Contributors to this project included a partnership between the Emory Center for Digital Scholarship and the Raymond Danowski Poetry Library at Emory University. Listed on the project’s website are technical lead Rebecca Sutton Koeser and researcher Aaron Goldsman. Rumor has it there was also a digital researcher Brian Croxall and his team on the project for a portion of the time, although organizational problems caused some team members to leave the project before it was completed.
The purpose of Networking the New American Poetry research was to test the taxonomy of Allen’s postwar schools of poetry, outlined in his anthology The New American Poetry, 1945-1960 and how they actually fit into the publishing community of the time period. Mainly, they wanted to examine if the networks of publishing, editorial work, and collaboration enumerated by Allen’s work truly matched the archival evidence or works of poetry published in those decades. According to Scheinfeldt’s theory, the project did have a central argument, although at the conclusion project members also wanted to open it up for more scholarly research and for other people to come in and ask questions. One useful purpose for this project is to enhance existing library catalogs but the main audience of this project is other researchers and scholars.
The main sources used for “processing”, as Posner would describe it, were twelve avant-garde literary magazines related to the main poets described in Allen’s anthology. They included Beatitude, Big Table, C: A Journal of Poetry, Evergreen Review, Fuck You: A Magazine of the Arts, Intrepid, J: A Magazine of Poetry, Measure, Origin, The Black Mountain Review, The Floating Bear: A Newsletter, and Yugen.
The data points were processed by the researchers who entered these categorizations into their online database by hand. Over 10.000 pieces of data were included. The network graphs displaying these datapoints were built by a program called zurnatikl, a Django web application developed by Emory University software designers. The network graphs are generated by the zurnatikl software that runs the website and are also the basis for the network visualizations displayed on the site. There are three different types of network graphs available, which represent both individual and overlapping content. One links the editors and poets connected by Allen in his book, another links all of the data points collected by the researchers in one large graph, and the third mode shows the specific connections of one artist, editor, or translator from these magazines.
These data visualizations are published as an offshoot of Emory’s university website. The website and the graphs are run by zurnatikl, developed by Emory technicians specifically for the visual connection of journal data. The researchers came to several conclusions about the project such as, “The shape of the communities that we see in the contributor graph suggests a network both less regimented and more capacious than Allen’s anthology indicates. In part, Allen has simply done the work of the anthologist, pruning and aggregating according to taste and other motivations” or “When we compare Allen’s categories to our graph, we can see that Allen’s poetic schools, which have come to define conversations about mid-century American poetry, were in practice quite permeable…”
This project enables similar scholarly primitives as the first, such as discovery, comparison, and representation. It has many strengths, especially in the data visualizations which are both aesthetically pleasing and easy to use. One could argue that the data included is also not exhaustive, and the singular graph of all of the data connections is harder to decipher because of the large amount of data. A tool to help narrow search on that graph could make it even more user-friendly in the future.
