by Andrew Piper
[FIGURE 7.2] An image of the interface FeatureLens, taken from Tanya Clement, “‘A thing not beginning and not ending’: Using Digital Tools to Distant Read Gertrude Stein’s The Making of Americans,” Literary and Linguistic Computing 23, no. 3 (2008): 364. By permission of Oxford University Press.
In John Mohr and Vincent Duquenne’s work on the history of poverty, by using quantitative techniques they were able to show how changes in the discourse of poverty—the words used to describe it—corresponded to historical changes in the institutional practices used to address it.26 Examining the clusters of words related to poverty (needy, distressed, worthy, indigent, destitute) and the actions to which they corresponded at two major turning points in the history of social welfare (between the end of the poorhouse era in the 1880s and the beginnings of the welfare state in the 1910s), their work is aimed at helping us understand the rationale behind changing types of institutional relief programs. Their research reveals a historical correlation between words and deeds, not only how a given moment understands what it is doing, but how words also help produce those actions.
Finally, in my own project that relies on computational textual analysis, I am interested in creating literary “topologies,” or historical maps, of literary networks. The first suite of maps will be called The Werther Effect, based on Goethe’s novel The Sorrows of Young Werther (1774), one of the most popular literary works of the eighteenth century (often referred to as the first European “best seller”).27 While we have a number of bibliographies that list the many adaptations of Werther that followed in its wake (over two hundred by some estimates), we still have very little sense of the extent to which Werther fanned out into the world of eighteenth-century letters more broadly. If Werther was, as one critic has put it, a “syndrome” of the emerging middle-class commercial culture to which it belonged, I want to know more about the reach of the text’s pathos.28 The question is not simply to what extent the novel infected the language of its age, but more challengingly, in what ways did it do so—what was the nature of Werther’s influence, in terms of its language and style, that suffused the closing decades of the period known as the Enlightenment?
To try to answer these questions, my collaborator, Mark Algee-Hewitt, and I are using statistical models to compare lexical “thumbprints” of Werther with other eighteenth-century works and then projecting those relationships as a network map (fig. 7.3). In measuring the prevalence of the most common words in Werther—its lexical identity—within other works from the period (eventually about five thousand), we can see where a work like Werther “goes” and the types of clusters or new arrangements of texts that it helps produce. We can observe a kind of structuring effect that Werther’s language has on the period. But we can also gain a better sense of how it travels, the specific aspects of Werther that contribute to these new textual communities. By focusing on smaller units such as the page, the paragraph, the sentence, or even the phoneme or morpheme (such as prefixes and suffixes), we can zoom in to see which aspects of Werther are doing the work of bringing certain groups of texts together. But we can also turn the question around and ask where a work like Werther comes from. If the novel supposedly undergoes a “rise” in the eighteenth century, where did it originate? Is it more or less like other novels, newspapers, biographies, or philosophy? How “new” was it? Whatever the answers may be, these are questions we simply cannot reliably answer by hand (though we have tried).
The kind of work I’ve been describing here, and again there is much more of it, largely falls under the heading of “distant reading.”29 In place of the “close reading” that analyzed a single work in great detail, the former bedrock of literary criticism, we are now reading a great many works together (or not reading them, as both detractors and promoters alike are fond of saying). But on another level, we could say that distant reading is in practice just another form of close reading, perhaps the closest kind. It is intensely word based, far closer to older reading practices like medieval gloss, which attempted to explicate a text word for word. As many have pointed out, it is indebted to textual instruments like the concordance, which was first used in the thirteenth century and which arranges words from a larger corpus (initially the Bible) into an alphabetical list.30
[FIGURE 7.3] Andrew Piper and Mark Algee-Hewitt, The Werther Effect (2011). Presented is a topological map of Goethe’s collected works, arranged in their relationship to his youthful novel The Sorrows of Young Werther. The map is color-coded according to genre, with novels and autobiography located mostly at the top, dramatic works to the lower right, critical essays making a streak down the middle, and his science writings comprising the far left. The further apart any two tiles are, the less alike they are, and the smaller the tile, the more it resembles those works directly surrounding it. Werther is the highlighted tile in the upper right corner. The map is reproduced courtesy of Mark Algee-Hewitt.
These projects are premised on the idea that reading “just the words” is not a primitive model of how we read, but a fundamental one. The attention to individual words is driven by the idea that our experience when we read is at its most meaningful prior to any theoretical abstractions that we produce through these words (as they coalesce into things like plot, character, or the text’s “message”). Words are what our minds embrace, follow, puzzle over. They are the clay, the raw material, the pleasure of reading. But such distant readings are also a means of understanding the meaning of linguistic recurrence, how words repeat themselves at a distance and how these gradual accumulations of repetition are essential for the way texts accrue meaning. Such distant readings are in many ways closer in spirit to the poetic device of the refrain, one of the oldest tools for lending words meaning. When Edgar Allen Poe was deciding upon the feature that would make “The Raven,” which would subsequently be considered one of the most important poems in the English language, universally accessible he chose the single word refrain, “Nevermore.”31 Understanding the differences within language’s repetitions is one of reading’s backbones.
For all of this newfound idolatry of the word, however, these types of projects also argue for a kind of reading that is deeply visual. We look at a topology or a graph in order to gain a sense of the whole. However visual our relationship to reading books may be, unlike a topology a book can never be read all at once. This is the point of the haunting images by the visual artist Idris Khan that consist of the pages of books superimposed onto one another and that are illegible. A book insists on either sequence or slice. A book takes time. The computational interface, by contrast, tries to give access to a totality, to present sequence as slice. It recalls reading’s other etymological origin, not as a riddle, but as a form of agricultural harvesting or gathering together. As the father of hermeneutics, Friedrich Schleiermacher, argued at the turn of the nineteenth century, “Without the whole no true understanding is possible.”32
As a consequence, computational interfaces like the topology immerse us into a world of likeness rather than one of distinction. They ask us to think about words and works in relation to one another. Where books are tools of distinction, “difference engines” as I called them in the last chapter, topologies teach us to understand the meaning of connectivity, of the “next to.” They place us in a more critical relationship to the “network” as one of the dominant figures of contemporary thought. As John Cayley intuited, language always goes overboard; its meaning runs past the artificial boundaries we establish to contain it. Distant reading is the attempt to hang on to the ballast of words against which we as readers inevitably, frustratingly, and at times joyfully founder.
. . .
At the opening of the sixteenth century, the great Renaissance man of letters Erasmus of Rotterdam published a new version of the New Testament. In two separate columns, he arranged the earliest Greek sources alongside Jerome’s Latin translation, known as the Vulgate Bible and the official edition of the Catholic Church. Erasmus’s edition, as one might imagin
e, was controversial. But in many ways he was following the lead of his fourth-century predecessor Jerome, who had returned to the Hebrew sources for his translation of the Old Testament. And Jerome, for his part, was following the lead of the third-century Egyptian scholar Origen, whose Hexapla, a six-column comparison of the Hebrew and Greek sources of the Bible, had served as an important basis for Jerome’s translation.33
Whether it was Erasmus, Jerome, or Origen, each of these scholars was taking apart an earlier text and putting it back together in new ways for the purpose of reading it in new ways. In the nineteenth century, the philologist Karl Lachmann would begin to do the same for the origins of Germanic literature. Undoing and redoing a rich vernacular heritage, he gave birth to what has come to be known as “the critical edition,” the foundational object for all professional reading.34
We are at a similar moment in terms of creating new textual instruments today. Current anxieties about the meaning of computational interfaces are no different than the controversies that surrounded the biblical translations of Renaissance humanists. Erasmus had provocatively entitled his edition Novum Instrumentum, not Novum Testamentum, a new instrument, not a new testament. For Erasmus, the book was indeed an instrument, not just a “mere tool.” Where some readers were shocked to encounter his edition rather than Jerome’s, so too are some readers today just as shocked to see their beloved Jane Austen heaped onto a giant pile of books and run through the mill of data mining. We continue to struggle with the idea of writing as an instrument and not as a testament.
Such computational reading is in large part a response to the growing quantity of things to read. There has, of course, always been too much to read. But with the digital preservation of Twitter (177 million tweets per day), Google Books (15 million books), or the Internet more generally (who knows), we have entered a new order of magnitude of “too much.” To read all of this, to preserve any sense of synopsis that Schleiermacher had suggested was the very condition of understanding, we will need new instruments and new methods of reading. The book isn’t enough.
It is, in my view, an exciting time. Whether as professionals or amateurs, there is an opportunity here for readers to get their hands dirty, to reimagine the shape of our reading instruments, like Peter Organisciak’s recent TAToo (for Text Analysis Tool), which can be imbedded alongside a text in your browser so that you have a running word cloud, concordance, and list of collocates for whatever you happen to be reading at the moment. It is the latest example of what Stéfan Sinclair calls “ubiquitous analysis.”35 If we are going to have enhanced books that distract us with animation and soundtracks, we might as well have ones that help us think more analytically about what we’re reading, too.
As I’ve tried to argue throughout this book, the point of such work is not to “overcome” older ways of reading. Reading can never be progressive. But it can show us how computational reading is different from book reading, how we can do different things with these different instruments, one no better than the other. The categories that I have laid out above—the lexical, visual, synoptic, and relational emphases of the computational interface—are a starting point for thinking through how computation will impact how we read. Perhaps the term we need then is not distant reading, but multiple reading—the way computation stacks different types of reading on top of one another. This too has its origins in book culture, not only in its indebtedness to the shape of the pile or stack, but in the way one of the book’s historical strengths as a technology has been its ability to be used in so many diverse ways. We don’t need more reading, just more kinds of reading. As in an ecosystem, diversity is a sign of a system’s health.
In my preface to this book I mentioned that when I was a child I went to computer camp. While there I was taught to read the language of a machine and understand its logic. I have since forgotten much of what I learned, but I am now in the process of relearning it, bit by bit (so to speak). Many years later, while in graduate school, I went to book camp, actually the Rare Book School at the University of Virginia, one of the premier centers for the study of the book. It was there that I learned to read the codes of how books work. In many ways it was surprising, but also refreshing, to know more about the instrument that I spent so much of my time unconsciously using. It was also the beginning of my understanding of the extent to which our digital future is indelibly linked to our bibliographic past.
When I finish writing today, I will pick up our children from school. Today is my turn for the whirlwind of packing, zipping, searching, undressing, and cooking that marks the end of “work.” My son is now completing first grade, and over the time it has taken me to write this book he has learned how to read. Like millions of other children his age, his mental universe has been permanently reshaped. He will never look on the world in quite the same way. And yet this is still not true for everyone. Whether because of disability or disadvantage, not all children (or adults) have equal access to reading. According to the United Nations, close to one-fifth of the world’s population cannot read. In the United States, 13 percent of school-age children have some kind of learning disability that negatively impacts their reading. And a growing portion of the adult population is falling behind in terms of functional literacy.36 Reading is not, and has never been, universal. Part of the aim of books like this one is to remind us just how much work reading requires to sustain itself.
When my son comes home today, he will play with the computer. Then he will go and do his homework, where he will use books for his reading, writing, and math exercises. My daughter, who is still in preschool, will simulate this process in reverse, playing with her notebooks, while the computer is still very much work for her. My hope is that these two categories, work and play, will remain as interwoven throughout their lives as the instruments that they use to engage in them, the book and the computer. I hope they are afforded the advantages of both, and that they pass on those advantages to others. My real hope, though, is that when it comes time to learn how these two very different instruments work (and play), I can send them to just one camp.
EPILOGUE
Letting Go of the Book
What? Still amusing yourself with a book?
This isn’t Sunday, you know.
MARCEL PROUST [in search of lost time]
Books are never finished.
PAUL AUSTER [red notebook]
What would the world be like without books? It turns out this is a very old question. We have already seen many examples over the course of this book. In the third century, the Egyptian scholar Origen created a six-columned reading device, a superbook, for comparing the Hebrew and Greek sources of the Bible. In the thirteenth century, the Majorcan scholar Ramon Llull imagined the book first in the shape of a tree and then as a series of spinning discs. In the nineteenth century, the German satirist Jean Paul conceived of the book as one long, unfurling piece of paper consisting of a single line of poetic prose (or prosaic poetry). He said it would make no small impression were it attached to the back of a child, like angel’s wings, or a juvenile sail made from diaper cloth.1
At the end of the nineteenth century, the French symbolist poet Stéphane Mallarmé suggested that “all earthly existence must ultimately be contained in a book.”2 Neither many columned nor extremely long, the book was to be as large as the entire universe. In the twentieth century, the Russian modernist filmmaker Sergei Eisenstein imagined a spherical book, the only proper form for his critical writings on cinema. There has even been an idea for a “book in a can,” an inspired student project of a scroll stuffed into a metal container now preserved in the rare books collection at McGill University. Ever since its inception, it seems, we have been dreaming beyond the book.
In 2005, as part of an installation that would later show at the Guggenheim Museum, the acclaimed Palestinian artist Emily Jacir shot one thousand books using a .22 caliber pistol. Material for a Film, as it was called, was designed to commemorate the assassination of the Palestinian in
tellectual Wael Zuaiter, who was killed by Israeli intelligence agents in reprisal for the slaying of eleven Israeli athletes at the Munich Olympics in 1972. At the time of his death, Zuaiter was carrying a copy of A Thousand and One Nights, which he was translating into Italian. Shot thirteen times, one of the bullets hit his copy of the book. The book now resides in the Wael Zuaiter Center in Massa Carrara, Tuscany.
In Jacir’s installation the book proved to be an affecting symbol for the defenseless scholar caught up in a world of violent exchange. It was a poignant reminder of Heinrich Heine’s famous remark that where books are burned, people will soon follow. But Jacir’s work was also part of a larger wave of contemporary projects that were performing aggressive, even violent, acts toward books.3 Cutting, drowning, soaking, unfurling, piercing, and shooting books have been some of the many ways that artists like Jacqueline Rush Lee, Jonathan Latham, Robert The, Cara Barer, and Sam Markham have over the past decade or more been enacting a collective sense of the book’s immanent demise (fig. 8.1). If we have forever been imagining our way past books, we have more recently begun to think about what it would be like to live in a world without them. We have begun the work of bibliographic mourning.