VizOR: Visualizing Only Revolutions, Visualizing Textual Analysis

Introduction
This paper describes VizOR, a new digital resource currently under development, that visualizes Mark Z. Danielewski’s 2006 novel Only Revolutions. VizOR is built on top of a MySQL database comprised of the complete text of Only Revolutions and is programmed in Python to produce a dynamic, database-driven visualization of the novel. In this paper, we discuss the methods and procedures we are currently developing to create this visualization and highlight the implications of these methods and procedures for theoretical concerns in both digital humanities and media studies. This project, we propose, is both a re-reading of and an exploration of the process of reading Danielewski’s novel: as such, it joins the novel in examining the interrelation of human and machine “reading” and “authorship,” pointing to a procedural understanding of reading and writing and suggesting that both activities occur across a wide variety of actors and platforms.
Context
The form of Danielewski’s novel is unique: the narrative consists of two parallel yet interrelated narratives, one by Hailey and one by Sam, and which one the reader reads depends on how the reader holds the book. If the reader is reading Hailey’s narrative, for instance, she must flip the book upside down to read Sam’s (and vice versa). Apart from the main narrative, each page also contains what we call a chronology section — a date with a list of historical people and events associated with that time or date, which Danielewski crowdsourced from his fans while writing the novel. Thus, each page is divided into four sections: Sam’s narrative, Sam’s chronology (which runs from November 22, 1863 to November 22, 1963), Hailey’s narrative, and Hailey’s chronology (which runs from November 22, 1963 to June 19, 2063).
In addition to its formal innovations, Only Revolutions is interesting for this project because it proliferates numbers, playing with the boundaries between “data” and “narrative.” For example, the numbers 360 and 8 and their factors and multiples are particularly important. Each narrative is 360 pages long, and each narrative and chronology section on each page contains 90 words (90 X 4 = 360). Furthermore, the narrative is divided up into sections of eight pages each, and the number of lines in each narrative section decreases, at regular intervals, from 22 to 14 as the reader progresses through these sections (22 - 14 = 8). “H,” for Hailey, is the eighth letter in the alphabet, and “S,” for Sam, is the eighth letter from the end of the alphabet; “Mark Z. Danielewski” has 16 letters; and Sam and Hailey are described as being “allways sixteen.” There are many, many more examples we could cite here. In this way, Only Revolutions encourages readers — human and machine alike — to count, and count again, the many different numbers that emerge from its pages.
In How We Think: Digital Media and Contemporary Technogenesis, N. Katherine Hayles includes a coda featuring visualizations of Only Revolutions produced with Google Maps. Collaboratively designed with Allen Beye Riddell, these three visualizations trace the geographical “place-names” of Sam and Hailey as they travel throughout the story, and then layer the two to create a composite map of the characters’ movement through the text (243-244). The resulting visualization shows that Hailey and Sam take similar paths across the map and that their overall directionality is nothing if not inconsistent. They move on a whim, together, wherever their overwhelming affection for one another points them. On one hand, VizOR is a response to Hayles’s map-based visualization and is concerned with the assumptions underpinning her choice of visualization platform. On the other hand, the project is addressing a larger trend within the digital humanities: the increased prevalence of data visualization as a mode of literary interpretation.
Due in large part to its often powerful and aesthetically pleasing visual impact, relatively quick learning curve, and overall “cool,” the practice of visualizing textual data has been widely adopted by the digital humanities. This prevalence is evidenced by, for instance, the high frequency of the term “data visualization” in the 2011-2012 Digital Humanities conference abstracts as well as the 2011-2013 Modern Language Association digital humanities panel descriptions. If the first wave of large-scale database projects in the digital humanities is exemplified by the practices of digitizing texts, constructing archives, and determining best practices for digital preservation, then the practice of data visualization is emblematic of the second wave of projects devoted to mining and interpreting this newly available data.
VizOR is influenced by the thinking of scholars and practitioners like Franco Moretti, Matthew Jockers, Jeremy Douglass, and Lev Manovich, as well as by visualization projects like UC San Diego’s “Cultural Analytics” initiative and Stanford’s “Gephi” visualization engine. Like other critical DH projects, VizOR is not only interested in engaging with literature via data visualization, but also in performing this engagement to ask what is at stake in this new mode of interpretation, both in terms of the individual scholar and the digital humanities as a field.
Project Description
The database is designed to be highly flexible, and this architecture allows us to enact the same linguistic layering that occurs while reading the text. We can query a specific word of a particular character’s narrative or chronology, the text from a specific line of a character’s narrative or chronology on a specific page, or the narrative or chronology text from a whole page for a specific character. Querying the database in this way allows us to instantly compare the words and phrases used by one character with those used by the other character on the same page and in the same position. For example, on Sam’s page seven, line five, we see “Gold Eyes with flecks of Green,” while the words in the same position on Hailey’s page seven, line five read,“Green Eyes with flecks of Gold” (Danielewski). VizOR surfaces this kind of reading through its very design, allowing readers to see the similarities and, perhaps more importantly, the differences between the narratives.
The visualization is currently still under development using the Python programming language. Python was selected as the programming language for VizOR due to its general flexibility and its ability to produce dynamic, database-driven visualizations. However, static mock-ups have been designed using Adobe Illustrator and are included at the end of this document.
The visualization reproduces the image of the page numbers in the novel, the two small rotating circles enclosed within the larger circle. In the visualization, however, each of the two smaller circles represents the narratives of Hailey or Sam. All three circles are comprised of absent centers with the text literally expanding outward. The larger circle is comprised of the chronological headings and entries. Clicking on one of the text strings in either narrative will query the database for the corresponding narrative’s string. This correspondence, as well as the inclusion of the Boolean double pipe (symbol for “or”) as link, is shown in the mock-ups at the end of the document. This, in effect, mirrors the layering that occurs during the act of reading. The visualization, though, has the ability to position the corresponding lines on the same plane at the same time, a phenomenon in the novel that is always already delayed or fading away. Rather than attempting to flip the book fast enough to see both sides of the coin, so to speak, VizOR attempts to freeze the text at each moment of mirroring. This moment of pause opens up the possibility of interpretation without disrupting the line’s relationship to other lines or removing it from its context.
Users can navigate the visualization in a number of ways, hovering over any of the terms to magnify chosen words or lines. Once users click on a given line, the circles will rotate to realign the corresponding terms. Further, rotating the encircling chronology forward or backward in time will result in the rotation of the two inner narratives, mirroring the motion of the flip-book page numbers of the print text.Users can navigate the visualization in a number of ways, hovering over any of the terms to magnify chosen words or lines. Once users click on a given line, the circles will rotate to realign the corresponding terms. Further, rotating the encircling chronology forward or backward in time will result in the rotation of the two inner narratives, mirroring the motion of the flip-book page numbers of the print text.
The finished form of the visualization will be searchable and will also contain external hyperlinks. In keeping with the novel’s data-driven construction, produced in some ways by a crowd-sourced or “collaborative” author, the visualization produces a similarly data-driven reading experience. The goal here is to mirror the outward push of the novel, its awareness and incorporation of external databases like online encyclopedias, and the uniquely distributed reading experience of excitedly setting the book down to search for one of its vague historical entries.
Conclusion
In the name of this speed and efficiency, however, these technologies often strip data of its context and idiosyncracies, creating what Tara McPherson has called “a system of interchangeable equivalencies” (35). VizOR, through its emphasis on the distributed processes of reading and writing across different technologies and media, pushes against this seamless homogenization. By highlighting particular, idiosyncratic moments of reading, we hope to activate the messy, "seamy" place where data meets narrative.
References
Barthes, R. (1977). Rhetoric of the Image. Image, Music, Text. New York: Hill and Wang.
Borner, K. (2003). Visualizing Knowledge Domains. Annual Review of Information Science & Technology. 37. Medford, NJ: Information Today, Inc./American Society for Information Science and Technology. 179-255.
Cartwright, L. (2009). Practices of Looking: An Introduction to Visual Culture. Oxford: Oxford University Press.
Danielewski, M. (2006). Only Revolutions. New York: Pantheon.
Drucker, J. (2008). The Virtual Codex from Page Space to E-space. In Schreibman, S., and R. Siemens, (eds). A Companion to Digital Literary Studies. Oxford: Blackwell. http://www.digitalhumanities.org/companionDLS/ (accessed September 21, 2012.
Hayles, N. K. (2012). How We Think: Digital Media and Contemporary Technogenesis. Chicago: The University of Chicago Press.
Lima, M. (2011). Visual Complexity: Mapping Patterns of Information. New York: Princeton Architectural Press.
Manovich, L. and J. Douglass (2007). Cultural Analytics. UC San Diego Software Studies Initiative.http://lab.softwarestudies.com/2008/09/cultural-analytics.html (accessed: September 21, 2012.)
McPherson, T. (2012). U.S. Operating Systems at Mid-Century: The Intertwining of Race and UNIX. In Nakamura, L. and Chow-White, P. A., (eds). Race and the Internet. New York: Routledge.
Terras, M. (2008). Digital Images for the Information Professional. London: Ashgate.
Tufte, E. (1983). The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press.
Vesna, V. (2007). Database Aesthetics: Art in the Age of Information Overflow. Minneapolis: University of Minnesota Press.
Yau, N. (2011). Visualize This: The FlowingData Guide to Design, Visualization, and Statistics. Hoboken: Wiley Press.