Networked NY Q&A: Micki McGee

This week, Micki McGee has graciously agreed to participate in our Networked New York follow-up Q&A series. For more on the Yaddo Circles Project, click here.

1. In “Creative Power: Yaddo and the Making of American Culture,” you write, “Although an artist’s work may seem to be authored alone, an artwork is a social activity, situated in a dialogue with the conventions that preceded it and, if widely enough disseminated, the artworks that follow.” Is this true of the Yaddo Circles project as well? Moving forward with the archive, how do you think other sites–both contemporary social networking tools like Twitter and Facebook and other digital archives like The Crowded Page–will inform your work?

Unequivocally the Yaddo Circles network-mapping project is situated in a dialogue with other such projects. The idea for the Yaddo Circles mapping project emerged in 2007. We wanted an interactive online media component for an exhibition I was developing for the New York Public Library, Yaddo: Making American Culture. I’d seen some of the work that others had done with mapping intellectual communities. Katy Börner’s beautiful information visualization projects with large citations networks were one source of inspiration. Another was the work on Valdis Krebs, who had been making his network mapping work transparent and accessible. Both Börner’s and Krebs’s teams were working with desktop applications that did not run live online (as were so many sociologists engaged in social network analysis), so the on-screen and print outputs were fantastic, but they were not exactly suitable for an interactive online tool that we wanted to build to accompany the exhibition.

As I was working in the Yaddo Records, Josh Greenberg came on board at the Library and became an enthusiastic proponent of working on developing this online tool. Around that time I also came across the newly developed Visual Thesaurus by Thinkmap.com and thought – “Wow, this is exactly what we want, only for mapping relationships of persons and institutions rather than the relationships between words.” Unfortunately, the Thinkmap software was not open source and was well outside of our budgets, so we tried to build our own mapping tool on the fly out of the thin margins of our exhibition and departmental budgets. This is not an approach that any of us would now recommend!

After the traditional brick and mortar exhibition was up and running, some of us circled back around to look for support to build our digital mapping tool. Jennifer Serventi and Jason Rhody, Program Officers at the National Endowment for the Humanities, introduced us to the work that Andrew Jewell (University of Nebraska, Lincoln) and Edward Whitley (Lehigh University), whom the Endowment’s Office of Digital Humanities had just funded for work to develop The Crowded Page. At that point if was clear that several of us were all dreaming the same dream: of being able to map the relationships of artists and writers to see the influences of one upon another. Other folks were doing this with the Muckety website, that maps relationships of political influence, but also allows you to upload your own data sets to their network mapping tool. Similarly, the computer scientists at IBM’s Many Eyes labs were making an online network visualization tool. And Jeffrey Heer and danah boyd had developed Vizster, for visualizing networks online. Heer’s beautiful open source Protovis and D3 software is part of the nuts and bolts of the tree-graph visualization that Aditi Muralidharan, Asik Pradhan, and Charles Forcey developed for Yaddo Circles. So many teams have been working on these network visualization tools. The magnificent Republic of Letters project from Stanford University is another example of this sort of mapping effort, though it maps the circulation of letters (and therefore a social network) over a geographical map, affording an understanding of both social and geographical spaces.

Another important parallel development to this rise of network information visualization has been the movement to harness the power of the semantic web to render network visualizations. In the world of archival science, Daniel Pitti from the University of Virginia was working with a number of people on the standards for archival data developing Encoded Archival Context-Corporate Bodies, Persons, and Families (EAC-CPF) guidelines that are beginning to make it possible to capture data from archival finding aids and render network maps from that information. His prototype project Social Network for Archival Contexts (SNAC) is really a model for the field. At the same time, other parts of the cultural heritage sector coalesced around the idea of developing linked data standards for libraries, museums, and archives when Jon Voss (now at History Pin, formerly at LookBackMaps) and Kris Carpenter Negulescu (Internet Archive), hosted an international meeting on Linked Open Data in Libraries, Museums, and Archives (LODLAM). The conversation that they ignited has accelerated the rate of discussion for appropriate data structure and capture that would allow us to leverage the power of the semantic web to conduct this mapping of ideas and influences on a massive scale. The Compatible Data Initiative that I am working on with Börner, Pitti, Whitley, as well as with Richard Edwards at Ball State University, is part of this effort. Cristina Pattuelli’s Linked Jazz project is another example of a network-mapping project that harnesses the power of linked data to create network maps of cultural communities, as is David R. Morrow and Chris Alen Sula’s Phylo project that looks at the networks of twentieth century philosophers.

Now there is also the Person Data Repository in Germany at the Berlin Brandenburg Academy of Sciences and Humanities, and the Linking Lives project out of the UK. Lots of people have been dreaming the same dream and the technologies to realize this dream are now becoming more widely available. It is mind-boggling to imagine what will be possible in terms of humanities research on intellectual communities within the next five years, and simply thrilling to be involved in these efforts.

Well, that was a very long answer. I promise to be more concise on your other questions!

2. Returning again to “Creative Power,” you write, “The fabric of a community is woven by…modest moments of admitting one and rejecting another, in the reinvitation or refusal of the next. An institution, although conceived in its papers of incorporation, is made and remade in the minutiae of such actions.” How does the team behind your site represent Yaddo’s exclusivity and make their own decisions about whom to include?

Our intention with the Yaddo Circles project is to make available all the data we have captured about the community and its admissions process in our network visualization. At this point our prototype models only the first, second, and third degree relationships of persons associated with the group in the 1942 summer photograph (which includes about 5,000 total records). This decision was made so that we could develop a working prototype and then begin to work out the issues of scaling that up to include the 6,000 personal records and more than 12,000 corporate records in our data set.

3. How, and by whom, do you envision the site being used? How do you think using a controlled vocabulary to describe relationships at Yaddo will inform the work researchers produce?

These seem like very different questions to me, so I’m going to answer them separately. We envision two complementary outcomes for the Yaddo Circles project. On one hand, we imagine the project serving as a tool for discovery, allowing anyone with internet access to explore the relationships between the artists, writers, and composers affiliated with Yaddo and the impact of this group on 20th century arts and letters. And on the other hand we imagine building a data-agnostic tool that would allow researchers to capture person and entity-centric data from any archival source, map it for their own purposes, and when they are done with their own scholarly projects, release their materials as structured open data with provenance information, allowing people to almost incidentally begin crowdsourcing the work of indexing archives. Several people are also dreaming the data-agnostic back-end dream. Among the most promising developments is the work being done by Alan Liu has put together at UC Santa Barbara on a project called RoSE, a Research-oriented Social Environment. We want what we are developing on the Yaddo Circles project to resonate with these initiatives so that we can contribute to advancing these efforts nationwide.

With respect to the use of controlled vocabularies for describing the relationships between people represented in the Yaddo Records, the Compatible Data Initiative group and my Yaddo Circles research team at Fordham will be experimenting with this in the next few months, and consulting with archival experts like Cristina Patuelli and Daniel Pitti to explore how our work at the level of the archival file might be captured in such a way as to allow for individualized, highly-granulated, folksomic tagging of relationships as well as for a controlled vocabulary that can be interoperable with projects such as SNAC, that work at the level of the collection and its finding aid (or even at the level of the card catalog, if one moves out of the archive and into the library more broadly).