“Standardizing Museum Provenance for the Twenty-First Century” was a lecture that took place at the Yale Center for British Art on February 27th in New Haven, Connecticut. The lecture was centered on the Carnegie Museum of Art’s (CMOA) NEA-funded Art Tracks project. Art Tracks is a digital provenance project that “turn[s] provenance in to structured data by building a suite of open source software tools. These tools transform traditional written provenance records into searchable data, with an emphasis on existing data standards and a strong focus on building tools that are useful (and usable) across multiple institutions.” David Newbury, the lead developer on the project, walked the audience through the process with insight from Louise Lippincott, fine arts curator at CMOA and Costas Karakatsansis, provenance researcher on Art Tracks. In the end, the project represents a definitive example of a project that could serve as a way forward for digital humanities in the museum world. Digital art history, a developing phenomena within academic art research, is much discussed, however, digital art history projects within the work of curators and museum professionals is not as prevalent a topic. Art Tracks is one project that can start this discussion.
Newbury begins with a brief history of Art Tracks in order to explain, “what it does, and what it intends to do.” Digitizing provenance, which is the history of the ownership of a work of art, came from the idea that it would be beneficial to researchers to be able to utilize a centralized resource where they can see where a work has been and how it passed from owner to owner. It began as a simple data visualization project; however, Newbury discovered that any project, given the information, would need to dig deeper. Moreover, the information each work had, the title, date, ownership, and notes in the form of sentences, would need to be converted to computer-readable data. Newbury and his team, in order to convert the data, created software called Museum Provenance, a text parser. The data that Newbury was able to convert suddenly became useful to art researchers looking to incorporate data visualization and digital humanities into their work. This tool eventually allowed researchers to see the history of ownership of a work of art through time and to map that ownership.
The next step in the development of Art Tracks was linked data. Newbury and his team wanted the data to be teachable and to be able to work across different museums. He asks the question, “Why do museums need linked data?” Essentially, Newbury’s linked data initiative was about the role that museums and their collections played in scholarship. Museums are publishers and researchers of scholarship and data, as such data should be shared between institutions.
For Art Tracks, phase two is linked open data, just one of the ways that data is expressed, along with a JSON data structure and standardized text. Phase two is essentially a structure wherein there are different forms of provenance and they all feed into each other. The first level of provenance is “the document,” in which the provenance is written and a timeline is created. The document is necessary in order to convey nuance and context of the written provenance.
The second level is “provenance with entities,” meaning, for example, provenance that shows the relationship between artworks of the same collection. It allows researchers to see the relationship between different artworks that may be owned by another museum. Newbury does warn that one museum, however large their collection, cannot be the authority on everything due to, among many other issues, budget.
The third and final level of provenance is “event-based provenance” in which in provenance is parsed through the events. For Newbury, linked data is important because it helps different museum develop “shared semantics.”
Linked data is a way for museums to communicate with each other and share terminology. However, no one person or even system can know everything there is to know about provenance or document the nuance of any transaction. Linked data could allow for this in the future and it is why he reveals that it could be fourth level of provenance.
The final section of Newbury’s presentation is about the future. How do museums collaborate? How does linked data allow scholars to do work that was previously impossible? He and his team are currently developing the Northbrook Project, in which seven Old Master paintings previously owned by the Earls of Northbrook were, over time dispersed and then later made their way back together when they were acquired by the Carnegie Museum of Art. Newbury and his team want to use collaborative tools such as linked data to reconstruct the collection by collaborating with other museums such as the Smithsonian, the Yale Center for British Art and fifty other museums from around the world who also own works that were previously owned by the Earls of Northbrook.
Newbury and the team at the Carnegie Museum of Art did not initially seek to create a digital humanities project, however they seem to have inadvertently created a project in Art Tracks that adheres to several tenets of digital humanities, including a deep engagement with technology and collaboration. What is not clear is whether or not a scholar, performing art historical research, who uses Art Track, is then, by transference, conducting digital art historical research. The lecture, in the end does not answer, or even fully raise these questions, but the project itself is evidence to cite in a discussion.