Order of Multitudes

“Deep, Broad, and Out”: Paul Messier on Collections as Datasets

Paul Messier is the founder and Pritzker Director of the Lens Media Lab at Yale’s Institute for the Preservation of Cultural Heritage (IPCH), and as of January 2020, Chair of the IPCH. The founder of several private companies dedicated to cultural heritage preservation, he has published widely on conservation and holds two patents for innovative techniques of characterizing cultural materials. He recently completed a multiyear initiative to establish a department of photograph conservation at the State Hermitage Museum in Saint Petersburg, Russia, funded by the Andrew W. Mellon Foundation. Paul and I sat down to discuss his ideas about cultural heritage collections as “datasets” and how eBay helped create a globally-important collection of photographic materials.

Sarah Pickman: Tell us a bit about yourself. What is your educational and professional background, and what are your research interests, broadly?

Paul Messier: I am an art conservator. Since 1985, I have specialized in the conservation of photographs. After acquiring the requisite hand skills and education, in 1994 I set up a private conservation practice in Boston, working mostly with galleries, auction houses, and collectors. I always considered my practice both as a means to earn a living and as a creative platform for training emerging conservators and original research. In 2015, I came to Yale to establish the Lens Media Lab within the Institute for the Preservation of Cultural Heritage. The move was fueled by a gift to Yale from John Pritzker, a deeply perceptive collector of photography with whom I have worked since 2007. 

SP: Part of your career has been spent assembling a collection of historic photographic papers. Can you talk a bit about how you became interested in these materials, and why you thought they would be an important historical source? 

PM: My interest in photographic papers as a scholarly resource began with my work exposing the infamous Lewis Hine authenticity scandal that broke in 1999. Through this research, I realized there was no great reference archive for this material and if I had a desperate need, in time, others would as well. Two other intersecting trend lines helped immensely: the emergence of eBay as a platform to aggregate the world’s ephemera (a.k.a. junk) and the breakup of countless analog photography darkrooms. I had a solid first mover advantage and swept the table for many years. It was thrilling. The collection, as it stands, arguably is the largest of its kind in the world and was acquired by Yale, again through the insightful generosity of John Pritzker. 

SP: Why do you think there had been so little attention paid to these papers when you started your collecting activities? 

PM: Principally the collection is boxes and boxes of blank sheets of paper, cast off as artifacts of an obsolete printing medium. That this material could shape up into a scholarly resource takes a real leap of faith.   

SP: The Lens Media Lab (LML) at Yale, which you founded, “approaches cultural heritage collections as datasets.” Can you speak about what this approach entails, and how it opens up new insights into collections? How does it differ from more traditional approaches to studying material culture?

PM: Photography, a medium of multiples if there ever was one, is the grounding of the LML.  Despite overproduction, singular prints, emerging for a host of largely arbitrary reasons, garner disproportionate focus in the academy and the market. There are many good reasons for this tendency, maybe including the art historical practice of “deep looking” as well as a lack of tools for “seeing” at scale. At the LML, we build these tools. With a grounding in the reference collection, we can visualize the material choices of a photographer over the course of a career and how these choices relate both to other photographers, like teachers and mentors, and to the range of possible materials choices available for a given period of time. In many ways, there is no difference from other ways of studying the material world. Natural history collections, for example, were formed explicitly as data and though the term “dataset” might not have existed in the nineteenth century, the concept of looking for trends within and across collections was foundational. Thinking about art history, this idea is embedded in scholarly practice where Berenson’s encyclopedic collecting of images, turns into university slide libraries, then ArtStor, and so on. Lacking the tools and understanding to meaningfully extract and interpret patterns, the visual and material cultural communities are maybe a little behind. Perhaps there is some resistance for art historians to “see at scale,” as it requires unfamiliar tools and collaborations, but we love that sort of motivation. 

SP: When it comes to material objects as datasets, how do you think about size and scale? How many objects do you need before you have a large enough dataset to work with? What are the considerations, such as media, age of the objects, or others?

PM: Bigger is better. The challenge is minimizing the time and cost required to safely and repeatably “peel off” data (whether physical, chemical, or visual) from cultural heritage collections. Unlike natural history specimens, acquired to be poked and probed, there is a different sensibility, and ethical obligation, regarding works of art. The LML reference collection is very much the baseline of our research. We have roughly 7,000 reference samples, most of which have measurements against a range of important characteristics. These data can contextualize other objects, whether a single print or a collection of thousands. Looking across media, we are interested in key visual markers, like texture, but also patterns that relate to manufacture, like the chain and laid lines in handmade papers. We also are seriously interested in surfacing patterns in the context of preservation and have a project underway to process lightfastness data to help understand display impact. The fact that this work also shows patterns indicating fundamental material properties is a major bonus.  

SP: In January, you became the chair of Yale’s Institute for the Preservation of Cultural Heritage, of which the Lens Media Lab is a part. What are some projects or directions you’re hoping to pursue at IPCH over the next few years?

PM: I like to say that IPCH is all about “deep, broad, and out.” Deep: we have the capacity to do forensic-level analysis on singular objects where combinations of elements and compounds often tell the story. Broad: we are imagining how these data are shared, structured and interrogated to reveal patterns and new insight. Out: through outreach and partnerships, within and outside the university, we seek to connect collections by revealing origins and histories that show civilization is interwoven and evolving. An ideal project would have all of these components. Right now, with colleagues at the Yale Center for British Art we are working hard to understand fascicles of William Henry Fox Talbot’s Pencil of Nature. Our data is a means of seeing Photography’s Big Bang – what this object was, what it is, and what it’s becoming. Equally important, this project is provoking the Institute to think hard about how we architect the pipeline of “deep, broad, out.”