The Datathon Talk was an idea of Diana Greenwald for the National Gallery of Art (NGA). Which purpose does it serve? The NGA has for mission collecting Art. Nevertheless, compared to other museums, the NGA has a pretty young collection, started in 1941. Of course, the collection has grown, notably thanks to gifts from donors. According to Greenwald, all of the aforementioned lead to new possibilities and especially the use of data science in order to understand the NGA collection. Data scientists and historians can gather around three central criteria: analyze, centralize, and visualize.
When Greenwald looked at the NGA collection, she wondered “Does this export to a spreadsheet?”. The answer is yes. In Art history, there are simple bias. For example, the idea that “tall means healthy” was a belief in the old times but it is not representative. There can also be a disproportionate attention to some artists or medias. The Datathon teams decided to combine collection data with methods analysis. As a matter of fact, questions emerged mainly around diversity and inclusiveness.
The teams participating to the Datathon Talk were formed by professors, historians, data scientists, but also graduate and undergraduate students. Each team had a subject assigned.
Team 1 talked about “Diversity on display: who is on the wall at the NGA?”. It shows that most objects in the collection are from male and white artists. In addition, depends on the building (east or west), the objects are from different group of people, ethnicity, and gender. The representation of women becomes concentrated in some rooms of the museums as well. The team also noticed that when doing data research, they encountered difficulties with transparency, accuracy, and availability. For example, the painting “Girl on Globe 2”, is not linked to any location, the data is not available. Data mining can be challenging.
Team 2 talked about “Influence of women as donor in the Collection”. Surprisingly, there were many women, some of them famous, who influenced the collection. For example, Julia J. Noorell, a collector in the South, Kathan Brown, and Ailsa Mellon Bruce who donated 182 paintings. When the NGA received consequent donations, it reshaped the NGA’s perspective, sense, and aesthetic.
Team 3 talked about “Acquisition History”. The team noticed that volume of art created by men stabilized over the years, but women’s work is increasing. They also observed that American art is in expansion since the 1980s. They studied the proportion of works by gender and by media type in the entire collection. For example, most female representations are in photography and prints. The different NGA directors, for a total of five, also influenced the direction of the collection.
Team 4 talked about “Display versus Holdings”. Indeed, they ask the question “what should go on the NGA’s walls?”. The biggest criteria for decision making is the artist popularity. This is something we can observe on Wikipedia pages for example and see how many times they were visited. For example, Van Gogh, Da Vinci, and Neil Armstrong. It is particularly interesting to notice that 97% of the page view correspond to 20% of the artist. There is a huge concentration of interest on a few famous artists.
Team 5 focused their research on a software called “Inception V3 Neural Network”. The idea is to see abstract art to be representative of life. They gathered thousands of pictures together in a visual goal. The result is surprising, and many interpretations are possible.
Team 6 analyzed spatial evidence. There are different scales of space. For example, widener Artworks Exhibition locations over time, relationships, building level, room level.
What should be the conclusion of this Datathon Talk? According to Greenwald, we should remember that the NGA evolves depends on taste and institutional priorities. Data research brought up questions related to curatorial work. Is it aesthetically pleasing? Is it worth collecting? Curatorial work is socially bounded and changes overtime. Data research should be playing a main role in understanding and reshaping the NGA or even other collections of museums. Data research can bring a new dimension to the curatorial work through the use of diverse and useful data tools.
