Introduction
On a recent trip to London, I made my way to the Tate Modern on my first full day in the city to see the museum’s Cezanne installation and an exhibit of Australian artistry. I was upset to not be able to visit the Tate Britain on my trip, but thankfully this midterm gave me the ability to explore the museum from afar. For my project, I have chosen to map the hometowns of Italian artists born between 1930 and 1949 who currently have work on display in the Tate. I have long had a fascination with Italian language and culture, and have a personal goal for myself to learn Italian as my next language, so it only seemed proper to explore this country over the other well-represented countries in the data set (i.e., UK, USA, New Zealand, France, Northern Africa, etc.). My aims were to reveal a spread of where the Tate’s Italian works are coming from, and form a hypothesis as to why some of the country’s more famous cities are seemingly underrepresented.
Sources
The data set used to create the above map consists of 629 artists who have work on display in the Tate, all of whom were born between 1930 and 1949. The set includes data on each artist’s name and gender, as well as their Tate-specific ID, their year of birth and year of death (if noted), their place of birth and place of death (also if noted), and a link to their bio on the Tate’s website. I knew when I began looking at possible data sets that I wanted to use one of the two Tate sets (artists or artworks), but there was such a wide array of people and works to choose from that I was quickly overwhelmed by it all. I figured I needed to pair it all down somehow and focus on one specific aspect to visualize.
Processes
After deciding I wanted to focus on Italian artists and their birthplaces, I imported my data into OpenRefine and performed some minor data cleaning to make everything uniform and organized. I then applied a Text Filter to the Place of Birth column and filtered out any rows not containing the word “italy.” This left me with 18 matching rows: 18 Italian artists, 16 of which were male and 2 female, and 3 of the the 18 being from the same hometown (Torino, in NW Italy). With these 18 data points, I exported the custom set as a CSV file and added it as a layer in a new ArcGIS map.
Presentation
Using ArcGIS’s internal tools, I was able to manipulate my map in a number of ways. I experimented with creating a heat map to show where the highest density of artists was, but there weren’t enough units to create differentiation in the map’s heat zones. I also considered creating some relationship between the different points, but realized there wasn’t enough similar/differing data to create relationships that weren’t already obvious. In the end, I started my customization with a map full of data points, similar to the one you see above, and decided it would be best to let them stand on their own. I increased the size of the Torino point to 3x the others because there are 3 artists from there, and only 1 from everywhere else. I also changed the colors of Northern Italian points to red, and all Central and Southern Italian points to green, because there were less of them and they were less clustered than the Northern points. I decided to remove any labels, because some points were too close to each other which caused for overlap between some labels, and I edited the pop-up box for each to display only the data I deemed significant to the visualization. I edited my base-map again to make sure it reflected my aim, added an interactive legend and drop shadows to each selected point, and published it as an Instant App to embed and share here.
Significance
When a museum chooses to represent an artists by displaying their work, they are making a conscious decision to use art from a specific place, created at a specific time, and by a specific person. Knowing what I already know about Italian art, I was surprised that my data set created a visualization that leaned heavily towards Northern Italy, and was confused as to why a major museum like the Tate would mainly be representing these lesser-known regions. With only 1 Roman artist, no Sicilians, no one from Naples, and even no Venetian artists (yes, Northern), I was taken aback and figured “there’s no way these are the only Italian artists on display here. It’s impossible.” I then remembered an entire aspect of this assignment is to choose a portion of data to present digitally. It also occurred to me in that moment that the original data set I was using was also only a part of a larger whole, and that there were certainly loads of Italian artists in the Tate who were not included in the data set, because they were born either before 1930 or after 1949. My guess is, were I to map a complete list of Italian artists in the Tate, not only would there be much more representation in Central and Southern Italy, but there would also likely be enough data to create an effective heat map like I’d intended. What I have so far taken away from the field of Digital Art & Humanities is that no matter how large your data may be, you usually end up only visualizing a part of that data, and the choices you make along the way all add up to the art you produce on the other side. Data science, on the other hand, is more about collecting and analyzing data holistically, and less about how one chooses to present those analyses.
