It’s beginning to seem to me that the best thing these digital tools and modeals can do for us is offer a kind of critical distance. They let us view works from above, not from within; Elizabeth Gaskell might say, from a “mount of observation.” Voyant (and even more than voyant, woodchipper) gives us unbiased analyses of important themes and threads. Maps, graphs, and trees let us see data in new patterns that we weren’t necessarily looking for; the tools give us distance. They remove some of the subjectivity that has always been integral to literary criticism. But Daniel Rosenberg’s talk left me convinced that there was no such thing as raw, unbiased data. If data is rhetorical, then it is all biased or skewed in some way. Even by choosing what information to record, we skew the results. Analysing texts in Voyant only works well if we remove certain common words, but what do we miss when we remove “it” and “he” and “she” and “I”? Data models pretend to offer critical distance, but I’m not sure whether they can—or whether eventually they will, when digital humanists have developed a basic set of data points that can be gathered. When a chemist does an experiment, he measures temperature, time, mass, volume, color, and so forth; what are the literary scholar’s equivalents?

I’m not at all sure Franco Moretti gives us an answer to that question. The examples he sets out are suggestive, but to me, Moretti seems to be explaining only one part of a larger theory. I like to think of literature as a conversation of sorts. Moretti’s idea that form comes from force is amazing: social forces shape a writer, who creates a work. The work then goes out into the world and is mediated and received.

Moretti skips over half of that conversation: he stops once the work has been written and published. His graphs of genres, and all the reasoning from those graphs, are based on publication numbers—how many books in each genre have been published. But how many of those novels were flops? How many sold well? Were books read once and then discarded, or did readers reread novels—and how often? Did some novels—or some genres—have longer reading lifespans (as opposed to publishing lifespans) than others? That’s a much more complicated question, of course—but perhaps not an impossible one to tackle. Maybe there’s information about sales in bookseller’s archives or publisher’s archives, and information about circulation in circulating library archives. This is also, of course, where annotations and mediated copies come in—this is the sort of information that the books we’re putting on Omeka can give us, on the large scale.

I loved the section on maps, even though it did seem to cover only half of the conversation again. He talks about the way village story collections display the social geography and mindset (mentalité) of the time; but that’s an explanation of how social forces shaped the work. Did the works have any influence on society? Did they provoke any response? I can’t imagine how to figure this one out. In any case, this sort of intersection of real-world humans and art is exactly why I study English, and exactly why I think it’s an important discipline: the humanities show how we process our world, and are how we process our world. But that means that it’s even more necessary to study the second half of the conversation—not just what the author produces, and why and how it was produced, but what happened to it once it was produced. The trickiest bits might be the most important.