Travel log: A new digital afterlife for museum exhibitions

Art and other cultural heritage objects in Yale’s renowned collections tend to travel. The university’s museums and libraries frequently loan objects to one another, as well as to institutions across the country and abroad, for display in public exhibitions.

The journeys of these pieces can generate commentary in the form of exhibit labels, catalog essays, critic’s reviews, media coverage, and even social media posts by museum patrons. All this information, if amassed and made accessible, could provide humanities scholars valuable insights into how works of art are circulated, contextualized, interpreted, and perceived.

A group of Yale scholars and data engineers is doing just that. The Yale University Library, the Yale Center for British Art (YCBA), and Yale University Art Gallery (YUAG) have partnered with Edinburgh University, Oxford University’s E-Research Center, and Ashmolean Museum of Art and Archaeology in compiling a digital project that will use advanced computational techniques — such as text mining and machine learning — to capture all sorts of exhibition data and allow museums to make it easily accessible and shareable to scholars and the public.

“We look forward to working closely with Edinburgh and Oxford to provide our cultural heritage expertise and data to this international project on connecting exhibitions,” said Susan Gibbons, vice-provost for collections and scholarly communication. “It furthers collaborations and research across the community and contributes directly to internal keystone projects here at Yale, such as our cross-collection discovery system, which will allow scholars and the public to easily search Yale’s collections across our museums and libraries.”

The project, funded in part by a grant from the National Endowment for the Humanities (NEH) and the Humanities Research Council in the United Kingdom, will gather exhibition records from Yale’s museums and other institutions using a data model created by the Linked Art collaboration, a community of museum and cultural heritage professionals focused on making it easier to maintain and share art-related “metadata” — information about the objects in a museum’s collection. The project team members not only will collect the data but will reconcile it with the differing data-management systems that cultural institutions use so that scholars and the public can access it seamlessly across those systems and institutions.

“Museums have a long tradition of collaboration when it comes to exhibitions, lending objects nationally and internationally at significant risk and expense,” said Robert Sanderson, director of cultural heritage metadata projects in the Office of the Vice Provost at Yale and a member of the project’s team. “This project will establish the tools needed to ensure the digital records of those objects are connected across organizational boundaries, bringing added value for researchers and patrons.”

A painting at the YCBA, “British Gentlemen in Rome” by 18^th-century artist Katharine Read, illustrates how artworks travel — and the potential benefits of capturing and sharing information on when they are exhibited. The painting, which depicts a half-dozen British men congregating near the Colosseum, was one of 19 pieces the museum lent to Ashmolean Museum in 2012 for an exhibition. It was also shared across campus, to YUAG, for an exhibition in 1965, and to the University of Arizona Museum of Art in 2015 for an exhibit on Rome. Each of these shows generated catalogs, coverage, and commentary.

Mining these texts will provide useful information on the objects presented in exhibitions and how the audiences reacted to them, said Peter Leonard, director of Yale’s Digital Humanities Lab and the principal investigator on the NEH grant. This information can help researchers discern an art object’s provenance, how artistic styles developed and changed over time, trends in exhibition themes, and patterns in lending among museums, he said.

“Acquiring the data needed to understand how artists and their works are presented, contextualized, and discovered through exhibitions currently takes a lot of effort,” Leonard said. “Printed exhibition catalogs and social media posts are two examples of what we’d like to transform into linked, open and usable datasets for art history. Our goal is to make a broad range of information about exhibitions available to anyone who’d like to access it.”

Museums will be able to incorporate this new metadata into their collection management systems and maintain datasets that can be shared with scholars, Sanderson said. For instance, exhibition data could appear with the other metadata in a painting’s collection record in a museum’s online database. And it could provide links to objects from other institutions that were once exhibited alongside the painting, he added.

The Yale members of the project team will focus their initial work on compiling information on exhibitions held at the YCBA and YUAG, which each have distinct collection management systems and practices, he said. That data will eventually be incorporated with data from the Ashmolean Museum and other partners in the United Kingdom.

Data sources for the initial phase will include the full text of exhibition catalogs published by Yale University Press as well as Yale’s London-based Paul Mellon Centre’s “Chronicle 250,” an online publication that provides searchable, full-text versions of the Royal Academy’s annual summer exhibitions going back 250 years. Edinburgh University will take the lead on the project’s text mining component.

In addition to the Yale museums and the Ashmolean Museum, a handful of other major institutions, including the Museum of Modern Art in New York, the National Gallery in London, and the National Gallery of Art in Washington, D.C. have agreed to participate in the project. Once the team has established connections between the objects and exhibitions at these museums, it can begin to incorporate data from others into the network, Sanderson said.

“Yale’s museums will provide a strong dataset to get us started,” he said. “Once we prove the concept, we can invite more and more museums to share their data and make connections.”