I've been thinking some about an upcoming discussion on digital humanities in which I'll be participating at Computers and Writing, which has been fueled by a recent discussion on the Tech-Rhet list about the role of rhetoricians in supercomputing, data-mining projects.
On the list, Bill Hart-Davidson pointed to Brian McNely's proposed research as an example of what might be done. Here is a meaty part from Brian's proposal:
Following Grabill and Hart-Davidson (2010), my research is interested in "what writing does, not in what it means" (p. 1) in knowledge work environments. In other words, surfacing and tracing the literate activity of knowledge work can help us determine writing's formidable role in enabling "sociotechnical networks to hold together" (Spinuzzi, 2007, p. 268). For Grabill and Hart-Davidson, writing practices and activities are "epistemologically productive" (2010, p. 1), an assertion that foregrounds writing as heuristic, actionable, and explicitly social in the organizational and networked ecologies (Nardi and O'Day, 1999; Spinuzzi, 2003) of knowledge work. Attention to the practices of rhetoric and writing—what writing does—is paramount to our understanding of the ways knowledge moves at web-scale.
Movement, then, is a key characteristic of knowledge work in hybrid spaces. Though we have come to accept a burgeoning "ambient awareness" (Thompson, 2008; Spinuzzi, 2009) as a function of always on, always connected devices and networks in hybrid spaces, I argue for more intentional practices of situating knowledge work within heuristic frameworks of ambient informatics for the purposes of research and production. What does writing do at web-scale, how does it draw and hold people together, and how might it be leveraged?
I very much share Brian's interests here in what writing "does." The underlying question for me is how does studying a million texts rather than one assist in that investigation. Generally speaking, the premise of supercomputing is that the ability to examine very large data sets offers one a better or more accurate understanding of something. If what we are constructing here is a model of what writing does, then a million examples of writing activity might do that better than one example. Or at least this is the premise of supercomputing as I understand it. Here I think we are verging into a kind of Latourian space regarding how knowledge is composed. This interests me a great deal. I am curious to see how my colleagues will go about composing knowledge using a large data set, a supercomputer, and an array of computer science colleagues. Along with Latour, when I say this, I don't mean to suggest that the knowledge is "made up" as in being untrue, but rather that work must be done in order for the knowledge to be created.
To return to Brian's root question though, how does writing draw and hold people together? One answer is to say it doesn't. That is, what holds people together isn't "in" the writing, neither the activity nor the text. I don't intend that as a critique of the project necessarily but rather as a way of suggesting that the question may be different than it appears. Perhaps it is better to ask how people draw and hold writing together. Or perhaps we might ask how people and writing withdraw from one another through relation, creating spaces for thought and agency. That is, in part what I am suggesting is that maybe supercomputers can be a kind of Copernican moment for rhetoric, playing the role of the telescope. As with the telescope, seeing is not enough.
So that brings me back, unsurprisingly, to the horse I've been riding here for a while now. What rhetoric is and what it requires is a theory of minimal rhetorical un/relation in which thought and action, hence composition, occur. The capacities to mislead and misinterpret, for slippage, that trouble Plato are precisely the spaces that rhetorical relations open for thought. Without them, we are all (not so) super computers. These quasi-causal assemblages of affective relation shape cognition without predetermination, un/linking us to one another. I am not sure that we can see that in a supercomputer, but then again, one cannot "see" gravity through a telescope either.
Ultimately then, I suppose my curiosity lies in asking this question: can the super-computed parsing of a "million texts" finally shift rhetoric into a new theoria of writing?