Walt Whitman's "Song of Myself" (1892 Edition)

An Object Ecology Visualization by Zach Horton and his students

 

click on object categories to zoom in (performance is best in Chrome on a laptop or desktop)

Walt Whitman's "Song of Myself" is an ecstatic affirmation of difference within unity. Its efficacy for any reader lies in its ability to navigate this knife's edge between a cosmic perspective and the singularly intimate, the great whole and the miniscule detail. Whitman's strategy is to build, over the course of the poem, a speculative ecology of objects, each affirmed in its uniqueness as well as its place in the whole. The result is a poem that evokes over 2000 encounterable objects: people, places, animals, and other things (including speech and concepts). This Digital Humanities study of "Song of Myself" seeks to explore its rich ecology of objects in ways not accessible to standard reading practices.

The interactive "sunburst" visualization above presents the poem's object ecology in hierarchical, categorical form, eventually drilling down to individual objects as they are enumerated in the poem. Readers interesting in exploring the classes of objects and their relative frequency in the poem as a whole are encouraged to spend some real time with this tool.

The image below is a time-series plot that visualizes the distribution of high-level classes of objects throughout the poem. The horizontal axis is the poem's progression from first to final section (Whitman's final version of the poem includes 52 numbered sections). For more fine-grained distributions and trend relationships between different sorts of objects as the poem progresses, see more distribution plots here. You can also view the full marked-up text of the poem here.

All data was manually encoded, through a Herculean effort, by Sue Yang, Haley Wilson, Lizette Alejandre, Jordan Cherkes, Sahana Arkalgud, Erin Cantrell, Amanda Small, Alison Iwashita, Benjamin Lee, Joshua Gomez-Zavala, Daren Spears, Melanie Searles, Stephen Read, Miranda O'Mahony, Claire Breen, Paige Yamron, Kore Busath-Haedt, Laura Montalto, Alexandra Pingree, Jasmin Rios, and Grace Nevins. Yaojian Shen and Jackie Zintel did a great deal of the JSON file wrangling. The "zoomable sunburst" visualization above is based upon examples by Mike Bostock and Martin Etmajer. It was made using the D3 javascript library.

I welcome your thoughts, suggestions, and collaborations to expand this study! Please email me at zhorton [at] english.ucsb.edu.