Matt Kilbane, a Ph.D. candidate in Cornell's English Program, was selected to participate in the Summer Graduate Fellowship in Digital Humanities in 2016. Through the course of the fellowship, Kilbane developed a set of experiments to compare and measure modern poetry and presents the findings in a public website.
Project Title: Computational Experiments in Modernist Poetic Diction
Project URL: http://matthewkilbane.com/cempd/
With the support of the 2016 Cornell Summer Graduate Fellowship in Digital Humanities, I began a series of “experiments” applying computational tools to the study of poetic diction, particularly that of English-language modernism. These experiments include attempts to calculate and compare vocabulary “richness” across a series of twentieth-century long poems, to measure the distribution of parts-of-speech in canonical modernist works, and to discern in Langston Hughes’s blues poems the linguistic traces of contemporary blues practice. When I’m optimistic about this work I imagine a kind of hybrid methodology that would pursue the reciprocal and dynamic relationship between the particulars of any poetic language and its larger social and historical grounds, its grounds in speech. Wallace Stevens, in his Notes Toward a Supreme Fiction (1942), describes this relationship: “The poem goes from the poet’s gibberish to / The gibberish of the vulgate and back again.” If we agree with Stevens, then we need a scholarly approach that is founded on the close reading of “the poet’s gibberish,” but that is willing and eager nevertheless to zoom out, with a machine’s help, and access something of the “gibberish of the vulgate,” in order to return “back again” to the poet’s language with the profit of this new view in tow. That’s the methodological horizon toward which I’m aiming in these experiments, for which I'd be ill equipped without the generous assistance and guidance of the summer fellowship program.