Creating Digital Versions of Walt Whitman's Marginal Annotations
Travis Brown, Project Manager (Univ. of Texas at Austin)
Nicole Gray, Editorial Assistant (Univ. of Texas at Austin)
Erica Fretwell, Assistant Editor (Duke U.)
Kevin Webb, Consultant (Tackle Design)
Brian Pytlik Zillig, Consultant (U. of Nebraska-Lincoln)
Brett Barney, Consultant (U. of Nebraska-Lincoln)
Duke University Libraries
Research Assistants:
Kristen R. Davis (Duke U)
Anna Dudenhoeffer (Duke U)
Melissa J. Miller (Duke U)
Start date: September 2007
Projected publication date: ongoing
Cohen Lab received a Digital Humanities Start-Up Grant in 2007-2008 from the National Endowment for the Humanities to develop software tools and protocols for digital representation of static multimedia documents. Using Walt Whitman's manuscript marginalia--his annotations and other scribblings on other writers' printed works--we have built prototype tools for marking up such documents as well as for displaying interactive search results for such documents using images and text.
The project unfolded in several stages. First, with the help of Duke University Libraries, we scanned and transcribed some of the most interesting documents in Duke's Trent Collection of Whitman materials. Whitman used pasted-on clippings, drawings, mathematics, annotations, and underlining in pencil and ink as he interacted with printed materials and even photographs and drawings. Research assistants transcribed all such information and generated descriptions of non-textual content. Then, using the Whitman Archive's Poetry Encoding Guidelines as a starting point, we began to encode these transcriptions. Brett Barney of the University of Nebraska-Lincoln's E-Text Center came to Duke to help us create Encoding Guidelines for these static multimedia documents that extended the extant Whitman Archive DTD. These documents and standardized image scans of the originals were passed on to Tackle Design, a Durham, NC software design firm, for use in generating two tools: one to help encoders link images to XML files through a coordinate system that spatially locates each manuscript word or image element in a document; and one to help users search for such content online. It's a kind of "Google Earth" for documents that mix print, manuscript, and drawings or other visual elements.
The next phase of this project will involve creating a demonstration site and creating connections with the archives that hold large or particularly rich collections of Whitman's annotated material, including among others Duke University's Special Collections, the Harry Ransom Center at the University of Texas at Austin, the Library of Congress, and Middlebury College Special Collections.
A white paper by Erica Fretwell, Matt Cohen, and Kevin Webb on the project, published at the NEH web site for digital humanities projects, can be downloaded here:
NEH_White_Paper.pdf
Demonstrations of the coordinate capture interface, and a video showing how they function, can be found here:
http://www.structuralknowledge.com/markup_demo/goethe/
http://www.structuralknowledge.com/markup_demo/dryden/
http://www.structuralknowledge.com/markup_demo/
Our preliminary encoding strategies--extensions and tweakings of the general Whitman Archive Encoding Guidelines--may be found here:
Annotations_Guidelines_Draft.pdf
Demonstrations of the search interface and the text-based interface will be available soon. The text-based interface, which will use ImageMagick to generate interactive layers of textualized versions of the document scans, is being built by Brian Pytlik Zillig at the University of Nebraska.
Conference presentations on the project may be downloaded here:
STS_09_TalkInterfaceDesign.pdf