"Technological Revolutions and Art History": The Ethical Challenges of Digitization

The Frick Art Reference Library’s 2020–21 four-part symposium “Technological Revolutions and Art History” explores how recent advances in computer vision technologies are impacting the field of art history. In the third session, held on January 14, 2021, participants turned their attention to digitization and discussed how recent technologies from 3D digital imaging to artificial intelligence have the potential to expand audience engagement with cultural heritage while also, ironically, limiting understanding and community connection by omitting context and introducing bias.

To frame these issues of access and bias, former Digital Art History Lead Ellen Prokop interviewed one of the symposium’s organizers, the Frick’s Associate Chief Librarian for Preservation, Imaging, and Creative Services Luciano Johnson, alongside the Andrew W. Mellon Chief Librarian of the Frick Art Reference Library, Dr. Stephen Bury.

Register here for the fourth and final session of the symposium on Thursday, March 11, from 11:00 a.m. to 1:00 p.m. You can watch recordings of the symposium’s sessions on our YouTube channel.

Ellen Prokop (EP): Two of the greatest challenges facing researchers and the staff that supports them are the issues of context and access. What are some of the ways that the Frick Art Reference Library is tackling these areas?

Luciano Johnson (LJ): At the Frick Art Reference Library, we endeavor to stay at the forefront of technology, with the goal of increasing access to our material and data. The most prominent project my colleagues and I are engaged in to that end is our involvement in PHAROS, a consortium of art research libraries working together to consolidate and share our photo archive collections. As part of that initiative, my colleagues are working on transforming data and images using the semantic data framework CIDOC-CRM and utilizing sharing technologies for images, like IIIF. We make as much of our materials as possible available on our website and by contributing data and images to other repositories, like the Internet Archive and Artstor.

Stephen Bury (SB): As the Frick continues to digitize its Photoarchive, a collection of predominantly black-and-white photographs of works of art, the associated metadata records contextual information such as the size of the work of art. Our use of computer vision, moreover, allows similar works to be discovered through an online image search—so context and access are built into this process. As a founding partner of PHAROS, we contribute to a repository of digitized photographs, and image searches across different institutions can offer broader contexts for objects, such as differing attributions or locations.

Library staff member scanning an archival photo at a workstation with lights and an overhead camera
Digitization of Photoarchive materials in action at the Frick Art Reference Library’s Digital Imaging Lab. Photo: George Koelle, 2018
 

EP: What are your thoughts about crowdsourcing and building this strategy into the Frick’s workflow? Is this a valuable way to ensure community engagement, or do the downsides outweigh any advantages?

LJ: Crowdsourcing, especially with recently developed platforms like Zooniverse, can serve as an essential part of strengthening image access. In our case, the primary use has been to increase the ADA accessibility of our images by having interested users create alt text for images and transcriptions of handwriting. There are now impressive tools for natural language processing (NLP) and optical character recognition (OCR), but they lack the accuracy and sophistication of human intervention. There are a lot of people who find satisfaction in working on projects that benefit public education, and having people who are intimately engaged with your collections and can serve as ambassadors for your institution is tremendously advantageous.

SB: Crowdsourcing is an important means of getting audiences involved in a digital project. There is a time penalty, but it pales in comparison with the positives. And crowds can come in many sizes: in one of our archival transcription projects it was a crowd of one. Recently, because of the pandemic and many staff working from home, crowdsourcing projects were one way to keep staff involved.

“Digitized objects increase the demand for connecting to the physical materials—those things that encapsulate Walter Benjamin’s ‘aura.’”

EP: There is a lot of talk about “digital natives” and how they will engage with cultural heritage institutions and their resources in the future. Is the Frick thinking about ways to connect these audiences to more traditional—i.e. analog—research?

LJ: This is a genuinely fascinating question, and in large part it is why we have included speakers like Alonzo Addison [Special Advisor to the Director, UNESCO World Heritage Centre] and Thomas Flynn [Cultural Heritage Lead, Sketchfab] in the symposium. There are questions of authenticity, of experience, that maybe cannot adequately be answered just yet. What does the museum, archive, or library mean as digital becomes the default experience? Consider the flexibility digitization affords in terms of something as basic as zooming in on an object—already we interact differently with an object in its digital form. Now, from our current experience, I think the typical refrain you’ll hear from cultural heritage institutions is that putting digital versions of analog objects in the world generally increases the demand for seeing, feeling, and connecting, in a very human way, to the physical material. There is something about being around like-minded peers, about being in the building, about the journey to the object—those things that encapsulate Walter Benjamin’s “aura.”

SB: The library’s strategy is to introduce the digital native to the real object and the analog researcher to the benefits of the digital. Digital natives, for example, may not understand how an archive is structured, or what the size of a book—folio or duodecimo—says about its intended audience and its availability.

Thank you, Luciano, for the Benjamin cue. In 2019 I wrote an article for a special issue of the open-access journal Arts, “The Artist’s Book in the Age of Digital Reproduction: Walter Benjamin and the Artist’s Book,” in which I argued that digital reproduction had freed the book from serving as our main conduit of information, and that its qualities of weight, texture, size, shape, smell, and reading experience—the lag between recto and verso, for instance—were all revalidated in the digital era. The digital revolution is not just a story of loss but of transformation, of the old as well as the new.

Digitization project mapping data points on Van Gogh's "Starry Night" to create a 3D image of the painting
Screenshot of photogrammetry 3D imaging process of Vincent van Gogh’s Starry Night, courtesy of The Museum of Modern Art, co-sponsor of the symposium
 

EP: What about bias? Are there discussions and projects underway at the Frick to determine how bias might be eliminated in the digital archive?

LJ: The question of bias is significant to any cultural heritage institution. The Frick Art Reference Library is certainly focused on addressing the issue, but it is a slow climb. As we are focused primarily on artists from Europe and the Americas working in the Western tradition from the fourth to the mid-twentieth century, there is an intrinsic level of bias in the collections themselves that we must acknowledge and address. There is a lot of work to be done to understand how our assumptions and inherited understanding of the art represented in our collections often ignores the context and perspective of the rest of the world. So right now it still feels like there are a lot of questions: How do we first break the seal, and then how can we follow through to enrich and expand our understanding to a more holistic and inclusive discussion of art? How do we stop talking about art as if everyone has the same tools or same lenses on hand, or that there is only one appropriate lens or set of tools?

“We have to be very careful: What is digitized becomes the next academic canon.”

SB: Luciano is correct to point out that our historic scope of art in the Western tradition is limiting, as is the supposed West–East division itself. For instance, global trade has been going on probably since the beginning of history, and with it complex networks of cultural interaction. A title in our collection comes to mind: Anna Gannon’s The Iconography of Early Anglo-Saxon Coinage (2003), which traces the influence of Coptic, North African, and Byzantine designs on Anglo-Saxon moneyers—all the way back in the medieval period. The Frick Art Reference Library also has major holdings in costume history, which is rich in global trade influences. In addition to diversifying representation, a major part of decreasing cultural bias is reexamining the Western tradition we’ve inherited.

We have to be very careful: What is digitized becomes the next academic canon, and ensuring that what we digitize reflects an inclusive history is at the forefront of our minds.

Link: Tags: