Update August 2013
Visualization Tool for Poetic Elements 1750-1850
Laura Mandell presented Manesh Chaturverdi's poetry visualization tool at DH2012 in Hamburg; Luis Menses presented his work at DH2013 in Nebraska.
Update January 2012
Visualization Tool for Poetic Elements 1750-1850
Two new versions of the poetry visualization tool, one created by Manish Chaturverdi at Miami of Ohio for his MA Thesis, and one created by Luis Meneses, will be available shortly.
Update 9 September 2009:
Visualization Tool for Poetic Elements 1750-1850
Since the original "coming soon" posting (2006; see below), the PA has been working with Ira Greenberg and Jerry Gannod on developing this Poetry Visualization tool. Laura Mandell has given several talks on the topic, at MLA and Haverford's conference "Digital Archivalism and the Future of the Humanities," organized by Laura Mcgrane; the powerpoint for this last presentation is available for download here. You may also see and hear a slidecast on this topic.
The coded versions of John Keats's "Ode on a Grecian Urn" and schemas are NOW available with a TEI P5 schema and documentation.
Coming Soon: Visualization Tool for Poetic Elements 1750-1850
The process of information visualization demands a partnership between artist and data. The artist sets in motion the visualization, which, in turn, reveals truths the artist could not have known. (Martin Wattenberg, in John Maeda.)
Information Visualization: the use of interactive visual representations of abstract data to amplify cognition. . . . Visualization facilitates understanding of both large-scale and small-scale features of the data. It can be especially valuable in allowing the perception of patterns linking local features. . . . [Also,] visualization allows the perception of emergent properties that were not anticipated. . . . [Finally,] a visualization commonly reveals things not only about the data itself, but about the way it is collected. (Colin Ware)
About / Tech Specs / Research / Bibliography
About:
This web-based tool written
in Java will allow users to try out various hypotheses about poetry production
during this period. From a drop-down box listing
twenty items, users will be able to select as many elements as they wish
to "privilege" or
to view as part of their visual model. These elements will range
from metrical forms to semantics, publication venue to graphics on the
page, images, book boards, slipcases, etc. Additional information will describe in
detail what in each case is being selected or privileged. The elements
will apply to all the poetry in the database, again, carefully specified
in a general description and an accompanying searchable database -- that
is, the database will allow generating lists
of the poems, authors, poetry collections, editors, publishers, etc. that
occur for each selected element. A
3-D graphic and color will be pre-assigned to each element. Users will
be able to change style attributes.
Once selected and submitted, a visual model of the information will appear, spread out across the screen in 3-D (x, y, z coordinates) based on, for instance, chronological occurrence and places -- again, two elements selected by the user from approximately 10 choices, the third z-coordinate standing for percentage of total database items. Generated automatically with each model wil be a map-legend indicating what each shape means.
After generating the model, the user can decide to eliminate or add elements. Also, he or she will be able to zoom in on the model to look at the details of any given portion. Zooming into portions of the model will allow the user to see connections among elements in chronological detail. Most important, however, is that the user will be able to generate and compare multiple models, setting them side-by-side or in succession in a browser window. Eventually, we will allow users who are logged in to save specific states of the tool, to save the visual models they generate and use them in publications, online or print.
To give you a sense of what this tool might look like, the visual models it will generate, I have imported an award-winning visualization project available to see online:
(http://wwwcg.in.tum.de/Research/Projects/VisContest05):
Imagine objects here other than spheres, a legend designating what colors, size, and shape indicate, and a set of controls allowing you to zoom in as well as add and delete constraints. Imagine being able to generate one model with specific constraints and another to compare to it: the shape of the whole image will allow forming, modifying, discarding, and generating hypotheses about popular and canonical poetry.
My collaborator, digital artist Ira Greenberg, is developing a 3D visualization engine for the project. Here's an early protytype of a dynamic 3D node cluster built with the engine
The Oracle database operates through php commands. The Java visualization tool will be linked to it via extension, a php-java bridge. The list of elements from which users will be able to select will be encoded directly in the poetry and book document whenever completely factual and already allowable in TEI. Line meters, for instance, and elements of the book's material design are codable in TEI P4 (see also Syd Bauman and Terry Catapano). Some elements to be coded having to do with more specific literary analysis may not be supported by TEI: metadata specifying for instance, theme, mood, tone -- determined in accordance with fully specified rubrics available to all site visitors -- will be added to the database directly. Eventually, also, a php form for each poem appearing on its presentation page will allow users to submit metadata analyzing poems that will, pending approval by the editors, be added to the database. Semantic analysis of texts will not require hard coding: we will rely upon The Semantic Indexing Project undertaken at Middlebury College under the auspices of NITLE. NINES search engines already allow for word searches of all texts.
What kinds of generalizations can be made about differences in style and presentation between "high" or canonical and "low" or popular poetry?
What are the "publishing formulas" (Chartier) deployed by printers of poetry collections during this era? Do they institute associations between graphics -- not just engravings, but marks upon the printed page -- and themes or styles in various types of poetry? types of authors and graphics?
Do the stylistic devices we associate with specific themes in fact support those themes in a statistically significant way? What are the anomalies, and where are they published -- with what target public in mind? (That is, is high or popular art more idiosyncratic, or are they idiosyncratic in the same ways?)
Chartier, Roger. The Order of Books. Trans. Lydia G. Cochrane. Stanford, CA: Stanford UP, 1994.
Chen, Chaomei. Information Visualization: Beyond the Horizon. 2nd ed. New York: Springer, 2006.
Fayyad, Usama, Georges Grinstein, Andreas Wierse. Information Visualization in Data Mining and Knowledge Discovery. San Francsico: Mogran Kaufman, 2001.
Information Aesthetics. 3 May 2006. <http://www.infosthetics.com/>
Maeda, John. Creative Code: Aesthetics and Computation from the MIT Media Lab. London: Thames & Hudson, 2004.
Manovich, Lev. "Info-Aesthetics." New Media / Culture / Software. 3 May 2006. <http://www.manovich.net/>
Mitchell, William J., Alan S. Inouye, Marjory S. Blumenthal, eds. Beyond Productivity: Information Technology, Innovation, and Creativity. 22 May 2006. <http://newton.nap.edu/html/beyond_productivity/>
Tufte, Edward. Envisioning Information. Cheshire, CT: Graphics Press, 1990.
---. The Visual Display of Quantitative Information. 2nd ed. Cheshire, CT: Graphics Press, 2001.
Ware, Colin. Information Visualization: Perception for Design. San Francisco: Morgan Kaufman, 2004.