August 9, 2019
Since its founding in 2014, the Digital Art History Lab (DAHL) has served the public with workshops and symposia to introduce the possibilities that the digital world holds for art historians. During these events, DAHL staff have encountered a wealth of enthusiasm but a lack of workable art datasets. Thus, we are excited to announce the release of two datasets, the Montias database and a Vermeer dataset, on GitHub, an online repository and hosting service built for collaboration. The datasets are available in standard CSV and JSON format, making them universally accessible.
Dutch Art as Data
Compiled by Yale professor John Michael Montias, the Montias database records inventories in 17th-century Amsterdam. These inventories catalog 51,071 art objects with details about subject matter, artist, location, and even buyer and seller. Economists Frederico Etro and Elena Stepanova have already used this data to inform two articles, "The Market for Paintings in the Netherlands During the Seventeenth Century" (2013) and "Entry of Painters in the Amsterdam Market of the Golden Age" (2016). Both look at historical art markets to determine how price and profit might have affected innovation and specialization.
The Vermeer dataset, a compilation of all the works of art that have been or are currently attributed to Johannes Vermeer (1632–1675), is available as a complete dataset or broken up into individual files. The complete Vermeer dataset includes current attributions, past attributions, and information about lost works of art once believed to be by Vermeer. It also includes datasets on the provenance of these works, expanded citations for images, and information on the works of art sold at the Dissius Auction of 1696, a sale of the collection of Jacob Dissius, which included twenty-one paintings by Vermeer. When possible, we have included links to online museum catalogues with images of the relevant works of art. As Johanna Drucker points out in her article “Is there a ‘Digital’ Art History?” (2013), the creation of such a dataset is never unbiased. The images selected, the dates provided, the sources referenced — all privilege certain information. For this reason, we are eager to receive recommendations to improve the dataset. We hope users will contact dahlprograms@frick.org with additions, revisions, and new information so that this dataset can grow as a dynamic tool for researchers.
Demystifying Github
While the Montias database will continue to be available as a Filemaker database at research.frick.org, this new release will improve functionality for those looking to run data analysis research. With its release on GitHub, users will have new access to the data. The CSV documents can be opened in Microsoft Excel and used as a regular workbook. Users can create charts and graphs within Excel or a similar program. Click here to find an introduction to the dataset as well as recommendations for how to use it most effectively. Providing both CSV and JSON formats offers the benefit of more advanced data visualization with web tools as these formats can be plugged into data visualization tools like RAWgraphs, Palladio, and more. From other GitHub repositories, visitors can download plugins and tools that can be applied to these datasets.
Developing a Digital Art History Practice
As an emerging method of research for art historians, digital art history offers new methods — and new complications. Pamela Fletcher details some of these issues in her survey of the field, “Reflections on Digital Art History” (2015). While reviewing the three major branches of digital scholarship — literary analysis, spatial analysis, and network analysis — she concludes that digital art history should not occupy a separate niche in the field of art history. Rather, all art historians can learn basic tools to incorporate digital analysis into their research. The varied datasets included here will provide an excellent springboard for art historians to learn basic tools. As always, researchers can find many examples of existing projects and articles on the DAHL’s Zotero Library.
It is our hope that not just art historians but also economists and researchers in other fields will make use of these datasets. We look forward to seeing how people use the data and hope that it will inform more long-term scholarly research.
We would love to know how you use this data. Please tweet at us at @Frick_DAHL or send us an e-mail at dahlprograms@frick.org.
Stay tuned for an upcoming release of the Dictionary of Spanish Artists and an El Greco dataset.
Johannes Vermeer (1632–1675), Girl Interrupted at Her Music, ca. 1658–59, oil on canvas, The Frick Collection