Navigating in the Archive: Orientation, Disorientation, Reorientation

Navigation is first of all a labor of displacement. —Michel de Certeau

While the non-hierarchical or decentered structure of the archive reflects the fragments' irreducible singularity and insusceptibility to collection in a "book," the archive's system of nonlinear links reveals, on the other hand, their openness to and participation in multiple textual constellations and/or contingent orders.

In his recent book Open Sky, Paul Virilio calls our attention to the concept of trajectivity: "It seems we are still incapable of seriously entertaining this question of the path, except in the realms of mechanics, ballistics or astronomy. Objectivity, subjectivity, certainly, but never trajectivity." Paul Virilio, Open Sky, translated by Julie Rose (London: Verso, 1997), 24. The hypertext archive offers a new site for an exploration of trajectivity. At times, the close study of elements and attributes within a given document will delimit the trajectory of an exploration. At other times, the exploration of document constellations and especially the relations among the fragments and the texts in which they reappear creates a dialectic process leading the reader toward the discovery of new patterns. And, at still other times, a reader may act as a "shifter," cutting his or her way across the various levels of the archive by following a single tag or a combination of tags (that is, documents with canceled faces; documents written in ink; documents written on envelope fragments, etc.).

In the archive, theories of navigation are subject to change during the course of a research session as a result of unforeseen discoveries or new questions. The number of tags, types, searchable fields, and links is finite and editorially determined; the number of paths that can be traced through the materials, however, is almost limitless—or, rather, limited only by the reader's willingness to track individual attributes and elements and to collate search results, or by his or her imagination of virtual itineraries. In general, the best readers of Dickinson's fragments do not read linearly, but recursively, finding, often by losing, their different ways through the materials of the archive.

The paths traced by the reader through the electronic archive are the "traveler's tales" that Virilio fears will be lost, along with "the possibility of some kind of interpretation," on the information superhighway. Paul Virilio, Open Sky, 25. To be sure, there are risks. As Mireille Rosello writes, since the "screener's navigation. . . .[is] both read and written at the same time. . .meaning. . . [becomes] fragile, easily destroyed, [almost] impossible to record." See Mireille Rosello, "The Screener's Maps: Michel de Certeau's 'Wandersmänner' and Paul Auster's Hypertextual Detective," in Hyper/Text /Theory, edited by George P. Landow (Baltimore and London: Johns Hopkins UP, 1994), 121–58. Yet however fragile, illegible, or quickly erased these virtual itineraries (the inscription of which begins the moment the reader enters the archive, and ends only arbitrarily when he or she exits) may be, they are also new signifying (critical, creative) narratives that we cannot ignore and that require a new theory—a new method—of (speed) readingwriting (screening).

The power of the archive is above all the power to induce a departure: "And I wonder if it may not be desirable at this point to experiment widely with disorientation rather than safety." See Mireille Rosello, "The Screener's Maps: Michel de Certeau's 'Wandersmänner' and Paul Auster's Hypertextual Detective," in Hyper/Text /Theory, edited by George P. Landow (Baltimore and London: Johns Hopkins UP, 1994), 121–58.

Search: Document Features

The decision to transcribe all of the texts on a given document runs counter to the conventional editorial treatment of Dickinson's late fragments. In Thomas H. Johnson's Letters (1958), for example, apparently discrete text fragments inscribed across a single document are removed from their material context, divided between two editorially determined categories ("To Unidentified Recipients," "Aphorisms"), and then printed alphabetically by first line in a numbered series that no longer reflects their relationship either to the document on which they were inscribed or to one another. Respecting the integrity of the physical document and carefully mapping its zones (folds, seams, etc.) makes possible analyses of the relations between and among the texts each document carries. Moreover, since the document is often a metaphorical as well as an actual container for thoughts—an envelope shaped like a bird carries a text about flight; an envelope seal becomes the space for a meditation on secrecy; the two sides of a document are inscribed with rhyming texts; a torn edge corresponds to a textual verge, etc.—maintaining the integrity of the physical document facilitates further investigation into the relationship between Dickinson's medium and her messages.

The "Index of Documents Carrying Fragments" lists the core documents in the archive by catalog number and offers thumbnail images of the first page of each document.

The "Library of Search Paths (Browse Documents)" section allows further exploration of documents by feature.

Text Types

The analysis of Dickinson's late manuscripts brings to light a considerable number of text types: in addition to poems and letters there are fragments, notes and ideas toward poems, working notes and phrases, trial lines, and variant lines. The need to account for all of the possible texts that emerge during the transcription of a given document has been accommodated by the use of the DIV2 element. Each DIV2 identifies an apparently discrete text, which is in turn assigned a "type." The "typing" of Dickinson's late texts is a matter of interpretation (and debate); the types listed below should be treated as a "working list," rather than an exhaustive or definitive list. Many of the fragments are too brief to be classified definitively as "prose" or "verse," while others continually shift between prose and verse. Moreover, the assumption frequently made by textual editors that text types can be determined retrospectively—a text fragment that is incorporated into a poem is a "verse fragment," while a text fragment incorporated into a letter is a "prose fragment"—is also not tenable, first because it fails to account for the essential autonomy of the fragment and its resistance to classification, and second because it fails to recognize that the author's intentions toward a fragment may be multiple. Here fragments are typed as "extrageneric" except in instances where a fragment is clearly a message to a specific (and usually identifiable) addressee and cannot also be considered a freestanding, autonomous text. Notes regarding type, state (rough-copy draft, intermediate-copy draft, fair-copy draft, etc.), media (pencil, ink), and hand (rough, intermediate, fair) are provided at the level of the text (DIV2), rather than at the level of the document(DIV1). Text-specific notes about type, state, media, and hand are especially important given Dickinson's practice of inscribing multiple texts in different states, media, and hands on a single document.

Information concerning the original location of the documents is provided at the level of the DIV2. Manuscripts sent by Dickinson outside of her personal archive are marked "mailed"; unmarked manuscripts may be assumed to have remained among Dickinson's own papers. Similarly, manuscripts bound into volumes are marked "bound"; unmarked manuscripts may be assumed to be loose sheets or fragments.

The determinations of "state" made here are open to revision. The terms "rough-copy," "intermediate-copy," and "fair-copy" are not ideal terms with which to describe documents of an essentially private nature. In order to distinguish a fair-copy sent out of Dickinson's personal archive from a fair-copy housed within her personal archive I have added the word "draft" to the latter category. The vast majority of the fragments appear to be rough-copy drafts. In only a few special cases—when a fragment appears to have progressed from a rough-copy to an intermediate-copy draft, or when a fragment has been excerpted from a fair-copy draft, for instance—have I offered a more refined hypothesis about the state of the fragment at the DIV2 level.

Internal Textual Divisions

The textual field is continuously divisible into smaller and smaller parts, greater and greater detail. The need to account for all of the possible internal divisions has been accommodated by the use of the AB element. In verse texts, the stanza or verse line has generally been understood to be the base structural unit. Dickinson's writings, however, challenge the ways in which we understand and mark internal textual divisions. Her commitment to metrical experimentation, particularly her breaking (out) of traditional meters, often leads to considerable uncertainty about what constitutes a line of verse. The noncorrespondence in her poems, moreover, between "metrically complete" lines and physical line breaks, many of which appear to be intentional, further problematizes the choice of the base structural unit. Rather than dividing verse texts into verse lines, I have divided them into stanzas; in verse texts without clear stanza divisions, the text remains undivided and the notation in the AB element reads simply: "lines, verse." The AB element has also been used to mark "variant clusters," the strings of variant or additional word choices often appearing below the body proper of the poem.

In prose texts the paragraph or, in certain instances, the sentence, has generally been understood as the base structural unit. Here, too, however, this base structural unit is problematic. In the first place, Dickinson did not use paragraphs in a conventional way: in her rough-copy prose texts there is rarely any division of lines into paragraphs; in her fair-copy prose texts there is often evidence of division, but the divisions are not always clear both because Dickinson did not in general indent the opening line of a "paragraph" and also because while she often marked shifts of thought/subject by leaving several blank spaces at the end of a line, many shifts remain unmarked because they coincide with the far edge of the page. Second, Dickinson's idiosyncratic punctuation, especially her dashes, often makes it difficult even to say where a given sentence concludes. Rather than dividing prose texts into sentences and paragraphs, I have chosen to divide them into "passages," an inclusive term designating "phrases," "sentences," and "paragraphs." In those prose texts where there is no clear division of text into smaller units—a common situation in the rough-copy prose drafts—the notation in AB element reads simply: "lines, prose." Finally, in instances in which the text shifts between verse and prose, as many of Dickinson's letters do, the locations of the shifts are recorded with a notation in the AB element. In addition to dividing, wherever possible, Dickinson's letters into passages, ABs also delimit the different spaces or levels of a letter: the date (when added by Dickinson), the salutation, the closing, the signature, and any postscripts following the body of the letter.

The fragments, almost all of which are interpreted as generically undecidable, are rarely meaningfully divided below the DIV2 level; the notation in the AB element for fragments reads simply: "lines."

The question of how best to divide Dickinson's poems, letters, and fragments in order to reveal their linguistic and physical structures has not been satisfactorily resolved and is a matter for continued scholarly discussion and debate. The potential of XML to represent different structures within a single text will almost certainly offer interesting alternatives to the present schema.