“We write sentences. About sentences.”

or, towards a very conservative practice of the digital humanities.

The title of this post was spoken during the Q&A of a roundtable I saw at CSECS (the Canadian Society for Eighteenth Century Studies) at Montreal last year. The title of the panel was something like “How I practice.” Literary scholars love nothing more than a chance to exercise their existential dread in a professional context. And this was, intermittently, just such an occasion. The Q&A, accordingly, had taken a detour into how we do not practice; the perceived near-enemies of literary-historical work, of which the example du jour for many is digital humanities. (My favourite piece of writing about “Near Enemies” is Lisa Ruddick’s piece here. It is wonderful.) The speaker of the phrase in question — whom, and whose work, I like and admire very deeply —  was inveighing, quite reasonably, I suppose, against a particular kind of DH: one which is not practiced, to my knowledge, except perhaps in the minds of deans and donors. These are, of course, the most dangerous places of all to practice anything.

This speaker, like those who evince skepticism about the real probative value of DH, seemed to describe DH as misplaced science-envy. To argue that quantitative methods had no place in a discipline whose bedrock is semantic. And let’s be honest; it is a very elegant and compelling formulation. Everything that follows below, I should be clear, takes the form of perverse and spirited agreement.

Part of the appeal of this declaration is that it links what we (literary critics) produce with what they (our objects of study) produce. It’s a soothing ouroboros. Moreover, this formula (joke!) reassures by telling us, at last, what we actually study: sentences. In Two Aspects of Language and Two Types of Aphasic Disturbances Roman Jakobson delved into the question of what “we” study and hit an insurmountable snag. That snag is that the smallest observable semantic unit was either multiple, shifting in size, or infinitesimally small (in a turtles-all-the-way down, pure Saussurean way that links up the most hardcore book historians and paleographers with deconstructionists; all cussedly stubborn types near to my heart). Whether the fundamental unit of study was a paragraph, a sentence, a clause, a phrase, a word, or a letter was ultimately undecidable. These units co-exist happily, of course: they overlap, and the tensions, frictions, cantileverings and overlaps among semantic unit-boundaries produce many of the effects that we call “literariness” (see also: language games).

But! This scholar’s formulation had cut this semantic knot with a slightly positivist sword. There was a touch of “at-the-end-of-the-day”-, “let’s-be-real”-, “come-on-now”- weariness as the speaker intoned these words. Not only was the question of our objects laid to rest; so too was the nature of our productions: sentences.  The room seemed to wobble on its axis a little, and then settle to revolve around this new resolution, shorn of Wittgensteinian, Saussurean, Jakobsonian, or Derridean anxieties. We write sentences. About sentences.


There’s been a turn in the literary academy, perhaps now coming to its close, toward object-oriented studies. (When the first enthusiasm of a “turn” or “moment” burns out, that approach just goes in the general critical toolkit.) It would be invidious to mention particular scholars or monographs, but from it-narrative-studies to object-oriented-ontology (and its oddly unmentioned corollary, object-oriented-epistemology), objects have themselves been a key object of literary studies for some years now. Taking this approach seriously, as I know that the particular scholar who gave me my title at CSECS does, means taking seriously the notion that meanings, consciousnesses, arguments, intentions can be found in deeply unconventional vessels. The outer reaches of extended mind theory shade into a sort of secular pantheism (see Tegmark’s paper on consciousness in matter here, from a different angle). This new(ish) plurality of substantive, possibly agential, vectors of semantic content opens up new arenas of possibility. How can we reconsider the embroidered letter “A”,  or the strawberry-patterned handkerchief, or the cracked golden bowl, as entities in themselves?

Regular readers of this blog, such as they are, will know where I am going with this. It follows that as sentences might not be the only objects of study — think of the list of objects in Bishop’s “Crusoe in England” whose souls have “dribbled away” and “won’t look at me at all” — , so too sentences might not be our only products. There’s a DH slogan kicking around somewhere that goes “Projects, not Papers.” It suggests that an ongoing big DH project could or even should be significantly weighted in tenure review; that overseeing such an entity requires knowledge, insight, analytical skill and an investment of time and care easily on commensurate with that required for a peer-reviewed article or even a monograph.


The strong version of my response to that speaker last October would be this: at its most generative and plural, DH allows us to create objects about objects. I am going to be working on a project called Phono-Post soon, and we will be making visualisations of the lives lived by these objects, as we can reconstruct them from the archive we have of them. These visualisations, I hope, will help people think through, about, and with the histories of audio encoding, media shift, mail, distance, and so on. A more purely literary example would be the scholarly worth of manuscript visualisation projects (which I’ve written about before here), or of navigable large-scale text analysis sites. (For starters, try loading some texts into Voyant Links and see what you find out.) As scholars in the humanities we are able to radically expand “We write sentences. About sentences.” into a far more pedagogically and argumentatively fertile credo: “we create objects, about objects.”

The “weak” version of my response would be this. Even if what we do is and remains writing sentences about sentences, the question of selection (thank you, Jakobson) is paramount (thank you, culture wars). Which sentences do we choose to write about? And this is a question that DH can help us answer with at least a known degree of inaccuracy.

For example: in my work on Jacobite poetry, using network analysis I have been able to identify particular poems that are lynchpins of the community of Jacobite verse. On the basis of my network analysis I can make substantiated claims for the “importance” of a particular poem, substantiated by the frequency of its occurrence within the culture, by its co-occurrence in significant manuscripts, by the frequency of its distribution within a particular time frame, and so on. Without measuring and quantifying the network making up that manuscript culture, my claims for the importance of any one poem would have been more likely based on an impressionistic sense of its overall representativeness. Not that such a sense isn’t necessarily interesting or provocative; it is just less likely to be right. And I can measure the likelihood that my estimation of the representativeness of a single poem is correct or incorrect by comparing its clustering co-efficient with that of other poems. I would probably want to use a logarithmic ratio. But I have an idea of the extent to which I should go to the mat for that claim. (Answer: almost all of the way to the mat.)

So even if we reject the strong version, “Projects not Papers”, and even if we go only with the weakest defense of digital humanities that offers it the least transformative possible role in the future of the literary academy, then the digital humanities can offer us extraordinary supplements and methodological improvements through durable and measurable principles of selection. If we are resolved to write analytical sentences about historically situated literary sentences, we can at least use text analysis, corpus analysis, and network graphs to help us write about the most important sentences possible. Taken in this sense, the digital humanities can bring literary historicism closer to fulfilling its most dearly-held ambitions.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: