Conclusion

This suggests giving statistical material…a very sharp second look before accepting any of them. Sometimes a careful squint will sharpen the focus. But arbitrarily rejecting statistical methods makes no sense either. That is like refusing to read because writers sometimes use words to hide facts and relationships rather than to reveal them (121). Darrell Huff, How to Lie with Statistics

Certainly, for many exploratory data analyses, there are no “right” answers to the questions of which layout or color method is more appropriate—which is another reason why data visualizations are as rhetorical as they are quantitative. I provided arguments for why various methods and options made sense given the assumptions of what word clouds are intended to portray, but as with any rhetorical artifact (statistical or otherwise), these choices depend just as much on audience, delivery, and the broader rhetorical situation as they do the underlying data and project goals. In moving beyond the fact/deception binary for information rhetorics, the more difficult issue raised by the relationship of rhetoric to statistical reasoning is which choices—among an increasing variety of reasonably non-deceptive choices for data visualization—most effectively represent the underlying data? As more and more types of data visualization are made available, answers to questions like these will continue to venture further away from purely quantitative answers.

While this article focuses on data literacy and statistical reasoning more broadly conceived, these questions are becoming directly relevant to the actual methods and research conducted within digital rhetoric and writing studies. Significant progress has been made in the digital humanities to implement these methods for the computational reading of literary corpora, but much work remains to bring similar tools and methods into writing studies and digital rhetoric. As Jim Ridolfo and Bill-Hart Davidson argue in the introduction to Rhetoric and the Digital Humanities, the digital humanities and rhetoric have many shared interests and overlapping concerns. In my forthcoming chapter for Writing, Rhetoric, and Circulation Studies, I argue for developing applications of text mining and natural language processing tools to study how multimodal writing and digital artifacts circulate among social networks. This webtext provides a basic introduction to these tools, but text mining and natural language processing are scalable to many broad and diverse applications for digital rhetoric and writing studies.

In a recent publication for Enculturation, Eric Detweiler conducts an empirical analysis of the relationship between the terms “rhetoric” and “composition.” By mining the text from works cited lists from College Composition and Communication and Rhetoric Society Quarterly, Detweiler examines the pairing of these terms and their tentative relationship as portrayed in citation data. Responding to the work of Derrick N. Mueller mentioned in the introduction of this webtext, Detweiler’s research provides a “picture of the most influential citees, journals, and key terms—as well as the trajectories of these influences over the past decade.” Other projects, such as Jim Ridolfo’s Rhetmap.org use Modern Language Association Jobs Information List data to map trends in rhetoric and composition hiring practices. Research projects that help scholars understand trends in the field will continue to have a reciprocal feedback affect on the types of research that are prioritized in the future. An increased focus on data literacy will help scholars contribute to this important work and provide the necessary awareness to be critical of its effects.

In many ways, the renewed interest in empirical studies, data-intensive research, and information rhetorics is a productive variation of Jeanne Fahnestock’s research into the rhetoric of science. Jeanne Fahnestock works to complicate the idea that science and rhetoric are mutually exclusive territories: “the terms on both sides of the preposition in the phrase ‘rhetoric of science’ come out somewhat changed. A rhetoric of science puts pressure on both science and rhetoric” (Rhetorical 278). Fahnestock’s Rhetorical Figures in Science, understands the rhetoric of science as using stylistics, figures, tropes, and schemes to read and rewrite the persuasive elements of scientific inquiry and writing. Although she calls for a more integrated convergence of rhetoric and science, until recently the convergence has been relatively limited to the “formal possibilities, identified ultimately in both rhetoric and neurolinguistics, and [how] the historical particulars together constitute the ‘available means of persuasion’” (175).

In the areas of data literacy and information rhetorics, new possibilities for complex convergences of rhetoric and science continue to emerge. This webtext focuses on statistical reasoning and data visualization as rhetorical constructions, and as writing studies continues to expand its fluency (techne) with data literacy, quantitative and rhetorical methodologies will enmesh with new productive reciprocity. The questions that rhetoric and writing studies will have to continually re-examine in years to come are: How do we supplement well-established rhetorical methodologies with data-analytical frameworks? To what extent should statistical methodologies be present in advanced writing curriculum and graduate coursework? As more scholars utilize data-intensive methodologies, how will this affect peer review and the sharing of new knowledge in the field? There are no easy answers to these questions, but there are no detours around them either.