Bill Hart-Davidson, Michigan State University
Estee Beck, Bowling Green State University
Jennifer Warfel Juszkiewicz, Indiana University
Joe Warfel, Northwestern University
Session 3 of the 2015 Indiana Digital Rhetoric Symposium provided participants with an opportunity to consider seriously the role of computation in contemporary rhetorical encounters. While each presentation focused on a distinct approach to this consideration, as a whole the panel offered a consistent message: that computation–rather than, or in addition to, its effects–is worthy of and needs the close, sustained scrutiny of rhetorical scholars. While this message was sounded by a number of participants throughout the symposium, it was in this session where the argument was explored most extensively.
run_progymnasmata: What Should We Teach When We Teach Machines Rhetoric?
Bill Hart-Davidson spoke first, introducing to participants a project he and colleagues at Michigan State University have been working on in which they seek to develop a software program that can recognize and analyze rhetorical moves in writing. As he noted in his talk, there currently exists a difference in how humans and computers “read” texts, but this difference may not be significant for long. As a result, it is incumbent on those of us who claim to study digital technologies and their rhetorical impacts to involve ourselves in what Hart-Davidson called “automagogy,” the study of teaching machines.
“What type(s) of computational objects can a [given]text be?” #IDRS15
— Paul Cook (@paulgeecook) April 10, 2015
The notion of automagogy is crucial to Hart-Davidson’s project: how do our current technologies and algorithms allow us to direct machinic interpretation in particular ways? What is facilitated? What is constrained or restricted? (Among the examples Hart-Davidson provided in his initial demonstration of his group’s current dilemma was that it is difficult for computer programs to recognize pathos in the nuanced ways humans currently can identify and respond to pathetic appeals.) While Hart-Davidson and his colleagues were still exploring possibile avenues for improving machinic rhetorical analysis for the purposes that interested them, he recognized that the question we all should care about is not so much whether he meets his specific goal (although we shouldn’t ignore this!) but rather whether we can engage developers and development tools successfully to explain how and why textual analysis should attend to the myriad qualities of meaningful communication that make a given text particularly relevant and significant to the audience(s) analyzing and responding to it.
Who are the Real Digital Rhetoricians? Defining Persuasive Computer Algorithms
Estee Beck followed with a provocative call for scholars of digital rhetoric to recognize algorithms as “the true digital rhetoricians,” persuasive agents involved in the rhetorical activities that occur in computational environments. For Beck, the role that algorithms play in the kinds of digital behaviors we engage in exists far beyond that of a tool or vehicle through which “actual” rhetorical work happens; instead, algorithms as agents can be understood as co-rhetors that assist us in constructing certain arguments and that construct their own arguments–at least, they provide immediate and direct engagement that their authors may not be interested or concerned with.
However, as Beck noted, issues of agency “transfer” to computational objects (like software algorithms) must be complicated by a recognition of the diverse populations that these agents–and their authors’ intentions–impact through their activity. Questions of gender, race, and sexuality, among others, pervade seemingly objective technological considerations; however, if the algorithms that power digital technologies are not examined, dissected, queered, or repurposed, then those questions (and answers to them) will inevitably remain not only obfuscated from the populations a given algorithm impacts but alternative routes to persuasive engagements with those populations cannot occur. But, Beck observed, first we have to recognize that they serve as legitimate persuasive forces if we are to draw rhetoricians into conversations on technologies.
— e cram (@emcram) April 10, 2015
Expanding the Vocabulary of Mathematical Programming and Digital Rhetoric
Finally, Jennifer Warfel Juszkiewicz and Joe Warfel examined how rhetoric might be used more effectively to test fields that are commonly understood (whether by involved participants or by outsiders) to be arhetorical in scope–such as mathematics and computer programming. To demonstrate this potential, they explored problems related to the development and representation of a linear program–namely, how can one articulate the models they want to use to describe a given problem (or solution)?
Warfel Juszkiewicz and Warfel sustained an example case throughout their talk: they framed a linear program as the dilemma faced by a pasta factory seeking to optimize production of both macaroni and lasagna, each of which having its own unique set of production constraints. (This frame alone was illuminating for many, as the connection between persuasive description and mathematical procedure through the genre of the “word problem” became incredibly clear.) Representing this problem through different means requires–for both the mathematician or programmer and the rhetorician–an understanding of how audiences can understand the persuasive affordances of particular arguments (e.g., how does algebraic, geometric, matrix, or code representation each suggest a unique way of looking at and dealing with the case?). Ultimately, Warfel Juszkiewicz and Warfel argued, we should not shy away from exploring what disciplines and industries that seem (at least from a distance) unconcerned with rhetoric may be able to teach us about how to persuade through languages, modes, and media that provide powerful and compelling arguments for their communities.
— Nathaniel Rivers (@sophist_monster) April 10, 2015
As a whole, the members of this panel consistently asked symposium participants to attend to concerns about computation (whether as abstract or contextualized mathematical procedure, algorithmic agent, or means for rhetorical analysis) and its significance regarding rhetoric and persuasion for a contemporary age. If, the speakers have suggested, digital rhetoricians are legitimately interested in electronic technologies, then we are obligated to investigate and interrogate the mechanisms of those technologies. The means by which we might each choose to do so could be as broadly divergent as those approaches taken by the panel members; what matters is that we pursue a deeper and clearer understanding of our objects of inquiry–something that the panel demonstrated can be done in ways that (despite what some may fear) do not require an abandonment of existing methodologies or the incorporation of entirely new and unfamiliar knowledge sets to gain a critical awareness thereof.