Here’s a different way to think of museums as being social:
This graph shows the connections between works of art and users of the Walters Museum of Art online collection. The Walters website has a Community Collections feature where anyone can make and share their own selection of objects from what’s available. On the chart, the blue nodes are individual community collections connected by the lines to artworks (green nodes). I’ve included only artworks that occur in two or more collections, so the collections are connected to each other through the objects they share (and, inversely, the object are connected through the collections that share them)
I’ve long been skeptical about personal “collections” on museum websites. They’ve always seemed like dead ends to me—useful, perhaps, for some classroom activities but not much else. I’m still skeptical of the concept as we often see it, but as the website becomes less of a destination and more of a platform for other kinds of interaction I’m becoming more interested by the trails that those interactions leave behind. “My collection” may be a dead end if it’s just the result of picking objects out of a collections database interface, but if it’s the record of things I’ve tweeted, or tagged in a mobile app or used to create a mashup or something else more concrete as part of some activity then it’s starting to tell us more about how real people use a museum.
The Walters’ “Community Collections” are still just objects picked out of a database, but I like their redesigned online collection and they’ve made it very easy to share things that catch your interest. As I was reviewing it as an example for my students—and having just assigned Nicholas Christakis’ TED Talk "The hidden influence of social networks" the previous week—I began to wonder what the network of these community collections would look like.
Behold, the graph.
There is one thing that this graph can tell us right awy about how this community, acting as a community, sees this collection. The yellow node in the middle of the graph is the artwork with the highest "degree centrality". It is, in a way, the most popular work of art. It has the most and best connections, and if there were a cold or flu running through the collection, this work would be sure to catch it. It is Ingres’ Odalisque with Slave:
The next 4 most central objects are:
- Leon Jean Basile Perrault, Maternity
- Lawrence Alma-Tadema, Sappho and Alcaeus
- Erastus Palmer, First Disappoinment
- Asahi Meido, Courtesan with Brush and Scroll
What interests me about this list is that no one was asked to make it, it simply emerged from an activity with no particular goal. If you asked people to vote on “the most important work in the collection” I’m sure you would get a completely different answer. Of course, “central” on this graph doesn’t mean “important” in any conventional sense. What then does it mean? I don’t know—I just made the graph for fun—but if we can invite more meaningful interactions with collections and watch the connections and paths that are drawn along the way, I hope that it will become another area of curatorial interest, or community interest for that matter. Whatever it means, the position of Odalisque with Slave in this graph is not really a fact about the painting but as much as its a fact about the painting as a element in a human activity. At a time when we’re often wondering why people should, would or do go to museums and why museums should be funded those are important facts.
Some technical notes: The data was collected on October 23, 2011 by simply screen-scraping. The graph was made with NetworkX and if anyone knows how to get NetworkX to spread the nodes out to make a less cramped graph, please let me know!
Why the Walters? It’s just what I was looking at (because I think they’ve done a good job on their redesign), I could figure out pretty quickly how to get the data (an API would’ve been nice!) and it was a manageable amount.