Data Seeks A Story

Digitization has enabled the accumulation, storage, and manipulation of enormous amounts of data.  The numbers involved are mind boggling and we’re becoming familiar with ever larger orders of magnitude (remember when a gigabyte was a big deal?).  And we’ve been hearing similar claims long enough now that we hardly notice when someone like Google CEO Eric Schmidt tells us that every two days we create as much information as we did from the beginnings of civilization until  2003.  And, of course, we are told that the pace will only quicken and we will keep achieving ever larger orders of magnitude in data production.

So the question seems to be, what do we do with all of this data?  A good deal of it is of little or no value, and so filtering through it presents a significant challenge.  Representing data meaningfully can also be a challenge and here visualization can be quite helpful.  A couple of recent instances of visualized data come to mind.  The first is Google Lab’s Books Ngram Viewer.  The Ngram Viewer allows a user to search a database of  digitized books published from 1500 to the present for particular words or phrases.  The Viewer then generates a graph plotting the frequency with which the words or phrases have been used during a particular time period.  So for example, here is a graph tracking the occurrences in English books written between 1700 and 2010 of the names of three philosophers –  Rene Descartes, John Locke, and Thomas Hobbes:

One of the limitations of the approach comes to mind when you can’t be sure if a movement in the mentions of “John Locke” is owed to greater interest in modern political philosophy or a certain television character (or, more interestingly, both together).

Here is another graph, this one plotting the use of the words nostalgia and Nostalgia (the search is case sensitive) just because I’m intrigued by the idea:

Another recent and more elegant instance of visualized data comes from Facebook.  The graphic below was generated by potting lines representing a sampling of FB friendships.  What is most fascinating about this graphic is that no independent lines representing the continents were included, all shapes emerged from the data:

So we have two instances of data rendered intelligible, at least let us say manageable or usable. But there is still another question, what does it mean?  How do we interpret the data.  The charts and image above represent a tremendous amount of data, but what do we make of it?  That still requires judgment, context, and a story.  This is more or less the point Katherine Hayles makes in her response to Ed Folsom’s “Database as Genre:  The Epic Transformation of Archives.” Here are some comments I’ve taken from Hayles’ essay which apply rather well to both cases, especially to Google’s Ngram Viewer (keep in mind her responses are to Ed Folsom who maintains the Walt Whitman Archive online):

What it means that Whitman, say, used a certain word 298 times in Leaves of Grass while using another word only three times requires interpretation—and interpretation, almost inevitably, invokes narrative to achieve dramatic impact and significance …

These structures imply that the primary purpose of narrative is to search for meaning, making narrative an essential technology for human beings, who can arguably be defined as meaning-­seeking animals …

Manovich touches on this contrast when he perceptively observes that for narrative, the syntagmatic order of linear unfolding is actually present on the page, while the paradigmatic possibilities of alternative word choices are only virtually present. For databases, the reverse is true: the paradigmatic possibilities are actually present in the columns and the rows, while the syntagmatic progress of choices concatenated into linear sequences by SQL commands is only virtually present …

No longer singular, narratives remain the necessary others to database’s ontology, the perspectives that invest the formal logic of database operations with human meanings and that gesture toward the unknown hovering beyond the brink of what can be classified and enumerated.

In other words, data seeks a story because humans seek a story — it’s our primordial way of navigating the increasingly dense forest of data.  It is also worth bearing in mind Jerome McGann’s observations regarding databases (also in response to Folsom):

No database can function without a user interface, and in the case of cultural materials the interface is an especially crucial element of these kinds of digital instruments. Interface embeds, implicitly and explicitly, many kinds of hierarchical and narrativized organizations. Indeed, the database—any database—represents an initial critical analysis of the content materials, and while its structure is not narrativized, it is severely constrained and organized. The free play offered to the user of such environments is at least as much a function of interface design as it is of its data structure—whether that structure be a database structure or, as in the case of The Walt Whitman Archive, a markup structure . . .

Bottom line:  The interface is not neutral and for that matter neither is the data because it has already been tagged and marked up in certain way when the database architecture was designed and the information entered in accordingly.

If databases and interfaces that give us access to the immense amount of information being digitized are going to be useful to us, we need to make sure we understand the embedded limitations so that these limitations do not become immense blind spots for us as we try to do what we must always do with information — make a story out of it.  And the making of the story, a basic human drive, requires an awareness of context, judgment and discernment, and a certain wisdom that, as of yet, the database and clever, even elegant, means of representing the data stored in them, are not by themselves going to bring to the task.  It may be worth remembering the old adage, information is not knowledge and knowledge is not wisdom.

Theory and Electronic Literature

What is theory?  It is, at least in part, an effort to make explicit what is implicit and to expose what is assumed but unspoken. To paraphrase Niklas Luhmann, theory seeks to perceive the reality which one does not perceive when one perceives it.  If so, then Katherine Hayles is offering us a theory of electronic literature that renders electronic literature an embodiment of theory.  Not, however, in the sense early theorists of hypertext imagined, but in a way that is more consonant with the anti-speculative perspective advocated by Jerome McGann in Radiant Textuality: Literature after the World Wide Web.

Early hypertext theorists such as George Landow enthusiastically read electronic literature as an embodiment of poststructuralist theory.  It appeared that electronic literature manifested on the surface everything Roland Barthes painstakingly sought to reveal about traditional literature and the author.  This reading of electronic literature, however, appears now to have been stillborn because of its identification of the hyperlink as electronic literature’s “distinguishing characteristic,” a move which Hayles shows was beset by serious problems.  (Electronic Literature: New Horizons for the Literary, 31)  Likewise, McGann seems intent on moving past this mode of theorizing.  As I read him, it is not so much theory itself which McGann repudiates, but a certain way of constructing and applying theory.  As I’ve noted in response to radiant textuality, theory for Mcgann is best aligned with the kinds of knowledge that arise from performance/deformance because Mcgann envisions theory as poiesis rather than gnosis. The kind of embodied knowledge or theory that arises from acts of making or performance “makes possible the imagination of what you don’t know” because it elicits knowledge from failure and also from serendipity. (radiant textuality, 83)

McGann’s notion of imagining what you don’t know recalls Hayles’ clever appropriation of Rumsfeld’s Zen-like categorizations of knowledge.  As she puts it, “I propose that (some of) the purposes of literature are to reveal what we know but don’t know that we know, and to transform what we know we know into what we don’t yet know.”  Further resonating with McGann, she sees literature achieving this knowledge by “activating a recursive feedback loop between knowledge realized in the body through gesture, ritual, performance, posture, and enactment, and knowledge realized in the neocortex as conscious and explicit articulations” in much the same way that McGann sees theory arising through “the kinds of knowledge involved in performative operations.”  (Electronic Literature, 132; RT, 106)  In one sense I would argue that as Hayles describes the effects of electronic literature it functions in a way reminiscent of McGann’s practices of deformance.

For both Hayles and McGann, theory is bound up with the body, with the material, and with action.  For McGann, deformance is a practice which leads the reader to tap through performance the kind of embodied knowledge that helps them reckon with the materiality of the text.  For Hayles, electronic literature already requires or is intended to force the kinds of interactions that deformance is attempting to artificially elicit in the context of print.  Put otherwise, the productive disruptions code introduces into narrative awaken us, according to Hayles, to the reality of the human life-world’s integration with intelligent machines in much same way that the disruptions of deformance awaken us to the material realities of the text according to McGann.

Electronic literature as Hayles’ theorizes it draws into the open features of human existence that previously lay below the level of awareness.  At its best then electronic literature, like good theory, reveals what is not always perceived, but always present.  And this, according to Hayles, it accomplishes by “creating recursive feedback loops between explicit articulation, conscious thought, and embodied sensorimotor knowledge.”  (Electronic Literature, 135)  Or, as McGann might put it, through poiesis and not merely gnosis.

Deformance, Materiality, Theory

In radiant textuality, Mcgann introduces the practice of deformance by suggesting that it will help “break beyond conceptual analysis into the kinds of knowledge involved in performative operations.” (106)  As I understand him, Mcgann is reminding us that there is a kind of knowledge that cannot be “cast in thematic form.”  (106)  Interpretation that seeks to extract a meaning in thematic form assumes that meaning is something contained by the work of literature, let us say a poem for example, which can be sucked out and recast otherwise.  There is a place for such interpretative strategies, but they tend to ignore the knowledge embedded in the textuality/materiality of the poem that is acquired not by ratiocination but by performance.  This knowledge by its nature is difficult to articulate, but it is the kind of knowledge that deformance seeks to tap and the language of “intimacy” with a text gets at the sensuality of this knowledge.  This knowledge arises from the forming of the words with our mouths, the seeing of the words on the printed page (a seeing that is sharpened by distortion), the hearing of sounds made by the words (again sharpened by distortion), and even feeling the weight of the book in the hand.  To read the works backward or to displace or isolate elements of the text is a way of helping us see what ordinarily “escapes our scrutiny” (116) because with most interpreters we locate meaning not in the documentary features of the text but in a linguistic event. (115)  Mcgann’s dedication to John Unsworth, Deo Gratias, invites me to understand this almost as a kind of liturgical knowledge.

Referring to Dickinson’s backward reading as an example of the kind of performance that elicits this intimate knowledge, Mcgann suggests that this model is in a sense “antitheoretical: not because it is opposed to theory (i.e., speculative thought) but because it places theory in a subordinated relation to practice.”  (109)  This connects back to his earlier discussion of the “pragmatics of theory.”  (83)  There rather than equating theory with speculative thought, theory is distinguished from “hypothesis or speculation.”  Theory in this earlier case then is not what is being opposed by deformance when it is labeled antitheoretical.  Theory here is aligned with the kinds of knowledge that arise from performance/deformance because Mcgann envisions it as poiesis rather than gnosis.  This embodied knowledge that arises from making or performing “makes possible the imagination of what you don’t know” because it elicits knowledge from failure and also from serendipity.  (83)

My sense is that Mcgann sees this kind of knowledge that emerges out of the work of building something like the Rossetti Archive as a critical step toward envisioning what the digital humanities can be, but that he also is concerned that digitized texts will not be able to attain the capabilities of critical interaction with the materiality of our texts.  Or perhaps to put it differently, he is most concerned that forgetting what we have learned over centuries of interaction with material texts, we will fail to incorporate that knowledge into our emerging digital tools.