This talk explores some of the possibilities and limits presented by social network analysis for quantifying and visualizing histories of literary influence and affiliation. In the Japanese context, one of the most promising areas for experimenting with SNA is in the field of modern poetry—a field defined by the continuous grouping and regrouping of poets through small-scale journals typically organized around mutual acquaintance and common aesthetic purpose. Here, I work from an index of over 100,000 poems published by more than 4,000 poets between 1920 and 1944 in 166 different journals. Specifically, I map multiple layers of affiliation between two distinct factions in this dataset, one associated with the anarchist, provincially-based journal Gakko (1929-1930), and the other with the avant-garde, Tokyo-based journal Shi to shiron (1928-1931).
On the one hand, the resulting visualizations offer compelling images of how these factions evolved over the entire span of the index, exposing allegiances and fault lines intuitively known to scholars but which remain invisible when dealing only with discrete texts. On the other, the processes of abstraction that allow these images to be constructed raise difficult methodological questions: How does affiliation at the graphic level correspond with affiliation as understood at the historical and discursive levels? What information gets left out in the process of constructing our data, and what interpretive assumptions underlie this process? Finally, how do we design network visualizations in such a way as to encourage their productive dialogue with the literary archive rather than forestall interpretation at the image itself?
Copyright. All Rights Reserved.