Available online at www.sciencedirect.com

Computers and Composition 27 (2010) 59–70

Information Visualization, Web 2.0, and the Teaching of Writing Madeleine Sorapure UC Santa Barbara, United States

Abstract With the development of free, online, interactive visualization tools, the field of information visualization—or infovis—is being opened to diverse users and uses, and particularly to novice users who want to visualize personally relevant information. Indeed, Web 2.0 is making infovis increasingly viable as a medium for organizing, exploring, analyzing, and creatively deriving meaning from the deluge of information that we face in our everyday lives. For writing teachers, new developments in information visualization bring valuable opportunities to enhance our students’ digital and critical competencies. Projects that ask students to visualize text, personal data, and social data can provide compelling entry points into Web 2.0, as students learn about existing tools and sources of data, create their own visualizations, and then analyze the insight that they and others can gain through seeing data represented visually. Incorporating infovis assignments into writing classes can help us reinvigorate some of our standard assignments, encourage students to think critically about the software they use, and provide new opportunities for the production of digital artifacts. © 2009 Elsevier Inc. All rights reserved. Keywords: Web 2.0; Information visualization; Infovis; Data; Interactivity; Software

1. Introduction One effect of the outpouring of user-generated content in the blogs, wikis, tags, and tweets of Web 2.0 is to expand what had already seemed to be infinite: the amount of information available to us via the Internet. By inviting users to produce as well as consume content, Web 2.0 appears to be widening the gap between the amount of information we have available to us and our ability to process this information and derive value from it. However, we also find in Web 2.0 many new applications that can help us visualize and, literally, see and experience information in new ways. With the development of free, online, interactive visualization tools, the field of information visualization—or infovis—is being opened to diverse users and uses, particularly to novice users who want to visualize personally relevant information (Danziger, 2008; Pousman, Staszko, & Meteas, 2007). Indeed, Web 2.0 is making infovis increasingly viable as a medium for organizing, exploring, analyzing, and creatively deriving meaning from the deluge of information that we face in our everyday lives. For writing teachers, new developments in information visualization bring valuable opportunities to enhance our students’ digital and critical competencies. Drawing on examples from my teaching as well as referencing Web 2.0 infovis applications that help novice users visualize text, personal data, and social data, I argue that incorporating infovis assignments into our courses can be useful in at least three ways. First, we can recreate and perhaps reinvigorate some of our typical assignments by asking students to reflect, analyze, and argue largely with quantitative information. Students today encounter charts, graphs, and sophisticated visual representations of data with increasing regularity and in a variety of venues: not only in newspapers, magazines, and textbooks, but also in visualizations of personally

E-mail address: [email protected]. 8755-4615/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.compcom.2009.12.003

60

M. Sorapure / Computers and Composition 27 (2010) 59–70

relevant information such as their health or their finances. As with the other modes of communication beyond written text that are finding their way into our classrooms (e.g., video, audio, gaming), the introduction of infovis assignments can help us rethink the curriculum of our courses so that we are responsive to the needs and interests of our students. Assignments that ask students to visualize information offer fresh and highly relevant approaches to the writing and critical thinking involved in exposition, analysis, and argumentation. A second incentive for working with infovis applications is to develop students’ awareness of the limitations and biases of the software they use in our courses and elsewhere. We can encourage our students to see software not as a neutral tool but rather as an object of analysis. This goal is related to Lev Manovich’s (2008) call for the development of “Software Studies,” which would investigate “both the role of software in forming contemporary culture, and cultural, social, and economic forces that are shaping the development of software itself” (p. 4). Because the interfaces, features, and options of infovis tools are likely to be new to students, these tools are ripe for analysis in a way that more familiar software programs (e.g., Microsoft Word, web browsers) are not. The unfamiliarity of infovis tools and the fact that they often provide multiple ways to visualize the same information may help students see how software in general influences critical thinking and writing processes. Moreover, as Daniel Anderson (2007) argued, “experimenting with unfamiliar technologies can facilitate a sense of creativity that can lead to motivation” (p. 44). Finally, infovis applications offer new opportunities for engaging our students in design and production activities. The New London Group (1996) and others in our field have argued that the agency associated with designing artifacts yields analytical and rhetorical insight and ultimately can facilitate the development of cultural critique and social change (Anderson, 2007; Salinas, 2002; Sheppard, 2009). The goal is what Stuart Selber (2004) described as functional literacy: “Students should not be just effective users of computers, nor should they be just informed questioners.. .. In order to function most effectively as agents of change, students must also become reflective producers of technology” (p. 182). Particularly in discussions of visual rhetoric, the element of design is crucial; as students produce their own photographs, drawings, and digital images, they gain a better understanding of how arguments and ideologies are embedded in particular design choices (George, 2002; Salinas, 2002). Asking students to both analyze and produce visualizations of information provides yet another way of thinking about the role of the visual as it stimulates, accompanies, critiques, supplements, and/or replaces writing. In short, Web 2.0 infovis applications can enable us and our students to make the move from consuming to producing visual representations of information. 2. Web 2.0 A brief overview of Web 2.0 reveals features that make it a hospitable venue for infovis applications. Though Tim O’Reilly is generally credited with coining and defining the term Web 2.0, Paul Graham (2005) provided an interesting take on how its meaning evolved: “It was a kind of semantic deficit spending: they knew new things were coming, and the ‘2.0’ referred to whatever those might turn out to be.” In other words, Web 2.0 was conceived from the outset as a dynamic set of innovations that developed in large part to make the Web viable again after the dot-com bubble burst in 2001. Over the years, two defining elements have emerged. The first is that Web 2.0 is a platform, with applications and files stored on the Web rather than on a user’s desktop; in this arrangement, software is a service (and often a free service) rather than a product. The second defining element of Web 2.0 is participation; the Web is now the participatory Web, the social Web, the read-write Web. We see this in Wikipedia, YouTube, Facebook, blogs, tagging, product reviews, and elsewhere. The idea is that the Web “harnesses collective intelligence” (O’Reilly, 2005) and empowers users through the formation of communities and the mass publication of user-generated content.1 These two aspects of Web 2.0 are understandably enticing to writing teachers, who find a great appeal in the idea of free online tools that allow students to create new kinds of content. Indeed, writing teachers have jumped on board with blogs and wikis in particular because these kinds of projects stimulate writing for real and responsive audiences (Beach et al., 2008; Lundin, 2008). But teachers and researchers in Rhetoric and Composition and other fields are also cautious about the uncritical adoption of these technologies, both inside and outside of the classroom. David Silver (2007) concluded his otherwise skeptical review of Web 2.0 by describing our students in the following terms: “This is 1

In “Loser Generated Content: From Participation to Exploitation,” Søren Mørk Petersen (2008) offers an important critique of the participatory nature of Web 2.0. While fostering “user driven innovation and user generated content,” technologies associated with Web 2.0 also generate income for specific companies. In this way, the architecture of participation can turn into an architecture of exploitation, “a reterritorialization of free labor into a capitalistic structure of profit-making.” See also Lev Manovich (2008).

M. Sorapure / Computers and Composition 27 (2010) 59–70

61

Fig. 1. Information Visualization Process (Card, 1999, p. 9).

a generation of content creators, a generation of young people who with the help of Web 2.0 tools know how to create content, how to share content, and how to converse about content.” In fact, while students may be skilled users of Web 2.0 tools, the critical awareness they bring to the creation and publication of content for the Web is variable, to say the least. Stephanie Vie (2007) has argued that “Compositionists should focus on incorporating into their pedagogy technologies that students are familiar with but do not think critically about: online social networking sites, podcasts, audio mash-ups, blogs, and wikis.” In other words, writing teachers need to help their students cultivate a critical perspective on the Web 2.0 applications that they increasingly use. We are less likely to question the authority of data or to see the potential biases in how it was gathered, organized, and visualized; thus, it is particularly important to teach students to analyze the insights and perspectives allowed by or left out of infovis representations of data. Unlike other Web 2.0 applications, the tools and resources that are associated with information visualization have not received much attention from writing teachers or researchers. Stuart Card (1999), one of the key figures in the field of information visualization, defined it as follows: “The use of computer-supported, interactive, visual representations of abstract data to amplify cognition” (p. 7). Data is the starting point, but the end goals of infovis are, in Card’s words, “discovery, decision making, exploration” (p. 7). In other words, the information visualization process converts data to information in order to yield insight. Dominique Brodbeck, Riccarde Mazza, and Denis Lalanne (2009) described the utility of infovis in these terms: “Visual representation transforms a cognitive problem into a perceptual task, which is drastically more efficient” (28). Infovis is a relatively new field, with its first major conference, the Institute of Electrical and Electronics Engineers (IEEE) Infovis Symposia, held in 1995. It is related to the field of scientific visualization—that is, the visualization of physical entities—but also has roots in graphic design and static data displays (e.g., the work of Edward Tufte, 1990). Infovis essentially arose when technological advances enabled computer-based, interactive data displays to supplement and in some cases replace print-based, static data displays. Traditional infovis research is concerned with technical details in the design and implementation of computer applications that experts can use to explore and visualize large databases—for instance, information about world markets, items in a warehouse, or network operations. These kinds of infovis applications are intended for expert users in work-related contexts doing focused analytical tasks. Information visualization seen in this way is relevant for teachers of technical communication, for instance as regards usability testing (Mirel, 1998), but not for teachers of more general kinds of writing courses. However, in recent years, both data and the tools to visualize it have become more freely available on the Web. These new infovis applications are intended for non-expert users in everyday contexts—for example, visualizing the contents of an email inbox or the files in a photo collection. A number of infovis applications have made it fully into the mainstream: SmartMoney.com’s Market Map is a popular tool for visualizing and interacting with financial data about more than 500 stocks over different time periods and displayed on a single screen; Martin Wattenberg’s Name Voyager, one of Time Magazine’s 50 best web sites of 2009, has drawn millions of visitors interested in seeing and exploring trends in baby names over the past century. As infovis addresses these and other interests of mainstream users and takes advantage of Web 2.0 technology, it stands to become more accessible to students and more likely to find its way into their lives and our classrooms. Card’s diagram (see Fig. 1) shows the major stages in the process of information visualization. At each step, new resources and technologies come into play to create opportunities for non-experts to engage in the analysis and representation of data. As Michael Danziger (2008) observed, “this new relationship with information is breaking down previous models of data consumption that conceptualize raw data as something only trained analysts and other experts deal with” (p. 12). In the first step, raw data is transformed into data tables structured by specified variables, values, and other metadata. Data that was once the domain of governments and large corporations is now available to ordinary users via Web 2.0; we

62

M. Sorapure / Computers and Composition 27 (2010) 59–70

have increasing access to structured data sets on a wider array of topics as well as tools to restructure the raw data. In the next step, data tables are mapped to visual structures, charts, or graphs that visually express the data. Researchers in the field of infovis create effective visualizations by applying their programming expertise as well as their understanding of properties of human perception. But non-experts can engage in this step as well by working with an increasing number of tools that allow them to upload data and select from a range of visualization options. In the final step, views of the visual structure can be transformed interactively through various controls and probes—for instance, using search functions, zooming in and out, or brushing over areas of the graphic to reveal pop-up windows with additional information. Here again, the development of free, user-friendly, web-based visualization tools enables non-experts to explore data in a way that was not possible even a decade ago, when visualization software was proprietary, expensive, and challenging to learn. Charles Kostelnick (2007) has argued that being able to interact with the application “empowers [users] to visualize data on their own terms and.. . places them in a quasi-collaborative relationship with the designer” (p. 287). Though the designers of infovis applications clearly enjoy a greater degree of creative and analytic power in setting up the processes and parameters of the visualizations, non-expert users nevertheless bring into play a range of interests, skills, and goals that can open these applications to possibilities unforeseen by their designers. Infovis is entering the mainstream because of technological developments, clearly, but the influx of two groups of users—journalists and artists—has helped to extend the horizon of infovis beyond expert, task-driven, analytical applications and into our everyday lives. In the field of journalism, The New York Times has been at the forefront in developing interactive visualizations to accompany, illustrate, and extend news stories. For instance, Naming Names is an infovis application that shows who referred to whom and what each candidate said about the others in one of the early debates in the 2008 Presidential election. Another interactive visualization at The New York Times, Mapping the Human Diseasome, lets users zoom in to see connections between different diseases based on genes that those diseases have in common. More broadly, news visualization sites such as 10 X 10, Newsmap, Digg Labs, and DoodleBuzz present new ways of looking at the entire landscape of the news, so to speak. That is, these programs offer new opportunities for a general audience to interact with the news and view it differently. Artists have also helped to bring infovis into the mainstream and expand its sense of purpose and rhetorical complexity. While journalistic applications give us valuable experience interacting with data visualizations and seeing their usefulness, artistic projects support an idea of data as evocative, ambiguous, and open to interpretation, thereby inviting reflection, contemplation, and even pleasure. Andrea Lau and Andrew Vande Moere (2007) pointed out that “[A]n increasing number of visualization art (or data art) works have emerged that aim to express the subjective experience of our information society by artistically motivated but data-driven visual forms” (p. 1). Perhaps most notable among these artists is Jonathan Harris, whose We Feel Fine and I Want You to Want Me are aesthetically driven, interactive representations of data drawn from online blogs and dating sites. Other artists work with data to explore issues of privacy and self-representation. For instance, Brooke Singer’s Databody presents a silhouette filled in with data that Singer has collected about herself from a National Decision Systems database. Singer (2002) suggested that data has become the “perfect material for painting contemporary portraits.” In addition to artists who work with data as their medium, there are information visualization researchers who explicitly include aesthetic considerations in their projects (Lau & Vande Moere, 2007). As with other Web 2.0 phenomena, then, the influx of new users of visualization tools brings with it substantial shifts in traditional practices and conceptions. Writing teachers can therefore take advantage of opportunities for new assignments, activities, and course goals as research and practice in infovis become more public-facing, more open to diverse users and uses. 3. Text visualization Writing teachers might be particularly interested in infovis applications on the Web that allow student users to visualize text written by themselves or others. Users can upload the text, visualize it in different ways, and then analyze it to come to greater insight about the text. These applications evoke some of the other ways that we ask students to visualize texts—for example, sentence diagramming, outlining, and clustering. We are familiar with the benefits of seeing text disassembled and visually rearticulated and reconnected in some way. Computer-based infovis applications for text visualization multiply the possibilities both for re-seeing text and for developing analytical insight into it. As Geoffrey Rockwell (2003) explained, these applications enable the creation of a hybrid text that is “authored not just by the original author, but also by the user’s choices and the procedures used to generate it” (p. 213).

M. Sorapure / Computers and Composition 27 (2010) 59–70

63

Fig. 2. Word clouds from Wordle’s Gallery at .

The word cloud is probably the most common text visualization on the Web; we see word clouds at news sites and on blogs and elsewhere, being employed to visually represent a text by displaying its most commonly used words. The New York Times, for instance, offers an interactive set of word clouds for all of the Inaugural Addresses at “Inaugural Words: 1789 to the Present.” In a related project, Jonathan Feinberg, an IBM researcher, has created a comparative analysis of word clouds of inaugural addresses at . The underlying and certainly debatable principle of word clouds is that frequency is the measure of importance—that is, that the key ideas of a text will become evident by isolating the words most frequently used in it. With Wordle, also created by Feinberg and made available online in June 2008, word clouds have taken on new aesthetic and creative dimensions. Subtitled “beautiful word clouds,” Wordle is described on its site as “a toy for generating word clouds from text that you provide.” Users enter the text they want to visualize; then they can manipulate fonts, colors, layouts, and other visual elements to yield a word cloud that is both visually appealing and informative. Wordle’s Gallery provides examples of visualizations created by users, some of whom use the tool against its grain by focusing on the design rather than the analysis of text—that is, they play with the font, color, and layout features of Wordle to craft an invitation or birthday card, to write their name, or to create a personal message (see Fig. 2). Wordle was recently incorporated into an infovis site called Many Eyes, created by other researchers at IBM (Fernanda Viegas, Martin Wattenberg, and others). Many Eyes is a portal where registered users can upload text as well as other kinds of data and create visualizations using tools there; users share the data and data visualizations with others as well as receive and offer feedback on the shared content. In this way, there is a collaborative, social networking element to Many Eyes—a Web 2.0 element that makes Many Eyes something like Facebook for infovis nerds. In addition to Wordle, Many Eyes provides a word tree tool that allows users to select a word or phrase in a text and visualize the contexts in which that word or phrase appear. The tree-like visualization structure makes patterns and repetitions in the text more evident and thus can reveal recurrent themes and stylistic techniques. The richest set of text visualization tools on the Web is offered by TAPoR, the Text Analysis and Visualization Portal created by researchers at the University of Alberta. TAPoR tools allow for complex searches of large texts uploaded by registered users and provide options for both statistical analysis and visualization; the tools can be used

64

M. Sorapure / Computers and Composition 27 (2010) 59–70

Fig. 3. Visual collocates in Maya Angelou’s “Still I Rise” (Carolyn Biatek, redrawn).

for creating concordances, for showing words in context, for listing collocates of words, and for generating many other statistical results. For visualizing text, TAPoR offers both statistically based tools and other, more aesthetically driven and impressionistic tools. For instance, the Visual Collocates tool creates a network diagram showing words that are collocates of a specific search term. In this interactive application, each collocate can itself be expanded to show its own collocates, yielding a richly detailed network of collocate relations. A more impressionistic tool is the Word Brush, with which a user can set the color, size, rotation, and other variables and then use his or her mouse to “brush” the uploaded text across the screen. Taken together, the tools at TAPoR are meant to facilitate what Geoffrey Rockwell (2005), TAPoR’s Project Leader, described as a “targeted reading” of a text that tests intuitions or hypotheses a reader has already developed. Rockwell also notes that the use of text analysis tools encourages reflection on the questions we ask of texts and on the formalization of queries, a seemingly strong fit for composition pedagogy. In the Fall 2008 section of my “Writing in New Media” course at UC Santa Barbara, I asked my students to create one or more visualizations of a text of their own choosing using the tools offered by TAPoR, Many Eyes, and Wordle. I then asked them to write about what they learned in the process: about the text they chose, about the tools they used, and about information visualization in general. This assignment yielded interesting results, in part because students chose quite a range of texts to analyze; they evidently saw the possibility for gaining both personal and academic insight through text visualization. One student visualized the personal statement she had written for her law school application, entitling her corresponding essay “What Was I Really Writing About?” Another student compiled all of the lyrics from a recent album by the band Coldplay and used various applications to decipher common threads in different songs. One English major used visualization tools to analyze Mersault’s relationship with his mother in Camus’ The Stranger while another English major comparatively visualized segments from Tennyson’s “The Lotus Eaters” and Homer’s The Odyssey. The visual collocate diagram above (Fig. 3) shows Carolyn Biatek’s analysis of several word groupings in Maya Angelou’s “Still I Rise.” Specifically, the diagram shows negative collocates associated with “outsider” pronouns (“you” and “your”); more abstract or universal collocates of the “independent” article the; and positive, active collocates of the personal pronoun “I.” Biatek used the visual collocates tool at TAPoR as well as several other text analysis tools to both discover and express her sense of how Angelou’s poem achieves its defiant, confident tone. Most students came to some new insights or confirmed existing intuitions about the text they chose, and many also developed a better sense of the strengths and shortcomings of the visualization tools they used. For instance, one student chose to analyze the 2008 Vice Presidential debate transcripts in order to confirm her hypothesis that Joe Biden had won the debate and had delivered more powerful arguments in his favor; in actual fact, her comparative word cloud revealed very few differences between the words spoken by Biden and Sarah Palin. The student’s analysis caused her to posit the importance of other cues in the debate (e.g., the candidates’ delivery and demeanor) as well as to reexamine the influence of her own bias in how she interpreted the debate’s results. Moreover, the student came to see that word frequency and repetition, the sole factor measured by popular word cloud tools, may not be the best indicator of importance—for instance, that a specific fact mentioned only once or twice by one of the candidates may have carried more sway than the repeated words and phrases offered by both candidates. Another student, analyzing a

M. Sorapure / Computers and Composition 27 (2010) 59–70

65

short narrative poem by Raymond Carver, came to similar conclusions about the weaknesses of computer-based text analysis tools and their inability to deal well with context or nuance in a shorter text. Web 2.0 infovis applications that focus on seeing and analyzing text offer students new ways of interacting with and thinking about their own writing and the writing of others. Rockwell (2003) argued that computer-based text analysis tools are built on the principle of “research as disciplined play” (p. 213); they encourage exploration from multiple angles as they recombine and reconfigure texts in potentially interesting ways. Using these applications can also highlight significant differences between how computers and people “understand” texts and create meaning from them. 4. Personal infovis Another type of infovis application that might be interesting to writing teachers enables users to visualize personal information, data about themselves that they or others have collected. A personal infovis assignment remakes the personal, reflective, or autobiographical writing tasks that we frequently ask our students to do. With a personal infovis assignment, the material comes in the form of numbers that students gather, analyze, and visualize in order to develop greater insight about themselves. Drawing insight from quantitative rather than qualitative data may be a new activity in the writing classroom, but it is an approach taken by an increasing number of people in everyday venues, facilitated by Web 2.0 data collection and visualization tools. In a recent cover story for Wired magazine, for instance, Gary Wolf (2009) proposed that “a new culture of personal data” (p. 94)—a “personal metrics movement”—is taking shape, particularly in the areas of health and exercise. Indeed, there is no dearth of personal data being gathered about each of us, although access to it is not always simple. We may have relatively easy access to information about the specific purchases we make via our credit cards, but getting access to information about our own medical histories—information that is stored in databases of insurance companies or medical offices—may be significantly more challenging. Questions of privacy, ownership, and control of this information necessarily come into play. Web 2.0 complicates the issue in that users knowingly and willingly post a vast amount of information about themselves on the blogs, social networking sites, and other online venues to which they contribute. Asking students to gain access to data about themselves or to collect data about themselves (for instance, through observing and recording their own patterns of behavior or spending) is one way of drawing attention to issues of privacy and control about which students might not have given much thought. In representing themselves via data, they can, in a sense, gain an understanding of and some small measure of control over the data that circulates about them. A model for this practice can be found in the Annual Reports created by graphic designer Nicholas Feltron. Since 2006, Feltron has produced web- and print-based documents with representations of data that he collects about himself over the previous year. He essentially treats himself as a corporate entity, with professionally designed reports that contain visually striking charts, graphs, maps, and other representations. However, these annual reports focus not on professional matters, but rather on personal and often trivial data drawn from his everyday life, such as the number of subway trips he took, the number of meals he cooked at home, his favorite book, his Michael J. Fox sighting, and so on. Even as he parodies the genre of the annual report, Feltron claims ownership of his own personal information and control over the representation of himself created through it. There are a surprising number of Web 2.0 applications where people can upload data about themselves and eventually convert that data into some kind of visual representation. Some sites are more oriented toward business or other sorts of “serious” data, although users can and do upload personal data to these sites as well: Many Eyes, Swivel, iCharts, Widgenie, Trendrr, and Track-n-graph. Other sites are explicitly focused on seemingly trivial and mundane personal information. For instance, Feltron recently collaborated with Ryan Case to create Daytum, a site and widget where users can “collect, categorize and communicate” data from their everyday lives. A similar site, Mycrocosm, bills itself as “a web service that allows you to share snippets of information from the minutiae of daily life in the form of simple statistical graphs.” At Mycrocosm’s homepage, visitors are afforded visualizations such as “Eriktrips”’ sleeping patterns in the month of January and the weight gain of “GayHall” over the past year. Going one step further, the site Bedpost moves into the realm of the truly personal; its homepage describes Bedpost as “a personal web application that will give you some insight into your sex life. . .. Simply log in after every time you have sex and fill out a few simple fields. Pretty soon, you’ll have a rolling history of your sex life on which to reflect.” Perhaps indicating that there are limits to the perceived helpfulness of personal infovis applications, the site has been in beta for quite a while.

66

M. Sorapure / Computers and Composition 27 (2010) 59–70

In my “Writing in New Media” course, students used the personal infovis assignment to look at multiple sites that collect data about users. One student accessed her credit card bills and online banking statements to track her spending over the past year; she could target specific stores and even specific types of purchases at stores, and she was able to collect information as finely tuned as the amount of lettuce that she purchased during the year. Other students used iTunes or last.fm to track their music-listening habits; these applications keep detailed account of what genres of music and particular songs one listens to, largely in order to recommend new purchases. In examining this data as personally rather than commercially meaningful, students came to new insights into their music-listening habits and preferences. Another student, Gregory Burnett, looked at all of the metadata stored in his iPhoto account in order to gain insight into trends and qualities of the 20,357 photographs he had taken over the past three years. Fig. 4 shows the representation that he created in Photoshop and modeled on Feltron’s reports. Looking closely at the data associated with his hobby, Burnett tried several ways to represent its sheer quantity (number of photos, time span, size of library in gigabytes,

Fig. 4. Visualization of data from iPhoto (Gregory Burnett).

M. Sorapure / Computers and Composition 27 (2010) 59–70

67

number of photos per year, hypothetical length in miles); he also began to make meaning from the collection by identifying some key themes (location, time, best and worst, use of flash). In short, analyzing and visualizing data gave Burnett an entry point into understanding his “photographic life,” and this quantitative entry point certainly opened different kinds of insights and perspectives than Burnett would have gained from a more qualitative reflection on his hobby. More broadly, then, by asking students to find or collect data about themselves and to analyze and represent this data visually, personal infovis assignments can give students new opportunities to reflect on and rearticulate a sense of themselves as actors in an information society. 5. Social infovis The last type of infovis assignment that might be useful to writing teachers engages students in finding, analyzing, and visualizing data related to a social or global issue. Here the goal is typically persuasive, with the data and its visualization serving as evidence to argue a point or support a theory. Rather than incorporating the data in an essay, students use infovis applications as well as image editing programs to visualize the data and render it meaningful for a particular audience and context. This kind of assignment calls for the development of a visual argument and therefore has a more robust sense of audience than does the personal infovis assignment, where the data is likely to be much more significant and nuanced for the author than for the audience. Working with data about a social or global issue, students can experiment with different strategies for making information meaningful to others and can come to a better critical understanding of the charts, graphs, and other data visualizations they regularly encounter. One aspect of Web 2.0 that facilitates this kind of assignment is the proliferation of sites where data is freely available. Governments, non-governmental organizations, and other organizations are putting vast quantities of data on the Web and are also taking some time to make their databases searchable and manageable. For instance, UNdata is a searchable online database of United Nations statistics, with “14 databases and 6 glossaries containing over 55 million data points and covering a whole range of statistics including Population, Industry, Energy, Trade and National Accounts.” Infochimps, “built by data nerds, used by data nerds,” offers thousands of datasets from a range of sources (e.g., censuses, texts, sports statistics, stock market reports) and bills itself as “the world’s best repository for raw data—a sort of giant free allmanac, with tables on everything you can put in a table.” Files are available in a variety of formats, and, as with UNdata, users simply click to download the data. Other sites offer both data and visualization tools. For instance, NationMaster combines data from sources such as the UN, CIA World Book, and OECD; it offers a simple search function as well as dropdown menus from which users can select the category (e.g., Agriculture, Crime, Industry, Mortality), as well as a particular statistic within that category. For many of the statistics, there are options to view the data as a bar graph, pie chart, and/or map. There is also the option to create a correlation between one dataset and another, again with choices to visualize this data in different ways. Perhaps the best-known searchable data and visualization tool on the Web is GapMinder, made famous in the online world by Hans Rosling’s TED talk in which he demonstrated both GapMinder software and the persuasive power of infovis. On its FAQ page, GapMinder describes its goal as making data “more accessible and easier to use for instant visual analysis. . .. The idea is that all people, independently of their political agenda, should get free access to already existing statistics about global development to easily improve their understanding about the complex society.” In addition to exploring the datasets available through GapMinder, users can also upload their own datasets and visualize them by using a Google Gadget called Motion Chart. In my “Writing in New Media” course, students were asked to select data and present it in such a way as to guide viewers to a certain meaning or interpretation. In other words, the visualization of data needed to make a point about the data itself. One student drew information about poverty—particularly poverty among children—from the 2007 U.S. Census Report and included statistics along with images in a collage that was intended to add emotional, affective impact to the numbers. In the example included below in Fig. 5, Maggie Kanter used GapMinder to create a chart comparing US aid per person to the percent of the population at the poverty line in certain countries during the year 2000. However, Kanter modified the GapMinder interface, replacing the generic colored bubbles that represent each country. In place of the bubbles, she put images of consumer goods that cost the same as the US aid spent on each person in the country (e.g., a throw pillow for Egypt, a purse for Mozambique, a pair of sunglasses for Bangladesh). Kanter intended these images to convey a sense of the seemingly random inequities in aid distribution; replacing money with “gifts” raises the question of why certain countries get more or less aid money from the US. Kanter also intended her visualization to critique what she saw as a somewhat frivolous consumerist attitude in our country, as if

68

M. Sorapure / Computers and Composition 27 (2010) 59–70

Fig. 5. Visualization of US foreign aid (Maggie Kanter).

we can understand foreign aid only as a kind of Walmart-infused act of shopping and giving. Another modification that Kanter made to the GapMinder interface helped her to convey this critique more effectively: included alongside each image in Kanter’s chart is the country’s name and the amount of aid per person. This information only appears on the GapMinder chart if the user brushes over the country’s bubble with his or her mouse; as a result, the user can only see information about one country at a time and is therefore hampered in making comparisons. With the countries’ names and aid amounts visible in Kanter’s static display, aid inequities become more evident. As students find and analyze data and design their own visualizations, they can develop more effective strategies for reading the charts, graphs, and other data displays that they encounter in the news, online, and elsewhere. In addition to developing a general “graphicacy,” the ability to make sense of the different ways that data can be visualized, the act of designing a visualization that is intended to persuade or prove a point can lead to a sharper critical eye, or what Carlos Salinas (2002) described as “critical savvy regarding configured images” (p. 168). Salinas argued that producing images helps students see design as “the strategic composition of visual representations that have cultural significance” (p. 168). Design enables analytical insights, and a particularly important insight is that images, including data visualizations, are not neutral or objective representations of reality but rather are created as a result of choices that represent specific interests, ideologies, and values. As Kanter’s visualization shows, it is not a matter of lying about the data or misrepresenting it, but rather making design choices that effectively convey one’s perspective or change the attitude of one’s audience. 6. Conclusion For writing teachers, projects that ask students to visualize text, personal data, and social data can provide compelling entry points into Web 2.0 as students learn about existing tools and sources of data, create their own visualizations, and then analyze the insight that they and others can gain through seeing data represented visually. Experimenting with infovis applications can encourage students to explore both written texts and their own identities in new ways and from multiple angles. Moreover, assignments that draw on infovis applications open up new possibilities for writing

M. Sorapure / Computers and Composition 27 (2010) 59–70

69

teachers to present the standard modes of analysis, personal writing, and argumentation in the context of data-driven inquiries. By incorporating these kinds of assignments, we can help prepare students to be savvy participants in the world of Web 2.0 as well as critically informed undergraduates and professionals. Indeed, Charles Bazerman (2007) recently argued that “The competence to find one’s way within relevant data, to know how to manipulate and think about it, licenses the professional as professional, in industry and public practice as well as in the academy” (p. 99). Infovis applications provide a way for writing teachers to begin to address this competence, while also giving students practice in other key skills involved in the production and analysis of digital work. Determining the most appropriate entry point into the array of possibilities offered by infovis depends on the classroom context, including such factors as the skills of students and teachers, availability of technology, and curricular goals. Some of the infovis applications available to students are what Daniel Anderson (2007) termed “low bridge technologies”; as Anderson observed, “the entry-level nature of low bridge technologies ameliorates difficulties that can shut down flow, but the challenge of composing with unfamiliar forms opens pathways to creativity and motivation” (p. 43). Other infovis applications are far more sophisticated and require the kind of time and skill that would make them appropriate only in a specialized context. In either case, as students interact with these programs and find, organize, analyze, and visualize data, they are exercising rhetorical and technical skills that are increasingly relevant in this information age. Acknowledgments I’m grateful to Randall McClure for his helpful feedback on earlier drafts of this article. Madeleine Sorapure is a Lecturer in the Writing Program at the University of California at Santa Barbara, where she teaches composition courses and directs the Multimedia Communication track of the Professional Writing Minor. Her articles have been published in Computers and Composition, Kairos, Biography, and elsewhere, and she serves as editor of the Inventio section of Kairos.

References Anderson, Daniel. (2007). The low bridge to high benefits: Entry-level multimedia, literacies, and motivation. Computers and Composition, 25(1), 40–60. Bazerman, Charles. (2007). WAC for cyborgs: Discursive thought in information rich environments. In Takayoshi Pamela, & Patricia Sullivan (Eds.), Labor, Writing Technologies and the Shaping of Composition in the Academy (pp. 97–110). Cresskill, NJ: Hampton Press. Beach, Richard, Anson, Chris, Kastman, Breuch, Lee, Ann, & Swiss, Thom. (2008). Teaching writing using blogs, wikis, and other digital tools. Norwood, MA: Christopher-Gordon Publishers. Brodbeck, Dominique, Mazza, Riccarde, & Lalanne, Denis. (2009). Interaction visualization: A survey. In Denis Lalanne, & Jürg Kohlas (Eds.), Human Machine Interaction (pp. 27–46). Berlin: Springer-Verlag. Card, Stuart. (1999). Information visualization. In Stuart Card, Jock Mackinlay, & Ben Shneiderman (Eds.), Readings in information visualization: Using vision to think (pp. 1–34). San Francisco, CA: Morgan Kaufmann. Danziger, Michael, (2008). Information visualization for the people.(Master’s Degree Thesis, Massachusetts Institute of Technology, 2008.). George, Diana. (2002). From analysis to design: Visual communication in the teaching of writing. College Composition and Communication, 54, 11–39. Graham, Paul, (2005). Web 2.0. Retrieved January 14, 2009, from http://www.paulgraham.com/web20.html. Kostelnick, Charles. (2007). The visual rhetoric of data displays: The conundrum of clarity. IEEE Transactions on Professional Communication, 50, 280–294. Lau, Andrea, & Vande Moere, Andrew. (2007). Towards a model of information aesthetics in information visualization. 11th International Conference Information Visualization (IV’07): 87-92. Lundin, Rebecca Wilson. (2008). Teaching with wikis: Toward a networked pedagogy. Computers and Composition, 25(4), 432–488. Manovich, Lev. (2008). Software takes command. Retrieved January 10, 2009, from http://lab.softwarestudies.com/2008/11/softbook.html. Mirel, Barbara. (1998). Visualizations for data exploration and analysis: A critical review of usability research. Technical Communication, 45, 491–509. New London Group. (1996). A pedagogy of multiliteracies: Designing social futures. Harvard Educational Review, 66, 60–92. O’Reilly, Tim, (2005). What is Web 2.0? Retrieved January 14, 2009, from http://www.oreilly.de/artikel/web20.html. Petersen, Søren Mørk., (2008). Loser-generated content: From participation to exploitation. First Monday, 13(3). Retrieved January 20, 2009, from http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2141/1948. Pousman, Zachary, Stasko, John T., & Mateas, Michael. (2007). Casual information visualization: Depictions of data in everyday life. IEEE Transactions on Visualization and Computer Graphics, 1145–1152. Rockwell, Geoffrey. (2003). What is text analysis, really? Literary and Linguistic Computing, 18, 209–219. Rockwell, Geoffrey, (2005). What is text analysis? Retrieved February 4, 2009, from http://tada.mcmaster.ca/Main/WhatTA. Salinas, Carlos. (2002). Technical rhetoricians and the art of configuring images. Technical Communication Quarterly, 11, 165–183.

70

M. Sorapure / Computers and Composition 27 (2010) 59–70

Selber, Stuart. (2004). Multiliteracies for a digital age. Carbondale, IL: Southern Illinois University Press. Sheppard, Jennifer. (2009). The rhetorical work of multimedia production practices: It’s more than just technical skill. Computers and Composition, 26(2), 122–131. Singer, Brooke. (2002). Against data determinism in a networked world. Retrieved May 15, 2006, from http://www.bsing.net/thesis.pdf. Tufte, Edward. (1990). Envisioning information. Cheshire, CT: Graphics Press. Vie, Stephanie. (2007). Digital divide 2.0: ‘Generation M’ and online social networking sites in composition. Computers and Composition, 25, 9–23. Wolf, Gary. (July 2009). Know thyself. Wired, 92–95.