Electronic Textual Editing: Multimedia Body Plans: A Self-Assessment [Morris Eaves]

You've signed up for a seminar, ‘Editing in Multiple Media.’ It begins with a quiz and ends with a project, your own electronic edition. To start, answer a few hard questions about your proposed project in light of the examples offered.

Why does this material need editing? The emphasis falls on ‘need’ and ‘material’ because editing is a ruthlessly utilitarian craft that trades in sensible objects. In this realm even theory has consumable consequences.

Consider the William Blake Archive ( http://www.blakearchive.org ), which was made conceivable by the coincidence of several factors including the birth of the Institute of Advanced Technology in the Humanities at the University of Virginia (IATH) and the advent of exciting electronic scholarship, especially Jerome McGann's Rossetti Archive, but above all by the prospect of putting new media to work on an old, schizophrenic editorial legacy that emerged in Blake's lifetime (1757-1827).

The problem originates in a fusion of difficult content with difficult form that is unprecedented yet highly characteristic of the artist, who insisted on both the multifariousness and unity of his artistic identity and exercised a lifelong penchant for multimedia experimentation that was inhibited but rarely blocked by contemporary convention. Take the medium of much of his best-known work. He originally presented what he called ‘Illuminated Printing’—combining the tools, techniques, and materials of writing, drawing, painting, etching, printing and painting—as a multimedia solution to longstanding problems: ‘The Labours of the Artist, the Poet, the Musician, have been proverbially attended by poverty and obscurity . . . . owing to a neglect of means to propagate such works.’ Might the answer be his new ‘method . . . which combines the Painter and the Poet’ (692-93)? As he discovered, some new technologies merge seamlessly into the dominant system, others replace it, while any stranded but salvageable content from less imposing technical novelties becomes eligible for conversion into the dominant forms. Instead of a solution, then, illuminated printing became a problem: Blake's ambitious innovations produced a daunting overload of information and a trail of reader resistance and resentment. ¹

Would-be supporters instinctively countered with exercises in strong editing that soon became systematic. William Michael Rossetti was among the first to articulate the vital insight that studying Blake may involve some betrayal of his original forms: ‘Difficult under any circumstances, it would be a good deal less difficult to read these works in an edition of that kind, with clear print, reasonable division of lines, and the like aids to business-like perusal’ (see also Peattie; and Eaves, ‘Graphicality’). The printed edition became the centerpiece of the pre-electronic editorial settlement, which prepared the way for serious study and reflection on a scale previously unthinkable. Ultimately, however, dispersal, dismemberment, and translation seriously distorted the true picture of Blake's achievement. And that is primarily why we—Robert Essick, Joseph Viscomi, and I—undertook a new kind of edition.

Who needs your edition? The first imperative of editing is to meet the needs of audiences, and the audience that has guided our imaginations is the community of scholars. Our stated goal is to create an edition aimed at ‘scholars doing sustained original research’ (William Blake Archive Plan of the Archive: ). Blake specialists, then, are the hard core of a target demographic that softens toward the edges. Even the specialists are internally divided by discipline, usually art history or English. ‘Serious scholar,’ furthermore, includes experts who are not Blake experts. And the Archive aspires to be a ‘public resource’ (Plan of the Archive) to all those interested in Blake's work. In this respect our edition has more in common with libraries than with scholarly books and journals. But the thin end of our wedge, specialized scholarship, explains why the Archive is not primarily about reading or viewing Blake: it is about studying Blake.

Why are you multiplying media? From one notable perspective all editions are multimedia: every medium has digested other media that are its ‘contents.’ But, obviously, choosing that outer media shell is an utterly fundamental editorial decision. Choosing media of transmission and delivery dictates hard choices in hardware and software that inevitably box you into their particular corner. Editing forces such choices upon us because it puts ideas in concrete (though not necessarily inflexible) forms that are both pleasure and punishment. The Blake Archive attempts to break up a logjam of intractable old-media problems by exploiting deeply, but always within painful limits created by the choices made from a range of options, the capabilities of newer media to digest several old ones that were originally produced by the traditional tools and materials of painting, writing, and the graphic arts: etched and engraved prints, watercolor and tempera paintings, manuscripts, and typographic works, among others. But digital media use entirely different systems for processing words and pictures, much as print media comprise type and halftones. And, as I have argued elsewhere (‘Graphicality’), pictures have always been a problem.

What is your fundamental editorial strategy, and how are electronic media part of it? A strategy devised in light of the materials, the media of presentation, and the primary audience sets priorities and guides design. Editions are problem-solving mechanisms; without problems to solve, new editions would not be needed. Explicit strategy is essential, because the problems of producing hypermedia editions have not been fully faced or solved. The legacy of print offers much general but little specific guidance.

As an editorial solution, the Archive takes into account the uniqueness of Blake's work and the unusual needs of its scholars. But the basic elements, texts and still images, are fundamental to cultural memory. Hence in general form our aims are easily understood.

In one sentence, then: our strategy is to employ electronic media to achieve a level of consolidation, supplementation, and extension—overlapping categories—that will overcome significant disadvantages of both Blake's originals and printed reconfigurations. Departing from traditional strategies of substitution—as when letterpress editions replace watercolored etchings—we resituate the reproduced originals at the center of a complex but coherent structure of extensions and supplements, including tools. Consolidation empowers us to adopt a documentary approach that exploits the capacity of electronic media to digest carloads of data—edited documents with their supplements and extensions—on a scale highly impractical for printed books. This is neither to deny that electronic editing must adapt to major limitations nor to claim results consistently superior to print.

Providing redundant options—the original etched script, for example, our own documentary (‘diplomatic’) transcription, and even the transcriptions from Erdman's standard printed edition—provides multiple perspectives instead of a blinkered view. Historically, that frees us from Blake's perspective, which left us etched script to decipher in the first place. This is a central point. On the one hand, our strategy is anchored to highly controlled reproductions of Blake's original documents in order to restore his sometimes frustrating aggregations to positions of primary authority. But we are not retrospectively surrendering authority to Blake. Editors' rhetorical deference to authors' intentions can generate disorienting questions that mislocate the real source of power: ‘Would Blake have approved of the William Blake Archive?’ ² The history of editing suggests that we care primarily about our own needs and desires (which may include our desire to honor the artist's intentions). The ultimate strategic question is not what does Blake intend, but what do we intend for him and for ourselves? ³

It is useful to understand this crucial principle of choice when formulating strategy because, while some editorial features will represent the author's intentions and other features supplement and extend them, others may contravene them. Did Blake want his scripts transcribed for legibility? want his works enlarged, reduced, juxtaposed, categorized? want the words of other authors who have written on his prints and drawings reproduced for their documentary significance? Our aggressive determination to make these and other scholarly actions possible is authorized not by Blake's desires but by ours, backed by the powerful electronic medium that allows us to gratify them.

Would Blake approve scholarly approaches to his work? We may wonder, but our purpose and strategy do not address that question in practice—though, naturally, we persist in the hope that we do not murder our subject to dissect it.

How are your purpose and strategy carried through in design? Security experts have a saying: collecting information is one thing, analyzing it another. Your edition, like ours, will probably be a multifunctional apparatus of collection, recordkeeping, and analysis whose complexities (or problems) are drastically increased by multimedia commitments. Ideally, those functions will be embodied in efficient designs that optimize the advantages of the medium to achieve the interplay that sound scholarship requires.

Collection here embroils the Archive in the technologies of digital reproduction, representation, and cataloguing; analysis in the paraphernalia of search engines, elaborate indexes, numbering and measuring systems, comparison engines, etc.; and recording in various historical exercises (editors' notes, information about the physical object, provenance, etc.) that include recording our own editorial activity.

Consolidation bears multiple burdens. At its heart is an ecological mission of restoration. At a higher level it calls for the collection of multiple versions (two sets of illustrations to Milton's Paradise Lost: , several versions of the Songs). At a still higher level it aspires to the systematic arrangement of Blake's entire corpus—now the dispersed property of individuals and institutions around the world—in one virtual place with seamless connections to all related bodies of work. This is of course a vision, not a reality.

Editors of multimedia editions are in effect double-editing, first in discrete and then in integrated media. In a decade of intensive teamwork on the Archive we have evolved a fairly elaborate division of labor that follows suit, coordinating many moments of isolated specialization in a pattern that leads ultimately to integration. Securing accurate images of objects is basic, because they reproduce the evidence from which all else derives. This undertaking keeps us heavily invested in the techniques of digital image-processing: storage formats, scanning, compression algorithms, color and contrast correction, display resolution, scaling, and the like. ⁵ Likewise, creating accurate texts involves us in markup systems (SGML, HTML, XML) and the minutiae of textual criticism. ⁶ Our greatest challenge throughout is to establish appropriate benchmarks of quality with procedures to achieve and maintain them. Because these quality-assurance procedures have historical analogues in print scholarship—in the production of catalogues and facsimiles, for instance—we can compare results with our predecessors' if we adjust them to the advantages and limitations of our media. The advantages are enormous, the limits severe.

A huge pile of unsorted digital images and texts, even superb ones, would be too impoverished to sustain much serious scholarship. To make the images and texts meaningful and useful, we organize the information and supply contexts in three basic forms:

• ample bibliographical information (in the broad sense that includes, for example, art-historical provenances) compiled from the best current sources, including firsthand observation;
• self-documentation, ‘metadata,’ recording what is done, when, by whom (the ‘Info’ button under each image reveals the technical trajectory of each object from its source in a collection to its digital destination). ⁷ This exercise is motivated by the scholarly values of self-scrutiny and full disclosure—transparency for the sake of accountability;
• tools, which users employ to make useful things for themselves. With tools, applied to the primary textual and pictorial materials, users construct the contexts most relevant to their purposes.

The tools aid and abet the fundamental scholarly goals of close scrutiny fused with broad understanding. The techniques involved are complex products of an abiding need to see less of the object of investigation—to clear the field of clutter and to magnify the object—in order to see, paradoxically, more of it. The Blake Archive attempts, experimentally, to facilitate both. Our tools provide ways of:

• Searching texts (print precedents include page- and other unit-numbers, and the finding aids, such as tables of contents, indexes, and concordances, that use those numbers). Examples in the Archive include our text-searching engine that can query the contents of multiple versions as well as the editorial apparatus, including bibliographies, editors' notes, and illustration descriptions. Eventually it will be able to search even the markup codes.
• Searching images (print precedents include reproduction systems, such as engraving and photography, that aid recordkeeping, and systems of managing information about pictures, such as catalogues and indexing/description schemes, most elaborately manifested perhaps in the Index of Christian Art [1917-] and ICONCLASS [1972-]). ⁸ The Archive's experimental image-searching system, based on an extensive list of keywords and descriptions, mapped to visual details by means of an image grid, and viewed through Inote (see below), makes it possible to find most pictorial features.

• Inspecting by isolation and enlargement (with precedents in all manner of enlarged visual details—such as electron micrographs, maps that isolate particular aspects of a landscape—and verbal description of such details, close reading, selective quotation). Examples in the Archive include our high-resolution enlargements of every image; elaborate verbal descriptions of every version of every illuminated-book image; Inote, an image viewer/annotator that ties visual details to verbal descriptions; and Imagesizer, which allows users to control the scale of images. All these, individually or in tandem, help users answer the fundamental question, what is that?

• Comparison by juxtaposition (precedents include techniques of visually juxtaposing different versions of the same or the abnormal with the normal, which may in turn involve sequencing; and in the verbal counterparts of such comparisons). Examples in the Archive include the Compare feature, which will juxtapose any or all versions of individual works, such as the six versions of Blake's Marriage of Heaven and Hell currently available.

Much of this would apply as well to studying fingerprints as to studying art and literature. In general terms, what we face is a huge, dispersed stock of related information to be presented for study. In that light, the otherwise bizarre interface of a multimedia digital resource becomes a more understandable, at times almost comfortable, experience, or one at least as comfortable as multi-volume catalogues, concordances, and variorum editions but potentially—and sometimes actually—far more powerful.

Launching these actions and organizing this information on a single page—our OVP—is among our greatest challenges. More than one diagram is required to explain it (see figures 1 and 2), and it needs to be explained.

It looks only remotely like a printed page or like Blake's original productions—more like an instrument panel. It assuredly does not invite reading. But it invites investigative scrutiny, carried out through intensive searching, comparing, and other high-order manifestations of ‘scholarly primitives.’ Those, according to John Unsworth's anatomy of scholarly actions, are the elementary constituents of most of our labor. By combining access to documents with a set of contemporary tools, the OVP is designed to make complex scholarly procedures possible that are otherwise difficult if not impossible.

Out of this massive organizational undertaking has evolved what strikes me as the most remarkable feature of the new editorial settlement in its first electronic incarnation: its odd relation to the editorial legacy. Despite the heavily restorative, documentary slant of our project—which is ‘archival’ in that sense—we have by no means come full circle and arrived editorially back at Blake. Instead, his old aggregations have been restored (within strict limits, imposed mostly by the present limits of the medium conspiring with other limiting factors such as time and money) but then situated among the old disaggregations: transcriptions, verbal translations of images, search engines with functions segregated conventionally into texts and images, even a (searchable) version of the standard printed edition—all digitally reprocessed and tuned to the system, that is, to new positions as servants of the originals, and as concrete acknowledgements of two centuries of scholarly progress. The effect of the recuperation can be startling.

With some comparable sense, then, of what you want to do and why in this material realm, we can conclude with a glance at a few very material questions.

How will I finance my multimedia edition? One of the attractions of electronic editing has been the ready availability of the tools of the trade. Few editors own printing presses, but many own powerful computers with Web access. Another chance to bypass the middleman and cut costs? The early fate of Blake's own illuminated inventions may be a useful cautionary tale. To date, most good electronic editions have been expensive.

‘Electronic editing’ covers a multitude of media sins and opportunities, most prominently at this point CD, DVD, and the Web. Those can in turn be packaged and distributed in various ways restricted or unrestricted by payments and passwords. And ever more often, key components of ‘the’ edition can be treated as ‘content’ eligible for ‘repurposing’ and thus ‘delivery’ in multiple forms. Cutting the electronic deck depends upon a calculation of means, such as labor and funding, and ends, such as editorial vision and intended audience.

The Blake Archive has always been available via on the Web as a free site that has survived through the sponsorship of the Institute for Advanced Technology in the Humanities at the University of Virginia; the good will of the institutions and individuals whose property we reproduce; and soft money—generous foundation grants and university support. ⁹ It would be foolish to assume that any of these is permanent, and we will never be a set of volumes sitting securely on library shelves. We plow forward with no answer to the haunting question of where and how a project like this one will live out its useful life. Answers will probably begin to emerge eventually from publishers and libraries, the institutional cornerstones of print scholarship. Meanwhile, the need to persist without reassurance is one of the many unsettling conditions of life in new media. The costs and risks, however, should not be underestimated.

Are the materials you want to edit available? Availability is a sine qua non: if the materials can be had, strings will probably be attached in the form of permissions and/or payment. For printed editions this process is often made less onerous by long-established precedents that do not often transfer into the digital domain.

In this sensitive area it pays to formulate tactics early and take stock of your natural advantages. We based the Archive's early development primarily on several, including the cooperation of the Library of Congress, which has no permission requirement; the Blake Trust, which had made fresh transparencies of the illuminated books for a new series of printed facsimiles; and Robert Essick, one of the Archive's editors, who owns the largest collection of Blake in private hands. Those, plus a modest cohort of institutions whose recent participation in the Blake Trust facsimile project had demonstrated a commitment to scholarship that we, too, could draw on, were enough to give us a critical mass. At this writing, we have twenty contributing owner-institutions, ¹⁰ and negotiation with others has become a normal editorial duty.

How will you learn what you need to know? You are probably an expert in the subject matter of your edition. You can learn print editing from books, editors, or the MLA's Committee on Scholarly Editions. But the history of multimedia electronic editing is short, good examples are still scarce, and nothing stays put. This book should help, but continuous reeducation is the order of the day.

Electronic editing also requires technical knowledge that must be available in some form from someone. Of course editing books requires both editorial and technical expertise, but the understanding of bookish conventions is widely shared by editors and publishers. In the electronic realm, editors usually lack the basic knowledge of the medium required to conceive or execute their projects. We have addressed the knowledge problem in two ways, first by working hand-in-glove with IATH, which was fortunately founded to help humanists solve that very problem; and second by working collaboratively at great distances—via an email list (blake-proj), telephone, and the Web (Eaves, ‘Collaboration’). A burst of collaborative effort has been a surprising bonus of the humanists' entrée into the electronic domain. The Archive's radically collaborative design is based on a model that assumes, at the editorial level, (1) shared general expertise; (2) highly specialized competence in different sectors of Blake studies; and (3) specially developed competence in areas particular to the multimedia electronic edition; and then, in the most technical sphere, (4) expertise in computing (markup systems, programming, database management, etc.). We have bridged the gap between the first three and the fourth by appointing a graduate-student Project Manager as coordinator on the scene at IATH (see Kirschenbaum, ‘Managing’)—supplemented by assistants as required. Two previous Project Managers became our first Technical Editors: both Matthew Kirschenbaum and Andrea K. Laue are new-style Ph.D.s in English with concentrations in humanities computing, while our current Project Manager, Justin Scott Van Kleeck, is more interested in Blake than in XML.

How happily do you work with other people? Traditionally, humanists, if not quite Carmelites, are loners who find deep satisfaction in authoring, not ‘producing,’ books, not ‘content’ or ‘product,’ in disorderly cells they call studies. As an editor of multimedia electronic ‘objects,’ you will still have ample opportunity, perhaps more than you want, to use your study, but you will almost certainly have to share more of your time with others, taking almost nothing for granted, collaborating on decisions about everything you do. The aim is to produce precise maps of your production pipeline and everything in it. And if acronyms, technical terms, and phrases like ‘digital object’ and ‘production pipeline’ embarrass you with their taint of the worlds of technology and business on which you turned your back when you took the high road of humanities scholarship, you're probably in for some serious adjustments.

How much uncertainty can you bear? "Committed to print," we say, bookishly, because print stands for a kind of commitment we understand. This has shaped the modern history of editing, where the sense of prior commitment to a highly conventionalized medium has kept uncommitted editing—tentative editing—to a minimum. Digital editing is as unsettled as digital technologies. The dynamic editorial environment demands flexible theories, adaptable practices, and fast responses to unanticipated situations.

I experience it as X-editing, dominated by experimental action, Freud's label for the thoughtful hesitation before—or instead of—the leap from uncommitted thought to committed deed. ¹¹ The opening of a new arena for scholarship on the middle ground between hypothesis and action has produced a torrent of freewheeling experimentation characterized by multilateral problem solving, trial and error, approximation, compromise, revision, and, always, unintended consequences.

In this realm editing is speculation: by educated guesswork, we mold our practices to what seem the best bets at the time (which markup standards? which imaging algorithms?). Not that practices or consequences have ever been permanent and completely predictable in print, but the difference in editorial perspective is signaled, for instance, by the heavy traffic behind the scenes on our work-in-progress Web site, a combination editorial storeroom and testing area where we compare proposed solutions to present and future problems. The results of even the best solutions always seem to compromise our ideals.

Our protocols for diplomatic transcription are routinely insulted by our weak control over the display of texts as they travel from server to client. Every year we reconsider, revise, retest, and, if improvement seems possible, rework texts that last year seemed the best we could produce. Insider humor has generated a series of labels for our latest remodeling exercise: this year's is ‘de-uglification.’

On the imaging side, JPEG, a standard image-compression algorithm necessitated by present limits of memory and bandwidth, is tuned to the average picture, not to Blake's typical work. We must accept the JPEG compromise until something better comes along. So we cooperate eagerly with Vladimir Misic's experiments in combining JPEG2000—better but not the algorithm of our dreams—with ‘mixed raster content’ technology.

Editing on the electronic bleeding edge is like practicing medicine—hands always tied by present limits, one eye always scanning future developments. If you listen closely to editors editing, you will always hear the harsh sounds of primal conflict as visionary aspirations clash with reality. In a techno-commercial world the pressures of hard necessity bear down no less on editing with electrons than with ink, wood, or flesh. Evolutionary biologists think about body plans. Scholar-editors must be able to think as rigorously as they about our body plans and plans of action—including business plans—if we are going to bring adequate multimedia editions into being and keep them alive and healthy.

No question about it: ‘Hyperediting is what scholars will be doing for a long time’ (McGann, Radiant Textuality 68). The future of scholarly editing is electronic. Eventually we'll develop better coping mechanisms for dealing with its slights and shortcomings. McGann proposes a sharp distinction between theory that ends in ‘concrete acts of imagining’ (83) ¹² and theory that does not, the latter being what most people in the humanities mean by theory, the former being what electronic editors are doing. We can see what he means: editing in this theoretical spirit is a species of experiment that makes an object that then tests the idea that produced it, as buildings test architectural designs. For now we are engaged in a great collective experiment to find the most productive editorial paths. For now, electronic editions can be cool, but they lack the respect, the ‘dignity,’ that the Victorian poet and critic Swinburne perceptively saw in ‘the reprint,’ the posthumous edition he sought for the undignified Blake of the homemade illustrated poem. Electronic projects are subject to a splendid array of misunderstandings—mostly to their disadvantage. Such uncertainties make multimedia editing no game for those who need insurance. In time it may become a sanctuary of formula and routine. For the time being we wander in an editorial wilderness.