About the Archive

Project and Staff Information

Database as Genre: The Epic Transformation of Archives

Ed Folsom

Those of us old enough to have shopped for groceries in the early 1980s may well remember the strange sensation that genre, in its most reductive form, seemed to have conquered all. Suddenly, you walked down the aisle and, instead of the cacophony of a hundred brands, each bearing its identifying bright colors and trademarks, each arguing for its uniqueness, saw endless rows of plain white or yellow packaging, with black letters: Laundry Detergent, Beef Stew, Pinto Beans, Beer. Every week, the invasion of generic products took over a larger segment of American grocery stores. It seemed the apotheosis of the generic was on the horizon. Soon, or so it then appeared, wild variety would be tamed and we would all be buying the same plain packages. Category had prevailed; the borders were secured. I began to imagine that the generic revolution would inevitably take over the publishing world as well and that we'd soon enter a bookstore to see shelves of identical plain yellow covers with stark black titles: Poetry, Fiction, Drama, Essays, Novel.

If those generic books had in fact come to exist (and, of course, they have, even if dressed in multi-colored covers with various publishers' names on them, like Norton and Heath and Macmillan), I know how I would have found Walt Whitman. He would have been in the big yellow book with Poetry on the cover. But therein lies the problem. Our impulses always tend to funnel artists into one or another genre. Most authors work in multiple genres, but over time they get aligned with one category: not only do generic instincts pigeonhole literary works, they pigeonhole authors too. Rigidity is a quality of our categorical systems, not of the writers or usually the works we put into those systems. Most of my graduate students are still surprised to find Whitman wrote a novel and published fiction in some of the country's best journals; his stories appeared next to those of Hawthorne and Poe. Most are surprised to realize how he experimented throughout his life with mixing poetry and prose, sometimes on the same page, testing the boundaries of genre and performing typographical experiments that forced readers to engage the printed page in ways they were not accustomed to, precisely by slipping across the bounds of genre.

Even his work that we now call poetry did not settle into that category without a fight. Early reviewers of Leaves of Grass weren't sure what genre Whitman was writing in, and certainly Ralph Waldo Emerson wasn't when he wrote his famous letter in 1855 greeting Whitman "at the beginning of a great career" but never once mentioning poetry as the thing that made him rub his eyes "to see if this sunbeam were no illusion." Emerson, in fact, seemed to struggle to name what Whitman's dizzying new book was: he called it a "piece of wit and wisdom" and "incomparable things said incomparably well." It was left to Whitman, with his second edition of the book in 1856, to assign the word poem to every title in Leaves of Grass, from "Poem of Walt Whitman, an American" to "Burial Poem," and then, in his published response to Emerson, to gently chide his "master" for missing the genre by referring to his works as "poems" no fewer than seven times in his first paragraph alone. But Whitman's notebooks indicate that, as he was drafting the ideas that would become Leaves of Grass, he was entirely unsure how it would fit into a genre at all: "Novel?-Work of some sort [^Play?] . . . A spiritual novel?," he wrote, going on to describe some inchoate and absorptive work that would archive the full range of human experience:

Variety of characters, each one of whom comes forth every day—things appearing, transfers and promotions every day. There was a child went forth every day—and the first things that he saw looked at with fixed love, that thing he became for the day.-- *Bring in whole races, or castes, or generations, to express themselves—personify the general objects of the creative and give them voice—every thing on the most august scale—a leaf of grass, with its equal voice.—voice of the generations of slaves—of those who have suffered—voice of Lovers—of Night—Day—Space—the stars—the countless ages of the Past—the countless ages of the future."

Whitman, one of America's earliest huckster-authors, thought he knew how to sell his book, and one thing he needed to do was make it clear to consumers just what they were buying. If the first, 1855, edition of Leaves is the genre-bending edition, beginning with a preface that looks like prose in some ways but—with its cascading ellipses of various lengths and its lack of periods—reads more like the poetry that would follow, which, with its long cascading lines, mixed diction, and endless catalogs of the commonplace, itself reads more like some cross between journalism, oratory, and the Bible, then the second, 1856, edition is the generic one, shouting "poem" from the table of contents right through to the collection of reprinted reviews at the end. But, once Whitman claimed the genre for his work, he quickly began altering it, extending it, testing it again. He had an ongoing battle with genre. When he was toying in the 1850s with the idea of writing his own dictionary, he recorded his definition of the word genre in his notebook: "genre ja (zhän-r) peculiar to that person, period or place—not universal." Here we see clearly Whitman's discomfort with the concept, from his struggle with the pronunciation of this imported French word to the feudal mind-set that it encouraged: peculiarity to person, period, or place always leads to division and discrimination, always moves away from and against universality. Whitman's poetic project, of course, was to do the opposite—to move from a particular person, period, or place toward an absorptive embrace of all people, periods, and places. Could there be a universal genre? And, if so, wouldn't its realization be the death of genre? If genre was by definition not universal, then what would, what could, a universal genre be?

Wai Chee Dimock suggestively works with a universal sense of genre in her new book Through Other Continents, where she explores genre as a "world system." "What would literary history look like if the field were divided," she asks, "not into discrete periods, and not into discrete bodies of national literatures? What other organizing principles might come into play?" She looks to the "bending and pulling and stretching" qualities that are inherent in any generic attempt to contain and categorize, that make genre a "self-obsoleting system" because of what Wittgenstein called the "overlapping and crisscrossing" that define any "family resemblance." And genre, argues Dimock, is a kinship network, something like Gilles Deleuze and Félix Guattari's preferred image of the rhizome, that subterranean stem that grows every which way and represents the nomadic multiplicity of identity—no central root but an intertwined web of roots. Look closely again at Whitman's design of the floriated words "Leaves of Grass" on the cover of his first edition—the letters obscured with leaves and dangling roots, the title trope a continual reminder of surprising connections (leaves of grass as death emerging into life again and again), of transfer of atoms, of interpenetrating force fields. For Whitman, Eric Wilson argues, the grass is one of the "primary tropes for the rhizome," and Whitman's work—"a Manifesto of nomadic thought"—is impossible to track to the root. Instead it is casually related to a motley tangle of other work, from the Egyptian Book of the Dead to Homer to Shakespeare to Thomas Paine to nineteenth-century etiquette manuals. Emerson, always searching in vain for a category to put Leaves into, once called it "a remarkable mixture of the Bhagvat Gita and the New York Herald." Scripture and journalism, epic and etiquette manual, sublime transcendental philosophy and obscene filth.

What happens, then, when we move Whitman's rhizomorphous work into a database, put it online, allow for the webbed roots to zig and zag with everything the database incorporates? This is what we are gradually discovering on the online Walt Whitman Archive (whitmanarchive.org), which I coedit with Kenneth M. Price. Our goal when we began this project in 1996 was to make all of Whitman's work freely available online: poems, essays, letters, journals, jottings, and images, along with biographies, interviews, reviews, and criticism of Whitman. We plan to keep growing and altering the site as new materials are discovered and as we find the time and energy to follow other root systems into the unknown. Not only is Whitman's work rhizomorphous, so also is a database, and The Walt Whitman Archive is now a huge database. Our choice to try editing all of Whitman on the Web derived from our belief that, while Whitman was primarily a maker of books, his work resists the constraints of single book objects. It is impossible even to talk about Leaves of Grass as a book, since the entity we call Leaves of Grass is actually a group of numerous things—six books, three written before the Civil War and three after, each responding in key ways to a different biographical, cultural, and historical moment. Add to this Whitman's incessant revisions, many of which are scrawled directly into copies of his books, along with his array of thousands of poetry manuscripts, never gathered and edited; his letters; his notebooks; his daybooks; his other books; his voluminous journalism—and the database darts off in unexpected ways, and the search engine turns up unexpected connections, as if rhizomes were winding through that vast hidden web of circuits. We who build The Walt Whitman Archive are more and more, as Whitman put it, "the winders of the circuit of circuits," and Whitman's work—itself resisting categories—sits comfortably in a database.

Lev Manovich, in The Language of New Media, began the task of rethinking database as genre. His conclusions dovetail with Dimock's suggestion that fractals may be the most useful analogue for how to re-map genre, "a geometry of what loops around, what breaks off, what is jagged, what comes only in percentages." Fractals push us not away from the particular and toward the universal (to return to Whitman's own struggle to define the term genre), but rather toward a universality of particulars. "The fractal database," Dimock says, "thus comes as a spectrum, ranging from the microscopic to [quoting Benoit Mandelbrot] 'phenomena on or above Man's scale.'" This is how Manovich puts it:

After the novel, and subsequently cinema, privileged narrative as the key form of cultural expression of the modern age, the computer age introduces its correlate-the database. Many new media objects do not tell stories; they do not have a beginning or end; in fact, they do not have any development, thematically, formally, or otherwise that would organize their elements into a sequence. Instead, they are collections of individual items, with every item possessing the same significance as any other.
Manovich goes on to argue that "if after the death of God (Nietzche), the end of grand Narratives of Enlightenment (Lyotard), and the arrival of the Web (Tim Berners-Lee), the world appears to us as an endless and unstructured collection of images, texts, and other data records, it is only appropriate that we will be moved to model it as a database." The next step, Manovich suggests, is "to develop a poetics, aesthetics, and ethics of this database."

Database might initially seem to denigrate detail and demand abstract averaging and universalizing, but in fact the structure of Database is detail; it is built of particulars. "If fractal geometry has anything to tell us," Dimock says,

it is that the loss of detail is almost always unwarranted. . . . [T]he literary field is still incomplete, its kinship network only partly actualized, with many new members still to be added. Such a field needs to maintain an archive that is as broad-based as possible, as fine-grained as possible, an archive that errs on the side of randomness rather than on the side of undue coherence, if only to allow new permutations to come into being.
Dimock hints here at what becomes Manovich's most provocative claim:
As a cultural form, the database represents the world as a list of items, and it refuses to order this list. In contrast, a narrative creates a cause-and-effect trajectory of seemingly unordered items (events). Therefore, database and narrative are natural enemies. Competing for the same territory of human culture, each claims an exclusive right to make meaning out of the world.
What we used to call the canon wars were actually the first stirrings of the attack of database on narrative.

I have lately been reexamining Whitman's compositional techniques, now that we have gathered all the poetry manuscripts for the archive and can begin to see for the first time how Whitman conceived of the things he would come to call poems. For him, the world was a kind of preelectronic database, and his notebooks and notes are full of lists of particulars—sights and sounds and names and activities—that he dutifully enters into the record. In some manuscripts, we find drafts of poems that sound much like the published poems but contain the same lines arranged in a very different order. One manuscript of "Song of Myself" has lines that are dispersed throughout the printed poem: two lines appear on page 20 of the 1855 edition, another on page 24, one in the preface, one on page 42, one on page 16, one on page 34; another line appears in a different poem in Leaves, and yet another is part of his pre-1855 manuscript poem "Pictures." Whitman formed entire lines as they would eventually appear in print, but then he treated each line like a separate data entry, a unit available to him for endless re-ordering, as if his lines of poetry were portable and interchangeable, could be shuffled and almost randomly scattered to create different but remarkably similar poems. Just as Whitman shuffled the order of his poems up to the last minute before publication—and he would continue shuffling and conflating and combining and separating them for the rest of his career as he moved from one edition of Leaves to the next—so also he seems to have shuffled the very lines of his poems, in sometimes quite dramatic ways, right up to their being set in type. As Whitman once said, he was "always tempted to put in, take out, change" and he reserved for himself "the privilege to alter—even extensively." He was an early practitioner, in other words, of the database genre. Anyone who has read one of Whitman's cascading catalogs knows this: those catalogs always indicate an endless database, suggest a process that could continue for a lifetime, hint at the massiveness of the database that comprises our sights and hearings and touches, each of which could be entered as a separate line of the poem.

This battle between database and narrative that Manovich posits explains something about the way Whitman's poems work, as they keep shifting from moments of narration to moments of what we might call data ingestion. In "Song of Myself," we encounter pages of data entries that pause while a narrative frame takes over again, never containing and taming the unruly catalogs and always carrying us to the next exercise in incorporating detail. Henry David Thoreau struggled to articulate the tension between database and narrative when he described the experience of reading Whitman's work: "[Whitman] puts me into a liberal frame of mind prepared to see wonders,—as it were, sets me upon a hill or in the midst of a plain,—stirs me well up, and then—throws in a thousand of brick." Thoreau's description evokes Emerson's formulation of Whitman's work as a "mixture of the Bhagvat Gita and the New York Herald." The universal ("see[ing] wonders," the Bhagavad Gita) and the particular (the Herald, a thousand bricks) create the tension that Whitman sensed when he tried to define genre: the only way to represent the universal was through the suggestion of database, a thousand bricks, all the particulars with none left out.

Because photography captured these particulars, Whitman loved the medium and saw it as the new democratic art. It was the first technology that suggested database: early commentators were struck by its relentless appetite for details, for every speck that appeared in the field of vision. Many hated photography for that reason; it insisted on flaws and extraneous matter that a painter would have edited out of the scene to create beauty. But beauty, Whitman said, democratic beauty, was fullness, not exclusion, and required an eye for completeness, not a discriminating eye.

I experience this battle between database and narrative every day I work on the archive. We call it The Walt Whitman Archive, but that's a metaphor, meant to evoke the dust and texture and smell of the old books and documents themselves. The Whitman archive is, in actuality or virtuality, a database. Our database contains information from and can produce facsimiles of numerous archives; it can even reproduce a virtual single archive. Where before scholars had to travel to numerous individual archives to examine Whitman's poetry manuscripts, they are now able to access all those manuscripts from a single integrated finding guide and to display the manuscripts from diverse archives side by side, thus discovering lost connections (even reassembling notebooks that were long ago dispersed). Archive suggests physicality, idiosyncratic arrangement, partiality, while database suggests virtuality, endless ordering and reordering, and wholeness. Often we will hear archive and database conflated, as if the two terms signified the same imagined or idealized fullness of evidence. Archive and database do share a desire for completeness (though that desire can be and often is subverted by those who want to control national or institutional memories), but the physicality of archive makes it essentially different from database. There will always be more physical information in an archive than in a database, just as there will always be more malleable and portable information in a database than in an archive.

Initially, Price and I had ideas of how we would control the material in the database, and we knew the narratives we wanted to tell, the frames we wanted to construct. But the details of the database quickly exceeded any narrative we might try to frame the data with. Little roots shot out everywhere and attached to particulars we could not have imagined. Only if we insulated the narrative from the database could the narrative persist. As databases contain ever greater detail, we may begin to wonder if narrative itself is under threat. We've always known that any history or any theory could be undone if we could access the materials that the particular history or theory ignored, but when archives were physical and scattered across the globe and thus often inaccessible, it was easier to accept a history until someone else did the arduous work of re-searching the archives and altering the history with data that had before been excluded. Database increasingly makes inaccessible archives accessible from a desktop, and not just a professional scholar's desktop. On The Walt Whitman Archive, you can now place next to each other documents that previously could not be seen together. Already, notebooks that were once disbound and ended up in different states or different countries are being rediscovered, and manuscripts are fitting together like the rejoined pieces of a long-scattered jigsaw puzzle.

We are coming to recognize, then, gradually but inevitably, that database itself is becoming a new genre, the genre of the twenty-first century. Its development may turn out to be the most significant effect computer culture will have on the literary world, because literary genres have always been tools, families of technologies for exploring the realms of verbal representation as it moves from the lyrical to the narrative to the referential, from vision to action, from romance to comedy to satire to tragedy, from story to play to poem to essay, with all of the subgroups and various meldings that genre theory has spawned over the centuries. Participants in the recent American Literature Association Symposium on Biography frequently discussed how biography as a genre has managed to stay relatively untheorized, has clung to its unquestioned life-story narrative traditions, tapping into a Christological plotline involving deification of a mortal in a narrative that provides a kind of resurrection. In biography, all is sacrificed to the story of one heroic, flawed, and finally deific individual, who dwarfs everyone else. But what happens to biography when presented in the new genre, database? How does database represent a writer's life? Database biography is a genre very different from traditional narrative biography, as Price and I are discovering while we work on our biography of Whitman on The Walt Whitman Archive.

Our biography presents a traditional chronological narrative of Whitman's life and career, but the database hovers behind the biography and, as we develop it, will be made accessible with active links throughout the narrative. These links will dissolve the narrative back into the data out of which it was constructed, and the data that were left out of our particular narrative will be available to the reader as well. Each incident of Whitman's life might eventually link to previous biographies, so that readers can trace the history of how any incident has been told and embellished over the years. Each minor character, instead of staying secondary and flat, will link to biographies of that person. Links will take the user easily and quickly to the documentation that supports every fact or claim. Photographs and maps will link the user to rich environmental contextualizations that would be unwieldy or prohibitively expensive in the traditional biographical narrative (why not make available all known photographs of a writer, for example, instead of a tiny selection?).

Traditional biography grows out of archive, not database. Archive supports biography and history, but it does not become a genre, because it remains in place—difficult to access physically, often unreliably cataloged, always partial and isolated, requiring slow going. Database facilitates access, immediacy, and the ability to juxtapose items that in real space might be far removed from each other. When archive gets theorized or abstracted, it often sounds like database—some idealized hyperarchive that combines all the archives on a subject. But in reality archives are all about physicality, and such is their charm and their allure for researchers. Any of us who have spent time in actual nineteenth-century archives know the literal truth of Jacques Derrida's phrase, "archive fever." As Carolyn Steedman has argued, real archives may well produce something pathological in the researcher that might be named archive fever, because archives reify the period they record. They contain not only the records of a particular period but its artifacts as well, their dust the debris of the toxins and chemicals and disease that went into the making of the paper and glue and inks, that went into the processing of the animal skins that wrap the books we open and, in the dusty light, read and inhale. When we emerge from an archive, we are physically and mentally altered. We emerge with notes—photocopies if we're allowed—but never with the archive, which remains behind, isolated from us. Archive, if a genre, is one that only a relatively few ever read. Archive fever demands narrative as an antidote, and many of our books (and virtually all of our biographies) are tales of archive survival.

But database, as Manovich has argued, is the enemy of narrative, threatening it at every sentence, always shimmering, accessible, there. It threatens to displace narrative, to infect and deconstruct narrative endlessly, to make it retreat behind the database or dissolve back into it, to become finally its own sprawling genre, presenting a subject as it has never before been possible to present it. And, as it emerges into its own genre, database begins to reveal to us that it has been with us all along, in the guises of those literary works we have always had trouble assigning to a genre—Moby-Dick, "Song of Myself," the Bible. Dimock has examined how epic, broadly understood across cultures, is itself an unruly genre that now can be seen as an ancestor of database. Calling the epic genre "a prime candidate for fractal geometry," she finds its "linguistic fabric" to be "a rough cut, with dents and bumps, each representing a coil of time, a cystlike protuberance, in which an antecedent moment is embedded, bearing the weight of the past and burrowing into the present as a warp, a deformation." Epic loops and alters through the centuries and now survives "as a spilled-over phenomenon, spilling over into other dimensions of literature," like the novel. Or like "Song of Myself." Or, we might add, like database.

One of the most surprising realizations I've had while working on The Walt Whitman Archive is that, as it gets used, not only does our database of Whitman materials grow exponentially, so does a less visible database, the database of users. And those users cannot be corralled into a narrative either. We began predictably enough and were gratified to hit a couple of thousand users, almost all in the United States, almost all, presumably, scholars and students. But now we average around 15,000 hits a day, often spiking to well over 20,000, and our users have become increasingly international, with, over the past two months, 17,000 hits in South America, 21,000 in Asia, nearly 60,000 in Europe, and nearly a thousand in Africa. These are conservative figures, since a large number of users are not currently traceable. The archive gets a sizeable number of hits from twenty countries—from Lebanon to Brazil, Japan to Colombia—and fewer but still a substantial number from twenty others, including 1,100 from Turkey and 1,700 from India.

With this international usership, the database of users and that of materials begin to interact unpredictably. Since the site is currently entirely in English, users are at present limited by linguistic ability. But we hear from teachers in other countries who want Whitman translations included in our database. Why don't we make Leaves of Grass available in other languages, and why don't we include numerous translations from each language group so that students who speak Arabic, say, can compare Arabic translations and then look at the digital facsimiles of Whitman's original books, knowing some version of what the text means even if they aren't able to read the original? We now have editors beginning the daunting task of preparing early translations of Leaves, and the database will grow again in previously unexpected ways, and the possible narratives will increase and undermine any attempt at a grand narrative.

As Whitman has been read in other cultures and into other cultures, Leaves of Grass has become even more of a rhizomatic wanderer, looping into other traditions and finding its way back: in India, to the Upanishads and to the Bhagavad-Gita; in China, back to the foundational writings of Taoism via the twentieth-century poet Guo Moruo, who translated Whitman and rediscovered Chuang-Tzu through Leaves of Grass; in France, as an older sibling of symbolism; in Russia, as proto-socialist celebration of the proletariat. As the database grows out across national and linguistic boundaries, the ragged and rhizomatic structures of Leaves of Grass grow with it. Leaves of Grass as a database is a text very different from Leaves of Grass contained within covers, and, one senses, luckier, because database may well be epic's new genre.

This essay was the focus of a forum in PMLA, with responses by Jerome McGann, Jonathan Freedman, Katherine Hayles, Peter Stallybrass, and Meredith McGill, available in PMLA (October 2007), 1571-1579, followed by my reply (1608-1612). It is reprinted here with permission.


Published Works | In Whitman's Hand | Life & Letters | Commentary | Resources | Pictures & Sound

Support the Archive | About the Archive

Distributed under a Creative Commons License. Ed Folsom & Kenneth M. Price, editors.