Tag Archives: dh

Surfaces. About Surfaces

Claude Willan’s recent piece—“We write sentences. About sentences.”—has been rumbling through the spin cycle of my brain. The problem, as usually with Claude, is that he’s right and yet I disagree with him.

The climax of Claude’s argument comes in the following formulation:

DH allows us to create objects about objects.

Or, rather, to zoom out just a little:

[A]t its most generative and plural, DH allows us to create objects about objects.

As you might expect, given the structuring logic of this paragraph, my disagreement is going to land somewhere between zoomed-in and zoomed-out. But first, to gloss his idea somewhat, because it’s really tremendously productive. To say that DH creates objects about objects is to imply three suggestions about the future of the field:

  1. That the digital project—the “object” that is created—should be acknowledged as a serious mode of scholarship and respected in tenure review, hiring, and so forth.
  2. That the “objects” that we make (as well as at least some of the “objects” that we study) might be encountered productively through the rich lens of object studies. (We need to think still more, I think, about how scholarly databases, visualizations, and other such productions might be agential. How they work as “mediators” in Latour’s sense, transforming the data they represent.)
  3. That “aboutness” is a relation that one object can have to another. The close of his piece, as I take it, maps out one type of aboutness—a network analysis that demonstrates that one poem is more central to Jacobean social networks than another and therefore more likely to be historically important (in whatever sense we care to intend) than it.

(I’m with him on 1 & 2, and likely to roll my eyes at 3, even as I eagerly take his results at face value.)

But I want to articulate an alternative model of a plural and generative digital humanities, one that I think has a longer history within literary studies than we’ve been willing to address. A tentative formulation:

“We make surfaces. About surfaces.”

*    *    *

First, an object, or rather part of one.

<div1 xml:id="Son"> 
<div2 xml:id="Son-001" n="1"> 
<milestone unit="line" xml:id="Son-001-01" n="1" corresp="#w0000010 #c0000020 #w0000030 #c0000040 #w0000050 #c0000060 #w0000070 #c0000080 #w0000090 #c0000100 #w0000110 #p0000120"/> 
<w xml:id="w0000010" n="1.1">From</w> 
<c xml:id="c0000020" n="1.1"> </c> 
<w xml:id="w0000030" n="1.1">fairest</w> 
<c xml:id="c0000040" n="1.1"> </c> 
<w xml:id="w0000050" n="1.1">creatures</w> 
<c xml:id="c0000060" n="1.1"> </c> 
<w xml:id="w0000070" n="1.1">we</w> 
<c xml:id="c0000080" n="1.1"> </c> 
<w xml:id="w0000090" n="1.1">desire</w> 
<c xml:id="c0000100" n="1.1"> </c> 
<w xml:id="w0000110" n="1.1">increase</w> 
<pc xml:id="p0000120" n="1.1">,</pc>

This is a small snippet of the XML code that organizes the Folger Digital Texts edition of The Sonnets, encoding the very beginning of Shakespeare’s sequence. Even without much experience with encoded texts, one can make out a careful structure here in the patterns of anaphora and parenthesis. The code marks the beginning of the document, of the Sonnets text, and of the first sonnet, each enclosing the next like nested baskets. Then, it defines a new unit, a “line,” and lists the units that comprise it. Finally, each word, punctuation character, and space is given its own address. The code carefully recreates our mental model of a book composed of poems themselves composed of lines and so forth.

When we look at the resulting text on the website, it bears a close resemblance to the Folger* edition on which it is based. “From fairest creatures we desire increase,” we read, in familiar Times New Roman. Yet for this purpose, all that structuring hierarchy was unnecessary: one need not name every comma and space to display a line of text on screen. Rather, structuring information is encoded into the Sonnets so that we can read against and around it: we might ask for every line containing the word “desire,” the first word of every line that isn’t preceded by a comma (a punctuation character), or a list of the words in the document organized by frequency. (The first and third of these are already possible using the Folger Digital Texts site.) We might readily enough display the poems in reverse, or in the orders suggested by Brents Stirling, Denys Bray, or Arthur Acheson. Or, imitating Raymond Queneau, we might explore the billions of poems emerging when the first line of one poem is combined with the second of another, and so forth.

The object here—the Folger Digital Texts edition of Shakespeare’s Sonnets—enfolds into form a host of possibilities that were already present in the sequence. What we realize, looking at this encoded sonnet sequence is that structure (or form) is not the opposite of deformance, distortion, and rearrangement—it is its precondition. Without knowing what a line is, implicitly or explicitly, we can’t refer to a certain line and certainly can’t rearrange it. For an example, imagine a second encoding scheme organized not around the units of the Folger text (sonnet, line, word, character, punctuation) but some other constructs that would offer totally different affordances: say, syllable, phrase, quatrain or couplet, and poetic grouping according to the scheme of Neil Rudenstine. We could then algorithmically call up the closing phrase of each quatrain of the Rival Poet poems. Or substitute the closing couplets of the Young Man sonnets into the Dark Lady poems. Again, the categories that order the code enable what can be done with it. Someone interested in the material book might prefer a third schema—say, edition, copy, gathering, page, line, mark—and its own possibilities for reference and rearrangement. A given text admits many potential organizational schemas, and, in structuring a text, these models allow for its manipulation. This, I think, is the key insight of Michael Witmore in “Text: A Massively Addressable Object”: that what is fundamental about our conceptual engagement with texts is our ability to formulate many such schemas and to use them to engage with textual specifics at many levels of scale.

For Witmore (and, in a different sense, for Willan), what is useful here is the ways that massive computational power will allow us to use our encoded structures to test our concepts against the data. Willan writes that the digital humanities offer “extraordinary supplements and methodological improvements through durable and measurable principles of selection.” We might read, he suggests, “the most important sentences.”

But when we read recent monographs and journals, it seems clear to me that there’s another sort of argument that DH is teaching us to build, one that is not at all concerned with “methodology” as something that can be improved or “principles of selection” that can endure. A type of argument by juxtaposition, comparison, analogy, and suggestion, made possible by deep databases and long lists of searches. Some of our best critics work by collage, holding up fragments in a web of their own making that transforms the ‘objects’ of study as much as it illuminates them. Or, to try for a less badly mixed metaphor, I love it when critics offer me hand-written maps of impossible locales, detailing possibilities that weren’t or that are not yet.

You can probably hear that I’m borrowing from Johanna Drucker and Eileen Joy here—but there’s actually a long history of scholars of electronic textuality thinking less about knowledge-as-important-fact than how we choose to arrange and derange the infinitely-arrangeable text:

“[A] cybertext is not a static artefact. The conditions of its existence are tenuous, a feature that can be exploited for aesthetic purposes through continual morphing. If a text is in a profound sense produced in each reading, and if no text is ever fully self-identical, then cybertexts embody those premises with even greater flexibility.”Drucker, “Poetics of Electronic Textuality”

Always becoming, cyberpoems are emergent, heterological and heterogeneous in their constant spooling, transferences, hyperlinking and recomposition. the poem has shifted from bricolage to morphosis…. Made of textual typographic fragments constantly moving into and out of focus, resolution and degrees of proximity, the cyberpoem is more like an installation or event than a document etched in metal or printed on paper. The reader navigates through a sea of signs visiting information ports. There is no horizon line and any scratched in reference to one is nostalgic since we see beyond what the naked eye can see via satellites, microscopes, cable and data mirrors. Brereton, “The CyberPoetics of Typography”

Our scholarly inventio and dispositio, at this moment, at least, is mirrored in our artifacts: arrangement is both arbitrary and essential, an attempt to model in flat text our glimpses of morphosis. We read less the text than the work surface that we assemble around and with the text. Or rather: we make surfaces about surfaces.

And if we follow this line of thought backwards, from electronic text to cybertext to cyberpoem, we end up not with computers at all, but with form, that precondition of all disruption.

Writing’s forms are not merely shapes but forces; formal questions are about dynamics—they ask how, where, and why the writing moves, what are the types, directions, number, and velocities of a work’s motion. Hejinian, “The Rejection of Closure”

The unity of a work is not a closed symmetrical whole, but an unfolding dynamic integrity. . . . The sensation of form in such a situation is always the sensation of flow (and therefore of change)…. Art exists by means of this interaction or struggle. Tynianov, cited in Hejinian

Today’s criticism is always a scrambling and re-arranging of yesterday’s. Here, an account of the unfolding of a text’s form in time becomes spatialized in Hejinian, who is interested in the dynamics of field composition. Brereton allegorizes, turning the movement of characters on screen into a figure for such dynamics. And Drucker is one of many critics who abstracts this move once more, from typography to identity. What these accounts share is a sense of reading as navigating a constantly shifting field of significance, simultaneously prompted by the text and invented by the reader.

This is, I think, a fair shorthand for many of our critical practices now. We have searches open in tabs, database queries, poems, all arrayed on the darkling plains of our minds and we start to draw connections, start to assemble not objects exactly but fields of meaningful relation that we can sketch out again for others. Richard Blackmur writes, on another topic:

No one can improve upon the accidentally established order we possess; but everyone can invite himself to feel the constant interflow of new relations, of new reticulations–as if the inner order were always on the move…

This “constant interflow of new relations, of new reticulations,” is to me the great pleasure and achievement of criticism. Blackmur, of course, is talking about that early cyber-text, Shakespeare’s Sonnets, and its ability to constantly shift under the pressures of reading and interpretation, taking on new forms and new orders in response to our gambits about its structure and meaning. From a 21st-century perspective, Blackmur’s is just about the final word on Shakespeare studies’s own critical history of rearranging and restructuring, distorting and deforming, morphosis and bricolage. From Charles Knight in the 19th century, or even John Benson in the 17th, Shakespeare’s sequence has provoked editors and readers to reimagine it, moving, adding, deleting, and combining poems.

Reading descriptions of these rearrangers’ processes–spreading the poems out on a desk or table or floor and starting to trace connections—I see shades of Brereton’s cyber-reader, trying to bring order to a constantly shifting morphosis. I see Bruce Andrews’s active electronic reader:

We can think of the textual surface as an instrument panel, the screen as a flat & opaque workspace, given enormous fluidity, activating the user’s body. Action replaces both the passive representation of conventional literature & the passive spectacle of animated, programmed work. It embraces navigation, micro-evaluations, conceptual animation, freeze-framing, editing, blending, filtering, subliminal cut & paste, time compressions & expansions, frame resizing: practically everything we need to sidetrack closure.(from “Electronic Poetics“)

I see myself, cutting and pasting to see how one poem, one argument, one idea sparkles in the light of another. I see the invention and arrangement of an endlessly productive surface.

Let me be clear: these re-arranging critics have failed from the point of view of history, unable to convince us of their evidence, much less their claims. Their names (Arthur Acheson, Samuel Butler, Clara de Chambrun, and so forth) no longer appear in our editions. Their work now seems out-dated, responsive to historical speculation we find absurd and ethical concerns we no longer share. In the absence of evidence, we’ve largely given up the questions they’ve tackled as impossible (when we’re feeling charitable) or uninteresting (when we’re not). But, then, this will happen to most of us, too.

What remains, what attracts my eye when I’m reading their work and what I’m laying out on the table for you is this: their misguided work of ingenuity really does ignite the poems with new possibilities. Their criticism, by all standards of contemporary discourse, is flawed, suspect, unreadable, unimportant.

And yet it moves.

Thanks to Laura Kolb for comments on a draft of this post.
*An earlier version said “Signet,” because I got my editions mixed up. Thanks to Sarah Werner for pointing this out!