strategies that are based on data.[24] ies. Only multidisciplinary teams of re- Mapping spatial data and visualization of searchers together with data engineers can temporal aspects in timelines currently are build up comprehensive corpora and can seen as promising methodologies.[25] The take full advantage of their content, as was research on the interference of city geo- extensively shown by Diane Zorich in a re- Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 7

port for the Kress Foundation and the only partially feasible for scholarly work.[29] Rosenzweig Center.[27] As such – at least as A particularly promising alternative an option – the discipline should learn to seems to exist in the Digital Humanities labs collaborate more. As collaboration via the already in place in North American Univer- internet can be virtually global, concepts of sities, such as the Rosenzweig Center or open access and data sharing must be- Duke’s Wired lab, which are strong in the come more familiar. This should also in- development of appropriate tools. However, clude a culture of academic recognition. In there are few such laboratories doing note- the current academic climate, most gradu- worthy work in the field of art history. ate students running out of funding or as- In my opinion, the discipline can no sistant professors racing against the tenure longer avoid investing more in such concre- clock are understandably more committed te structures – ones that must be tightly to personal publications than to the com- connected with museums, libraries, and ar- mon goal of creating repositories. chives – in order to establish a stable con- Last but not least, one could name cept of digital art history. Digital art history another problem concerning the institutional is not just the momentary application of al- and infrastructural situation in art history. It gorithmic processing to particular research is the almost strict separation of the prac- questions, nor is it just the consultation of tice of art history into the two fields of mu- digital sources; it is also the continuous ag- seums and universities. Thus much of “di- gregation of knowledge in digital form, in- gital” – mostly cataloguing and digitizing, cluding a widespread digital virtualization of but also development of standards like artifacts that must be kept permanently ac- CidocCRM and database technologies – is cessible for any kind of research, like books done exclusively in the museum world, in a library. while classroom activity is generally limited to consultation of online resources and Notes blogposts with results. There might be 1. Even if one can note a shift towards non- many reasons for this, it is not simply be- textual matters within the last 10 years, there cause it is generally not possible to bring art is still a notably minor presence of object and image-related topics in the handbooks. Cfr. objects into the classroom, as the situation the considerations of an editor of one of the in archaeology shows. Something like ar- earliest digital humanities handbooks, Susan chaeo-computing seems to be an accepted Schreibman (Susan Schreibman, Ray field within the discipline. Siemens and John Unsworth (eds.), A Com- Research institutes with close con- panion to Digital Humanities, Oxford 2004 nections to museums could bridge this gap. (URL http://www.digitalhumanities.org/com- [28] There are some that are already doing panion/, 10-12-2017): Susan Schreibman, Di- so, including important ones like the Getty gital Humanities. Centres and Peripheries, in: Research Institute in Los Angeles or the Historical Social Research Vol. 37, No. 3, 2012, pp. 46-58 (DOI: 10.12759/hsr.37. RKD (Rijksbureau voor Kunsthistorische 2012.3.46-58, 10-12-2017). The illustration Documentatie) in Den Haag. Despite their (fig. 1), taken from Patrick Sahle, Digital Hu- impressive technical and organizational manities? Gibt’s doch gar nicht!, in: Grenzen capabilities, however, they are not equipped und Möglichkeiten der Digital Humanities, ed. to offer open services for the broad com- by Constanze Baum and Thomas Stäcker, munity. Openly available infrastructures 2015 (= Sonderband der Zeitschrift für digita- from private or commercial initiatives such le Geisteswissenschaften, 1) (text/html For- as Wikipedia, Wikidata or Google maps are mat. DOI: 10.17175/sb001_004, 10-12-2017) does not show “digital art history”. Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 8

2. James Cuno, How Art History Is Failing at the of CIDOC, Dresden, 6.-11.9.2014, (URL Internet, in: The Daily Dot, Nov 19, 2012, http://www.cidoc2014.de/images/sampledat (URL: https://www.dailydot.com/via/art-his- a/cidoc/papers/L-1_Kuroczynski_paper.pdf, tory-failing-internet/, 10-12-2017). 10-12-2017). 3. Hubertus Kohle, Digitale Bildwissenschaft, 13.Dominic Oldman, Martin Doerr, Gerald de Glückstadt 2013 (DOI: 10.11588/artdok.0000 Jong, Barry Norton and Thomas Wikman, 2185, 10-12-2017). Realizing Lessons of the Last 20 Years. A 4. Cuno 2012, How Art History Is Failing. Manifesto for Data Provisioning & Aggrega- 5. Johanna Drucker, Is There a “Digital” Art His- tion Services for the Digital Humanities (A Po- tory?, in: Visual Resources. An International sition Paper), in: D-Lib Magazine, July/August Journal of Documentation, Vol. 29, No. 1-2, 2014, Vol. 20, No. 7/8, (DOI:10.1045/ju- pp. 5-13 (DOI: 10.1080/01973762.2013. ly2014-oldman, 10-12-2017); Michael Pio- 761106, 10-12-2017). trowski, Digital Humanities, Computational 6. Benjamin Zweig, Forgotten Genealogies. Linguistics, and Natural Language Pro- Brief Reflections on the History of Digital Art cessing. Lectures on Language Technology History, in: International Journal for Digital Art and History, March 2016 (URL http://stp.ling- History 1, 2015, p. 38-49, (DOI: 10.11588/ fil.uu.se/~nivre/docs/michael_piotrowski_201 dah.2015.1.21633, 10-12-2017) and Max 6.pdf, 10-12-2017). Marmor, Art History and the Digital Humanit- 14.Cfr. Matthew Lincoln in Drucker/ ies, in: Zeitschrift für Kunstgeschichte 79, Helmreich/Lincoln/Rose 2015, Digital Art His- 2016, p. 155-158 (the latter is an answer on tory: “But physical objects resist both the Kohle 2016, which is rather a list of recent structured description and the abstraction projects). these methods rely upon. Compared to 7. Cfr. Edward Vanhoutte, The Gates of Hell. structured data for libraries’ generally homo- History and Definition of Digital | Humanities | geneous collections of books, museums do Computing, in: Melissa Terras, Julianne not yet have interoperable standards for de- Nyhan and Edward Vanhoutte (eds.), Defining scribing their heterogeneous collections of Digital Humanities. A Reader, London 2013, unique objects. While library data are pro- p. 121-156, p. 126ss. duced through broad consensus about the 8. Franco Moretti, Distant Reading, London facts of a book’s publication and classifica- 2013. tion, knowledge about historical objects 9. Cfr. the project’s website (URL http://www.- tends to be advanced through iterative and census.de/census/project, 10-12-2017). conflicting scholarly argumentation – a pro- 10.These early ventures are recently studied by cess that is difficult (though not impossible) Margarete Pratschke (cfr. Margarete Pratsch- to model as structured data.” ke, Wie Erwin Panofsky die Digital Humanit- 15.Cfr. the brief summary given by Diane M. ies erfand. Für eine Geschichte und Kritik di- Zorich, Transitioning to a Digital World. Art gitaler Kunst- und Bildgeschichte, in: kriti- History, Its Research Centers, and Digital sche berichte, 3, 2016, p. 56-66). Scholarship. A Report to the The Samuel H. 11.Matthew Lincoln, in: Johanna Drucker, Anne Kress Foundation and The Roy Rosenzweig Helmreich, Matthew Lincoln and Francesca Center for History and New Media George Rose (eds.), Digital Art History. The American Mason University, June 2012 (URL Scene, in: Perspective, 2, 2015 (DOI : http://www.kressfoundation.org/uploaded- 10.4000/perspective.6021, 10-12-2017), p. 5. Files/Sponsored_Research/Research/Zorich_ 12.Cfr. the endeavours of the working group “Di- TransitioningDigitalWorld.pdf, 10-12-2017), gitale Rekonstruktion” to reconcile these de- p. 22-24. manding aspects with classical 3D modelling: 16.CidocCRM was developed by the CRM spe- Piotr Kuroczyński, Digital Reconstruction and cial interest group under the direction of the Virtual Research Environments. A Matter of physicist Martin Doerr and became an ISO- Documentation Standards, in: Access and standard in 2006. ICOM is the International Understanding – Networking in the Digital Council of Museums. The responsible com- Era. Proceedings of the Annual Conference mittee was CIDOC (Comité International pour Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 9

la Documentation). Cfr. the very instructive Conference of the International Alliance of “manifesto” Oldman / Doerr / de Jong / Digital Humanities Organizations (DH 2017), Norton / Wikman 2014, Realizing Lessons of Montreal, Canada, 2017 (URL https: the Last 20 Years, which addresses a per- //dh2017.adho.org/abstracts/525/525.pdf 10- spective clearly beyond art history. 12-2017); Peter Bell and Björn Ommer, Digit- 17.Paul B. Jaskot, Anne Kelly Knowles, Andrew al Connoisseur? How Computer Vision Sup- Wasserman, Stephen Whiteman and Ben- ports Art History, in: Connoisseurship nel XXI jamin Zweig, A Research-Based Model for secolo. Approcci, Limiti, Prospettive, ed. by Digital Mapping and Art History. Notes from Alina Aggujaro and Stefan Albl, Rome 2016, the Field, in: Artl@s Bulletin, Vol. 4, Issue 1, p. 187-200. Spring 2015. 22.Of course, “data”, “facts” and “information” 18.Matthew Lincoln, in: Drucker/Helmreich/Lin- should be seen in a functional and heuristical coln/Rose 2015, Digital Art History. sense, not in an ontological one. 19.Big data is usually “flat” data on a low se- 23.An explanation for this could be the over- mantic level. This is also the case in the ex- whelming amount of available material. Case amples given by Lev Manovich, Data Science studies and almost generic narratives avoid and Digital Art History, in: International direct interaction with this material. Journal for Digital Art History, 1, 2015, p. 12- 24.Cfr. also the proposal of a more “forensic” art 35 (DOI: 10.11588/dah.2015.1.21631, 10-12- history of Nuria Rodríguez Ortega, Getty 2017, URL http://journals.ub.uni-heidelber- Voices: It’s Time to Rethink and Expand Art g.de/index.php/dah/article/view/21631, 10- History for the Digital Age, in: IRIS, March 5, 12-2017), and Harald Klinke, Big Image Data 2013 (URL http://blogs.getty.edu/iris/its-time- within the Big Picture of Art History, in: Inter- to-rethink-and-expand-art-history-for-the-di- national Journal for Digital Art History 2, gital-age/, 10-12-2017), and for “bridging art 2016, p. 14-37 (DOI: http://dx.doi.org/10. history, computer science and cognitive sci- 11588/dah.2016.2.33527, URL http://journ- ence” as said by Raphael Rosenberg, als.ub.uni-heidelberg.de/index.php/dah/art- Bridging Art History, Computer Science and icle/view/33527, 10-12-2017). Cognitive Science. A Call for Interdisciplinary 20.Rather critical and presupposing a highly Collaboration, in: Zeitschrift für Kunstge- biased concept of computational methods is schichte, 79, 2016, p. 305-314. Claire Bishop, Against Digital Art History, Hu- 25.Pamela Fletcher and Anne Helmreich, manitiesFutures, Franklin Humanities Institute Local/Global. Mapping Nineteenth-Century at Duke University, 2017 (URL: https://hu- London‘s Art Market, in: Nineteenth-Century manitiesfutures.org/papers/digital-art-history, Art Worldwide, Vol. 11, No. 3, Autumn 2012, 10-12-2017). (URL http://www.19thc-artworldwide.org/au- 21.Cfr. Meghan Rosen, Computer Program Re- tumn12/fletcher-helmreich-mapping-the-lon- veals Artists’ Influences. Algorithm Deciphers don-art-market, 10-12-2017), cfr. also Mat- a Painting’s Style by Contents, Composition, thew Lincoln in Drucker / Helmreich / Brushstroke, in: ScienceNews, October 13, Lincoln / Rose 2015, Digital Art History. 2014; Babak Saleh and Ahmed Elgammal, 26.Multilinguality is a critical issue. A more data- Large-scale Classification of Fine-Art Paint- driven research could shift communication ings. Learning the Right Metric on the Right from scholarly papers to exchange of primary Feature, arxiv.org, May 2015, (URL http://arx- source material, measurement data and short iv.org/abs/1505.00855, 10-12-2017). Well re- statements which are less tied to lingual and flected projects however admit that they cultural traditions. match just single aspects of the analysis of 27.Zorich 2012, Transitioning to a Digital World. images or prefer to speak about the com- As a result, the Kress Foundation published puter “supporting” art historical research: cfr. Stephen Bury (with Ralph Baylor, Samantha Leonardo Impett and Susanne Süsstrunk, Deutch, Sumitra Duncan, Julie Ludwig, Ellen From Mnemosyne to Terpsichore. The Bil- Prokop and Louisa Wood Ruby), Art History deratlas after the Image, in: Premiere Annual in Digital Dimensions. The White Paper, 2017 (http://dah-dimensions.org/report/, 10-12-2017). Georg Schelbert Art History in the World of Digital Humanities kunsttexte.de 4/2017 - 10

28.Almost identical considerations were already digital documentation. Only then sufficient data expressed by Matthew Lincoln in will be available for creating new insights and Drucker/Helmreich/Lincoln/Rose 2015, Digital knowledge in art history within a digital space. Art History; cfr. also Zorich 2012, Transition- ing to a Digital World. 29.Cfr. Martin Poulter, Wikidata – the New Hub for Cultural Heritage, in: Wikimedia UK-Blog, 20 Jan. 2017, URL https://blog.wikimedia.or- Author g.uk/2017/01/wikidata-the-new-hub-for-cul- tural-heritage/ 10-12-2017), and Jens Ohlig Georg Schelbert is the head of the media library and Georg Schelbert, Data Partnerships in of the Department of Art and Visual History at Wikidata. Project Durchblick, in: Wikimedia Humboldt-Universität in Berlin. DE-Blog, 21 Aug. 2017 (English version) (URL https://blog.wikimedia.de/2017/08/21/data- partnerships-in-wikidata-project-durchblick/, 10-12-2017). Title

Georg Schelbert, Art History in the World of Di- Figures gital Humanities. Aspects of a Difficult Relation- ship, in: Critical Approaches to Digital Art His- Fig. 1: No digital art history in the digital humanities? (The 3-spheres model of the Digital Humanities by tory, ed. by Angela Dressen and Lia Markey, in: Patrick Sahle 2015). kunsttexte.de, Nr. 4, 2017 (10 pages), www.- Fig. 2: “Taking possession” of the works of art (here kunsttexte.de. still with pre-digital technology): B. von Tieschowitz and R. Hamann-Mac Lean taking pictures in Vezelay, 1927, Foto Marburg. Fig. 3: The machine already better than human be- ings: Ahmed Elgammal and Babak Salehat, Rutgers University (MIT Technical Review May 11, 2015). Fig. 4: Digital Humanities as circular model of aggre- gation and analysis (G. Schelbert). Fig. 5: A data-driven project: Mapping nineteenth- century London’s art market (Pamela Fletcher and Anne Helmreich).

Abstract

This text can be considered among a number of recent statements regarding digital art history. It serves as a contribution to an ongoing discussi- on, certainly not as an exhaustive analysis of the field or its history. The basic thesis here is that the popular separation (Pias, Kohle, Drucker) between “digitized” and “digital” art history is disadvantageous to the discipline, particularly for the positioning of art history within digital hu- manities. “Digital art history” is not possible wi- thout extensive activity in both digitization and