The William Blake Archive
Plan of the Archive

Significance | History | Present and Future | Dissemination

1. Significance

Over the course of two centuries, respect for the poems, prints, and paintings of William Blake (1757-1827) has increased to a degree that would have astonished his contemporaries. Now, more than a century since the "Blake revival" was inaugurated by a cadre of Victorian writers and artists, he is universally regarded as a seminal literary and visual artist. Today both his poetry and visual art in several media are admired by a global audience. He is one of the most anthologized and studied writers in English, and "The Tyger" is the most anthologized poem in the language. Blake's poetry and prose have been translated into many languages, and the published scholarship is international. According to G. E. Bentley, Jr., there is Blake criticism in Bulgarian, Catalan, Chinese (Mandarin and Cantonese), Czech, Danish, Dutch, Esperanto, Finnish, French, German, Greek, Hebrew, Hungarian, Italian, Japanese, Korean, Latin, Latvian, Lithuanian, Norwegian, Polish, Portuguese, Romanian, Russian, Serbo-Croatian, Spanish, Swedish, and Welsh. Blake Studies in Japan (Bentley and Aoyama, 1994) alone contains nearly 1000 entries.

Likewise, Blake is one of the most exhibited and collected of British visual artists: in April 1999, one copy of his illuminated [First] Book of Urizen (24 plates) sold at auction for $2.5 million, widely noted at the time to have been the most ever paid for a work of British "literature," and in 2004 a color printed drawing, "The Good and Evil Angels," sold at auction for $3.9 million—perhaps a record price for any print by anyone. In 2006 some of his newly discovered watercolors for Robert Blair's The Grave fetched high prices at auction. In recent years large shows have been mounted in Australia, England, France, Germany, Israel, Japan, Scotland, Spain, and the United States, including major exhibitions in Melbourne (1989), Tokyo (1990), Barcelona (1996), Madrid (1996), Melbourne (1999), Paris (2009), and New York (2009-10). Tate Britain's enormous 2000-01 exhibition moved to the Metropolitan Museum of New York in 2001, and the Huntington Library and Art Gallery mounted an exhibition in 2003. More major international conferences, exhibitions, and other special events marked the 250th anniversary of Blake's birth in 2007.

Blake—in various configurations of his life, words, and images—has established and maintained a kind of multifaceted cultural currency that is exceedingly rare. His images are icons, endlessly reproduced wherever striking images are needed, and every year his words are incorporated into books on science, technology, history, religion, and art. Jacob Epstein sculpted a bust of Blake; Francis Bacon painted his portrait. Blake's color print "Newton," translated into a twelve-foot statue by Eduardo Paolozzi, greets all visitors to the new British Library. Notable composers, including Benjamin Britten, Ralph Vaughan Williams, Paul Hindemith, Gustav Holst, Virgil Thomson, and William Bolcom, have set his words to music hundreds if not thousands of times—more often than any British writer other than Shakespeare. There are significant Blake ballets (Ralph Vaughan Williams), Blake operas (Britten's The Little Sweep), and Blake movies (Jim Jarmusch's Dead Man). Since it was set to music in 1902 by Hubert Parry, the prefatory poem to Blake's illuminated book Milton—"And did those feet in ancient time"—has evolved into an alternative national anthem for the British. On the popular music scene Blake's influence has been profound for half a century; his words have been extensively recorded by major artists.

Blake's resonance now extends to the students of the wide range of biblical and literary texts that he illustrated. Recent interdisciplinary approaches in literary studies, art history, and religious studies, especially, have encouraged students to understand illustrations as visual commentaries. Blake's illustrations—those of the Book of Job and many other scenes from the Old and New Testaments; Dante's Divine Comedy; Virgil's Eclogues; Chaucer's Canterbury Tales; Spenser's Fairie Queene; Shakespeare's plays; Milton's Paradise Lost, "Nativity Ode," Comus, Paradise Regained, "L'Allegro," and "Il Penseroso"; Thomas Gray's poems; and the powerful illustrations to John Stedman's Narrative, of a Five Years' Expedition, against the Revolted Negroes of Surinam, for example—offer such critical opportunities.

In the broadest terms, the William Blake Archive, an archive of scholarly electronic editions freely available on the World Wide Web, is a contemporary response to the needs of this dispersed and various audience and to the reciprocal needs of the collections that currently hold Blake's original works. Both the audience and the collections on which the audience must rely share a strong interest in the accessibility and preservation of Blake's works. The Blake Archive attempts to serve both sets of needs at once by providing free access to its Web site, where Blake's works are accessible to a degree heretofore impossible. Whether our users' inquiries are inspired by a love of imaginative art and writing, term papers, or scholarly research, no other resource can match the depth or range of access provided by the archiving, searching, and viewing options at our Web site.

But we have designed the site primarily with scholars in mind. For them we believe that the Archive has become indispensable—as a handy reference, a point of departure, or a site of sustained research. The Archive has adhered to exceptionally high standards of site construction, digital reproduction, and electronic editing that are, we believe, models of their kind. They make it possible for the Archive to deliver reproductions that are more accurate in color, detail, and scale than the finest commercially published photomechanical reproductions and texts that are more faithful to Blake's own than any collected edition has provided. We have applied equally high standards in supplying a wealth of contextual information, which includes full and accurate bibliographical details and meticulous descriptions of the content of each image. Finally, users of the Archive can attain a new degree of access to these works through the combination of powerful text-searching and (for the first time in any medium) advanced image-searching tools that are made possible by the editors' XML markup systems, detailed image descriptions, and innovative text and image software. Although we have designed the Archive to function properly within the limits of existing systems, we have built in considerable allowance for future improvements in hardware and software.

The Blake Archive first opened to the public in 1996 with simple reproductions from two of Blake's early "illuminated books" in "illuminated printing," as he labeled them. The reproductions were lightly encoded (in HTML only) and accompanied by little contextual information, with no search capabilities. The Archive now contains fully searchable and scalable electronic editions of many copies of Blake's illuminated books in the context of full, up-to-date bibliographic information about each image, carefully edited transcriptions of all texts, detailed descriptions of all images, and extensive bibliographies of Blake's works and scholarship about Blake. This expansion has also seen the inclusion of examples of Blake's other work—commercial book illustrations, separate prints, drawings and paintings, and manuscripts—as well as a searchable electronic version of David V. Erdman's The Complete Poetry and Prose of William Blake, the standard printed edition for reference.

Meanwhile, the roster of major contributors has grown to encompass every major Blake collection in the world: the Library of Congress (now a sponsor) has been joined by the Huntington Library and Art Gallery; the Essick Collection; the New York Public Library; the Morgan Library and Museum, New York; the Houghton Library and Fogg Museum of Art at Harvard; the Yale Center for British Art; the Glasgow University Library; the Fitzwilliam Museum, Cambridge; the British Museum; the National Gallery of Art, Washington; the National Gallery of Victoria, Melbourne; the Metropolitan Museum of Art, New York; the Museum of Fine Arts, Boston; the Birmingham Museums and Art Gallery; the Whitworth Art Gallery, Manchester; the Victoria and Albert Museum; the British Library; the Tate Collection; the Louvre Museum; the Ashmolean Museum, Oxford; the Bodleian Library, Oxford; the J. Paul Getty Museum; the Royal Institution of Cornwall; and the Victoria Library at the University of Toronto (see Contributing Collections). Our success to date in obtaining the confidence and goodwill of owners has reassured us that a rather vast undertaking such as ours is possible at the present time. As it is difficult to know how long present opportunities will last, we are pressing forward now.

Further acknowledgement of the significance of the Blake Archive as a progressive scholarly and pedagogical enterprise has come from the Institute for Advanced Technology in the Humanities at the University of Virginia, which first perceived the value of undertaking such a challenging project; the Getty Grant Program, which provided major funding for the first phase of design and construction; and the Preservation and Access Division of the National Endowment for the Humanities, which helped to fund our second phase. We have benefited from additional grants from the Paul Mellon Centre for Studies in British Art, London, and the University of North Carolina at Chapel Hill, in addition to software and hardware contributions from Sun Microsystems and Inso Corporation. Our headquarters and source of major support since 2007 is the Carolina Digital Library and Archives (CDLA) at the University of North Carolina at Chapel Hill, significantly reinforced by the Department of English at the University of Rochester.

Scholarship: The Archive had its origins in the fortunate confluence of four phenomena during one brief span of time in the early 1990s: the completion of a broad base of mature Blake scholarship, capped by the publication of the first trustworthy map of the history of Blake's illuminated book production; the appearance of technology sufficiently revolutionary to alter some fundamental assumptions in scholarly editing; the emergence of new technical standards sufficiently robust to check, if not eliminate, the formidable threat of overnight obsolescence for large undertakings such as ours; and, finally, the creation of organizations specifically devoted to giving digital form to the ideas of humanists. Together these four developments combined to provide the cornerstone of integrated archival, editorial, and educational initiatives that would have been impossible ten years, and probably too risky even five years, earlier.

Despite decades of scholarship, the knowledge of Blake's work was fragmentary and unsystematic until the final quarter of the twentieth century, partly because the upturn in his reputation came so long after his death but largely because the study of Blake has often split into distinct institutional compartments. Literary critics and art historians operate with different assumptions, values, aims, and procedures even when investigating the same original materials, as they were in this case. The resulting lack of coordination delayed the appearance of the standard tools of reference and reproduction that represent what we know about Blake, a printmaker and painter who was also an author.

But by the late 1980s the standard points of reference had at last fallen into place: reliable printed editions of the poetry, prose, and letters (Bentley 1978, Erdman 1982); excellent, if rare and expensive, facsimiles of one (but only one) of each of Blake's illuminated books (Blake Trust-Trianon Press facsimiles, 1951-1976); documentary records of the life (Bentley 1969, 1988); and sound catalogues of the major categories of Blake's oeuvre, including the drawings and paintings (Butlin 1981), illuminated books and secondary criticism (Bentley 1977, 1995), complete graphic works (Bindman 1978), separate plates (Essick 1983), and commercial engravings (Essick 1991), augmented by numerous subsidiary catalogues that fill out the picture of Blake's multifaceted productive life. However, the crucial piece of this foundation that concerned Blake's illuminated books was weakened by errors that had generated a history of false inference and an essentially false overall picture. In 1993, building on the work of Robert N. Essick and G. E. Bentley, Jr. but reassessing the fundamental evidence, Joseph Viscomi's Blake and the Idea of the Book redrew the map of Blake's productions in his most famous and difficult medium, illuminated printing. This indispensable if uncoordinated legacy of scholarship in disparate disciplines set the stage for a new phase of radical editorial revision of a kind that would be almost unimaginable in the medium of print.

Technology: The technology that held the most promise in such a case was of course global network computing via the Internet and World Wide Web, which made it possible to conceive a long-distance professional collaboration and an "edition" of Blake that would transcend the limitations of conventional scholarly editing and, in the process, render irrelevant the gap between the original works in restricted collections, the incomplete sets of expensive facsimiles in the rare book rooms of some large university libraries, and the indispensable but highly misleading printed editions on which "readers" had relied for their "Blake" since the Victorian Blake revival.

Encoding system: But we were wary. Like everyone else in the humanities, we had seen grand scholarly hopes crucified on the cross of technological change and instant obsolescence. Not long before we began thinking seriously about the uses of digital technology, an unanticipated change in proprietary videodisc technology had pulled the rug out from under an ambitious scheme to reproduce a substantial selection of Blake's images at the University of Iowa. The PC-Mac wars were another reminder of the danger. But the promise of platform independence and portability represented by the codification of Standard Generalized Markup Language (SGML, the source of the Web's HTML and later XML) and its scholarly counterpart in the coordinated standards of the Text Encoding Initiative (TEI, 1987-) finally made it plausible to devote years of work to an electronic scholarly resource in the humanities.

Institutional base: But the promise of closing the editorial gap forced us, as humanists, to face the technological gap: we could half-envision electronic remedies that we could not execute. At that point, in 1993, we crossed paths with the new Institute for Advanced Technology in the Humanities at the University of Virginia (1992-). IATH's mission, we heard, was to help humanists use new information technology in carrying out their projects by supplying the requisite expertise and equipment at the research and development stage.

Our preliminary discussions with the staff of IATH introduced us to an exotic new world of markup codes, servers and clients, the Web, and Java. But the primary consequence was the conception of a William Blake Archive, which would be a comprehensive but coherent array of electronic scholarly editions to be made available free of charge on the Web. We came to believe that, given an elegant design and sufficiently powerful features—including an innovative way of searching for individual details in all the images in addition to the more conventional searches for specific texts—our project would help to set the pattern for serious art-historical and textual scholarship by electronic means at a key moment in their evolution. For a large international community of art historians and literary critics, among others, the Blake Archive would be a powerful reference tool, offering high-quality reproductions of an important body of work—much of it previously unreproduced, badly reproduced, or reproduced in rare volumes—and making that work accessible and usable in new ways that would improve interdisciplinary knowledge in areas where more and better knowledge was sorely needed.

Literary critics tend to favor the illuminated books in their research; art historians tend to favor paintings, prints, and drawings. While those disciplinary biases are of course natural and to some extent inevitable, the price of such specialized attention—in understanding the full artistic situation—has often been higher than it should be or needs to be. By incorporating as much of Blake's pictorial and literary canon as possible—with both images and texts organized, interlinked, and searchable in ways that only hypermedia systems will allow—the Archive would for the first time give scholars and students access to the major intersections between the illuminated books and Blake's other creative and commercial works. That is to say, by exploiting new information technology to deliver the historical, technical, and aesthetic contexts necessary to study Blake as printmaker, painter, and poet, the Archive would encourage a deeper, more responsible understanding of his aims and methods, which have been regularly misunderstood and misrepresented.

The concrete results would be:

  • a large, searchable hypermedia archive on the World Wide Web; and
  • eventually, once the architecture of the Web-based archive was substantially complete and a broad representation of Blake's work was in place, a series of the works on portable media such as CD-ROM or DVD-ROM disks.

Both products would be designed for use by a broad audience of scholars and students in their studies, classrooms, and museums. At this time, portability made disks a popular medium in classrooms and on desktop computers. But for scholars doing sustained original research, there would be no adequate substitute for global access to a platform-independent William Blake Archive published on the Web.

Thus all along our fundamental aim has been to construct a unified international resource out of highly disparate and dispersed original materials to which access is ordinarily limited by institutional and other restrictions and by the sheer cost and difficulty of travel. As a public resource, the Blake Archive would be maintained free and open to all those who have access to the Web anywhere. We hoped we could persuade major collectors and collections of Blake material to agree to contribute their works to the Archive in precisely that public spirit, especially considering that very few works by Blake are on permanent exhibition due to concern about the effects of handling and excessive exposure to light. Several major institutional collections had severely limited even scholars' access to the fragile originals. This included at least two of our contributors, the Huntington and the Fitzwilliam. Both institutions had disbound and/or rebound several illuminated books in an attempt to improve preservation. Our contribution to these attempts at preservation is very direct: by making available searchable and sizeable images of the highest quality, we can provide access without compromising preservation in any way. Institutions that contribute to the Archive can continue to provide scholars and the public full access to these treasures while at the same time taking all necessary measures to preserve the originals. As a curator wrote to us, "I look forward to viewing [his institution's illuminated books]. You will probably know that we have now made a link from our web pages to the Blake Archive so that users can easily view our copies rather than over-tax the originals. With a fast-enough machine they come up almost instantaneously."

Once archived digitally, structured and tagged (indexed for retrieval in SGML—and now in XML—adapted to the purpose), annotated with detailed descriptions, and orchestrated with a powerful search engine (in this case DynaWeb software), the images in the Archive could be examined like ordinary color reproductions. But they could also be searched alongside the texts, enlarged, computer enhanced, juxtaposed in numerous combinations, and otherwise manipulated to investigate features (such as the etched basis of the designs and texts) that had heretofore been imperceptible without close first-hand scrutiny of the far-flung original works. For example, the information necessary for doing good art history would enable scholars and students to draw sound conclusions about the differences between what Blake etched on his copper plates and what was added or changed afterwards in printing and coloring the impressions. But the published reproductions upon which much art-historical study necessarily depends simply cannot record such details with sufficient accuracy. Even scholars who are able to globetrot from collection to collection end up relying heavily upon their inadequate memories, notes, xeroxes, and photographs to compensate for the distances in time and space between collections. Seeing the originals is good in itself; but seeing them in fine, trustworthy reproductions, in context and in relation to one another, is the scholarly ideal. Difficulty of access to originals and reliance on inadequate reproductions have handicapped and distorted even the best efforts. Again, the inevitable result has all too frequently been distortions of the record, misconstructions, and the waste of considerable scholarly labor.

We began by tackling the multitude of challenges presented by a single category of Blake's work, his illuminated books. The illuminated books had their genesis in a series of graphic experiments that Blake began around 1788 and quickly evolved into a program of combining visual and textual elements in printed pages that he could control—design, write, etch, print, and color—himself. Though he produced a great deal of important work in other media, the illuminated books (c. 1788-1827) span most of his productive life and reflect its characteristic patterns. These much-discussed books are fundamental to his artistic reputation for several good reasons: they are spectacular examples of the illustrated book at one extreme of its development; fascinating explorations of the interactions between texts and designs at a level of narrative maturity seldom matched and never exceeded; major instances of the transformations of traditional iconography in the late eighteenth century; and central documents in British romanticism, both as historical period and as ideology.

In his lifetime, Blake produced about 175 copies of his 19 illuminated books. About 20% of those—40 or so—have been reproduced in print, sometimes well, sometimes execrably, but in no coherent historical order. By the end of the first phase of our project, in June 2000, we had reproduced 41 copies, about half of which had never been reproduced before. This constituted, for the first time, an archive of reproductions suitable for serious research. Only since the mid-1990s, largely as a result of Viscomi's extensive research, has the history of the production of the illuminated books been correctly understood. That in turn made possible for the first time a rigorously constructed scholarly archive, including numerous copies of the illuminated books that had been neglected because their place in the history of production was not understood.

Our next task was to add drawings; paintings; and several kinds of prints, manuscripts, and rare or unique typographical works to the Archive. Doing so yields an augmented "Blake" considerably larger than the one most familiar to students and scholars, especially those who approach Blake from the literary side. Without sacrificing the Blake of the Songs of Innocence and of Experience and The Marriage of Heaven and Hell, the extended Archive reveals the painter-printmaker whose illuminated books emerged from the materials, work routines, and imagery of eighteenth-century history painting, water color drawing, and graphic arts, as well as from the literary routines of Milton, the Bible, Swedenborg, and Boehme, which students of Blake have more often investigated. Commercial illustrations, for example, can bring into focus a major convergence between Blake's illuminated book Visions of the Daughters of Albion (1793) and his engravings to J. G. Stedman's Narrative, of a Five Years' Expedition, against the Revolted Negroes of Surinam (etched 1791). Even the familiar profile of a poet-Blake can be deepened by the inclusion of Blake's extensive group of literary illustrations, such as those to the Bible and to the poems of Dante, Milton, Gray, Blair, and Young, among others.

We have come to see the Blake project as a pacesetting instance of a fundamental shift in the ideas of "archive," "catalogue," and "edition" as both processes and products. Though "edition" and "archive" are the terms we have fallen back on, in fact we have envisioned a resource unlike any other currently available—a hybrid all-in-one edition, catalogue, database, and set of scholarly tools capable of taking full advantage of the opportunities offered by new information technology.

Uses of the Blake Archive as a prototype: The Blake Archive is not designed to be an isolated resource. We foresee a time when it will be only one of very many cooperating electronic resources for scholarly research. Thus we have been members of the groundbreaking NINES (Networked Infrastructure for Nineteenth-Century Electronic Scholarship) consortium from its inception in 2003. But until the standards for such resources evolve, projects like ours must be in the business of setting standards by designing prototype tools and techniques. A guiding aim of our project is to solve problems with methods that are widely applicable. The collaborative procedures we are developing, which we hope will become useful prototypes of "distance editing," depend upon intensive day-to-day teamwork among the three editors and the staff of CDLA to integrate the textual, art-historical, critical, and technical expertise necessary for the construction of a scholarly resource as complex as this one.

We see the products of our collaboration as similarly prototypical: in facing new technical and editorial challenges, the Blake Archive will make available new tools for future archivers, editors, and cataloguers. The Document Type Definition (DTD) developed by the Text Encoding Initiative (TEI) has become the default standard for encoding text-based projects in the humanities. We believe that the Blake Archive DTD has the potential to become the standard for encoded electronic images, as the TEI has for encoded electronic texts. In this image-oriented non-TEI DTD—an art historian's DTD if you will—we are constructing the most comprehensive applied product of its kind, one robust and flexible enough to be adapted readily to the needs of many other projects that use images extensively. In addition, software developed by IATH, such as our Java applets Inote and ImageSizer, is adaptable to virtually any project in which images are important.

The perception that our editorial and technical experience is directly relevant to other projects has generated steadily increasing interest in the Archive that has nothing to do with Blake per se. Hence, for example, we have been invited to speak at numerous meetings where the subject was not Blake but humanities computing or editorial theory.

Because the signal advantage of electronic editing and cataloguing is the open-endedness that makes it possible to add materials, correct errors, incorporate new discoveries, and construct new relationships, we have believed from the first that our principal objective—the most significant contribution we can hope to make—should be the creation of a sound and durable foundation for decades of future scholarship. Since 1995, when the Getty Grant Program underwrote the initial phase of our project, we have worked to shape the foundations of the Archive in strict accordance with our original ideals and priorities.

Top | Significance | History | Present and Future | Dissemination

2. History

Background, 1991-1994: In 1991-93, at work on two printed volumes in a new series published by the Blake Trust, Tate Gallery, and Princeton University Press (see Editorial Principles), we came face to face with the limitations of even lavishly illustrated books for the kind of Blake edition we had envisioned and began to conceive the outlines of an electronic edition that might overcome many of these limitations.

With this in mind, at the urging of Jerome McGann we visited IATH in the summer of 1993 to see his Rossetti project (see "The Rossetti Archive and Image-Based Electronic Editing," 1996; "The Rationale of Hypertext," 1997; and "Imagining What You Don't Know: The Theoretical Goals of the Rossetti Archive," 1998) and to meet with the staff of the Institute, including John Unsworth, who was the new director. After extensive discussions and demonstrations, we concluded not only that our concept of a rather primitive electronic edition was technically feasible but also that a scholarly resource far more ambitiously transformative was within the realm of possibility.

In 1994 we applied to become Associate Networked Fellows of IATH and drew up a preliminary proposal for a Blake "archive"—appropriating McGann's term for our somewhat different purposes—in three major phases that would tackle, first, the difficulties presented by the illuminated books; second, the remaining categories of work in Blake's oeuvre (prints, paintings, drawings, manuscripts); and, third, such issues as secondary publication (on disks), interpretive supplementary material, and educational applications. A great deal of further planning over the course of the year developed this initial proposal into the original blueprint for the Archive.

Blake Archive, phase one, 1994-2000: In 1995 we received a grant from the Getty Grant Program to underwrite the initial three years of planning and execution focused on the illuminated books. The Blake project was the first opportunity for IATH to work intensively with researchers from outside the University of Virginia community. The Institute provided the team with full-scale technical assistance and archive-design consultation, along with the necessary equipment to establish the foundations of the Archive.

Year 1, 1995-96: The Institute phase of the project began with Joseph Viscomi as a resident fellow for the year. With our first project manager, Amy Sexton, and a student technical assistant in place, the three editors met with IATH staff in the summer of 1995—in retrospect, the first "Blake camp," as we came to call these annual planning and problem-solving sessions. We drew up a two-phase plan with illuminated books in the first phase and non-illuminated works in the second. At the end of the second phase the architecture of the Blake Archive would be complete and all its wings would have substantial content.

We settled easily into our division of labor as an outgrowth of the editors' experience with the Blake Trust volumes. We would make all final decisions collectively. We would deal with institutions according to our individual experience with them. Beyond that, Eaves and Essick would share major responsibility for generating the bibliographical information and image descriptions. Viscomi would take major responsibility for generating digital images and transcriptions of Blake's texts. Among other duties—parsing the SGML markup, moving works from testing to publication, etc.—the project manager would coordinate activities at IATH, including our project's access to the technical staff. Everyone would proofread and test.

We compiled a prioritized list of illuminated books (for our principles of selection, see Editorial Principles) and began to seek cooperation from key collections of Blake material. After reading and consultation that included sessions with a member of the Getty/MESL (Museum Educational Site Licensing) project in digital imaging, we conducted extensive trials to determine the optimum balance of photographic format, scanning resolution, and file size for an archive of this type. Key benchmarks were arrived at (see Editorial Principles), and enough fundamental design work was completed to move us to the next stage. We established blake-proj, the online discussion group that has proven essential to the collaboration that has been a hallmark of the Archive's development (see Morris Eaves, "Behind the Scenes at the William Blake Archive: Collaboration Takes More Than E-Mail," The Journal of Electronic Publishing 3.2 (Dec. 1997).

By the end of the inaugural year we had concluded agreements with four contributors that control access to thousands of Blake's images: the Library of Congress (which became our official co-sponsor), the Huntington Library and Art Galleries, Glasgow University Library in Scotland, and the Essick Collection, the largest collection of Blake and his followers in private hands. The editors personally supervised a six-day photographic session at the Library of Congress that yielded 620 images. All the new photography was in our benchmark format, 4" x 5" color transparencies with color bars and gray scales.

During the year we made our first public outings. We presented our plans at the Society for Textual Studies meeting in New York and did our first demonstration (using mockups) at the meeting of the Society for Documentary Editing in Baltimore. Finally, we opened a Web version of the Blake Archive to the public with two copies each of Blake's Book of Thel and Visions of the Daughters of Albion, using the first version of our page design, lightly coded in HTML only, with no search capabilities or Java applets.

Year 2, 1996-97: In retrospect this was a make-or-break year. At the second summer Blake camp we discussed our goals: to pursue negotiations with additional major collections; to complete the first version of the Blake Archive DTD—the Document Type Definition (DTD) for the SGML-encoded illuminated books; to add search capabilities for images as well as for texts; and to ready for testing a single work for publication in something approaching a fully operative form. New hardware (a server), software (DynaWeb, a search engine), and technical support were donated by Sun Microsystems and Inso Corporation.

A major initial challenge was to find a system to base our image searches on. We began by giving careful consideration to two leading candidates. "Computer vision" or "pattern recognition" searches attempt to identify objects in digitized images by recognizing patterns (wheat vs. waves). Controlled-vocabulary searches operate by assigning words to pictures (sheep, Urizen, finger, contrapposto). The most formidable iconographic classification system is Iconclass, developed by a Leiden art historian, Henri van der Waal, and published after his death in 17 volumes, 1973-85.

Computer vision is a fascinating research subject in computer labs but, we quickly discovered, completely incapable of making the precise discriminations we had in mind. Iconclass is more capable—with the advantage of being a published point of reference. But though it aspires to be a universal system of image description, Iconclass was first used to describe medieval and Renaissance iconography, and it is still best suited to images that are typical of a certain school of art. Its classifications are very different from our intricate descriptions of individual images and their numerous components.

Blake's works are full of unconventional elements. His uniqueness and the specificity of Blake studies require a more detailed and precise semantics, we concluded, than a general system like Iconclass provides. Thus we set out to develop our own controlled vocabulary of descriptive terms empirically, in the course of describing the images for the first time—a procedure we have followed throughout the project with excellent results, we believe. We remain open to alternatives, however.

We decided to incorporate two significant reference works: an extensive bibliography of works useful in the study of Blake and David V. Erdman's standard printed edition, The Complete Poetry and Prose of William Blake (complete with Erdman's textual notes). We published the first bibliography in 1997. The Erdman edition presented more formidable challenges, including a separate TEI-compliant DTD. The first round of encoding was completed in 1997, but, in order to create a searchable text, integrated with the rest of the Archive, a good deal of further work was required. (We tested two early versions and published early in 2000.)

A sophisticated final design for the site, complete with search engine and a Java applet, Inote, was refined over several months of discussion, experiment, testing, and revision. Inote, a tool designed at IATH for image viewing and annotation, was integrated with the search engine to zoom in automatically on particular visual details and display the editors' descriptions in image searches. We decided that users must also have a way of controlling the size of images. The result was ImageSizer, a second Java applet, which allows a user to view images at their original size or to enlarge and reduce them at will.

Meanwhile, the three editors at their separate outposts and the project manager and technical assistant at IATH proceeded according to the division of labor that had been worked out the previous year. Digital scanning, painstaking image-by-image color correction on specialized professional equipment, and elaborate SGML markup of all images began in earnest. As always, the Archive discussion group blake-proj remained the place where all momentous and trivial issues were hashed out. In addition, we created a work-in-progress Web site (our WIP site), accessible by password, where we could conduct all our pre-publication testing.

During the year we made public presentations of the Archive in New York; Washington, DC; New Haven; Cambridge, UK; and Oxford, UK. We also reached final agreements with two new contributors, the New York Public Library and the Yale Center for British Art, and continued the process of acquiring and scanning 4" x 5" transparencies as we approached other major repositories.

Years 3-4, 1997-99: Whatever remained unresolved in our usual forums—the work-in-progress site and blake-proj—was moved onto the agenda of our third Blake camp, where we were able to get to the bottom of thorny problems—the logic of searches, the operation of Inote, the consistency of displays, the structure of the SGML hierarchy, editorial formats—that were blocking the first publication of a fully searchable and resizable work in the Archive. After two months of intensive tweaking and testing, we added a detailed Help document and opened the new site to the public in August 1997 with a single work, The Book of Thel copy F, from the Library of Congress—a mere eight plates but, as far as we were concerned, a landmark.

We determined that 1997-99 were to be years of production and publication that would move us toward the primary goal we specified in the original plan: to publish at least one copy of each of Blake's 19 illuminated books, along with multiple copies of several books. As of June 1998, 23 copies of 13 books had been published; as of June 1999, 33 copies of 18 books had been published. (For information on the significance of individual copies, see Blake Archive Updates).

In early 1999 we also added a new wing to the site. About the Archive made available a large body of documentation and supplementary material: the Archive at a Glance (for a quick overview), Editorial Principles, a Technical Summary, a detailed Plan of the Archive, Frequently Asked Questions, and a list of published scholarship about the Archive along with published reviews of the project. In June 1999 we added a long-anticipated Tour of the Archive. Through a sequence of several dozen graphical screenshots linked to narrative commentary, the Tour introduces users to the basic organization and structure of the Archive, the features of its interface, its search options, and the function of the Inote and ImageSizer applications. Since then, we have continued to revise and expand About the Archive.

During these years we reached agreements with the Pierpont Morgan Library, New York; the Houghton Library, Harvard University; and the Fitzwilliam Museum, Cambridge University. We were well along in the process of acquiring the necessary photographs from these collections, some by generous loans of file transparencies, others by new photography: at the Morgan Library, for instance, the editors supervised a three-day photographic session that produced 255 images. In 1997-98 the Archive received a modest grant from the Paul Mellon Centre for the Study of British Art, London, the first ever given by the Mellon Centre in support of an electronic project. We gave public presentations of the project on several occasions (see Presentations and Demonstrations about the Archive); a well-attended session on romanticism at the annual MLA meeting (San Francisco, Dec. 1998) eventuated in the publication of several papers on the Archive (Wordsworth Circle 30.3 (summer 1999); see Articles about the Archive).

Year 5, 1999-2000: We spent the year consolidating our gains and preparing to initiate, in 2000-03, a second phase of development.

Matthew Kirschenbaum, who had served as our project manager while a graduate student at the University of Virginia, became our technical editor when he moved to his new faculty position in humanities research computing at the University of Kentucky (since 2001, at the University of Maryland). Kirschenbaum's appointment, one of the first of its kind in the nation, allowed him to continue overseeing the technical development of the project in collaboration with IATH, while also working to refine some of its potential contributions to humanities computing and digital libraries, such as the Archive's prototypes for searching and manipulating structured image-based data and the expansion of the Archive's DTD into a more generalized resource usable by other projects.

After final testing, we published our TEI-compliant, searchable electronic version of David Erdman's edition of The Complete Poetry and Prose of William Blake, which had been in production for more than three years.

We continued publishing additional copies of illuminated books in order to clear our second hurdle: after publishing one copy of every illuminated book, to publish at least one copy of every printing of every book, according to the historical rationale that sets our priorities (see Editorial Principles). We also added multiple copies of works from several printings—mostly copies that have seldom or never been reproduced or reproduced badly—and began to create the structure that would eventually allow us to incorporate related materials that help to document the history of production in context (individual proofs, early states, sketches, associated drawings, prints, and paintings).

The last illuminated book to be represented in the Archive, Blake's 100-plate Jerusalem (copy E, the only complete colored copy, from the Yale Center for British Art), continued to serve, while in development for several years, as an object of editorial inquiry that led gradually to major changes in our protocols.

Our newly convened advisory board—a broad group of museum curators, art historians, textual critics, romanticists, authorities in humanities computing, and Blake scholars—agreed to participate in the testing of new works at our work-in-progress site. To facilitate communication, we set up a new online discussion group for them.

Meanwhile, we reviewed our procedures for reproducing images in discussions with experts in image compression at the University of Rochester, Kodak, and Xerox, including members of the international team that developed JPEG 2000 (see Technical Summary).

Finally, of course, we continued to acquire additional works from our contributing institutions and to establish cooperation with others. But we already had access to more than enough material to represent very amply the artistic range of Blake and his contemporaries. We also presented our project in a variety of public forums in cities including Dublin, Ireland; Charlottesville, Virginia; and Halifax, Nova Scotia.

Phase two, 2000-03: As our initial three-year funding from the Getty Grant Program allowed us to initiate the first (five-year) phase of our work, an NEH Preservation and Access grant allowed us to initiate a new phase of development with three principal aims: to continue adding illuminated books to the Archive; to extend the image-oriented Blake Archive DTD and stylesheets to accommodate the other categories of Blake's works; and to incorporate a significant representation of work from these categories into the new wings of the Archive, including a preview wing—a provisional area that would allow a more expeditious movement of works from the editorial pipeline into the Archive proper. In the process, we aimed to greatly increase the size of the Archive and make it fully representative of the range of Blake's work.

We began by assessing the requirements, scholarly and technical, of each category of work required to complete the structure of the Archive: prints (original and reproductive), paintings, drawings, manuscripts, and typographical works. From that assessment flowed the design for our original SGML architecture. Although we started with the illuminated books because they required us to address the problems of both texts and pictures (discrete, juxtaposed, and fused), the other categories of Blake's oeuvre are distinct in important respects. For instance, the illuminated books are book-like in scale; none is as small as Blake's smallest works or as large as his largest paintings. The issue of scale—to take only that one—raises art-historical questions (should we continue to privilege the actual size of the images?) and technical questions (should we limit the use of ImageSizer or extend its capabilities?) that we have had to deal with. The task at hand was to anticipate changes in the DTD and stylesheets, interface, image descriptions, and art-historical and textual information, and arrive at an initial set of blueprints that would allow us to add new subcategories of Blake's work to the Archive.

Much of the work of the second phase was behind the scenes—making editorial changes, creating prototypes, and running tests. These were precursors to basic revisions of the elements in the DTD on which the Archive rested. We adopted a radical documentary-style line-numbering scheme capable of accommodating not just conventional printed literary texts, such as poetry, but any text—inscriptions, labels on engravings, signatures and imprints on engravings, monograms, and numbers. We revised the stylesheets that control the display of works, and we modified the "copy headers" and "object headers" that deliver fundamental bibliographical information about the images to the user. We also implemented new transcription standards in the display of several of Blake's works. These standards attempted to adhere to the shape of Blake's text more closely than we were previously able to reproduce in a digital medium. New servers installed at IATH in 2003 noticeably improved access times, which were increasingly important as the Archive continued to expand the number of works on the site and as more users came to the site.

Along with the redesign and extension of the DTD came problems of rendering and description—how to render images across various new categories in ways that would be coherent with the rest of the Archive and yet remain faithful to the images themselves. Blake's work beyond the illuminated-book canon includes the drawings, paintings, and prints to which the artist devoted most of his productive life, along with important typographical works and manuscripts. Although the Archive has been designed with the aim of accommodating all of Blake's work (and, for that matter, the work of his contemporaries as well), these additional categories have required special attention at each stage: scanning and correcting line engravings, for instance, present new technical problems, such as a tendency toward optical distortion in areas of close crosshatching, that we have had to solve; single "paintings," separate "plates," and continuous "pages" of type involve us in structural relations distinct from those characteristic of the illuminated books. Similarly, we have been working out protocols for describing images that are in many respects unlike the images of the illuminated books for which our controlled vocabulary was developed.

We addressed these problems in the process of publishing two extensive bodies of work that we identified as the most logical extensions of the Archive: Blake's illustrations (in several media) to the Book of Job and to the poetry of John Milton (almost 200 items). (Users had written us asking for Blake's designs to Job, Milton, and Dante.) These images, which are often historically, thematically, and formally related to the illuminated books, meet two criteria of "significance": they are representative of Blake's oeuvre as a whole, and they are among the works most often studied by Blake scholars. Hence they constituted a logical, large, but manageable first extension of what we had already done. The Huntington, one of our contributors, holds the world's largest collection of Blake's illustrations to the works of Milton.

Our second phase involved the difficult transition from a specialized to a generalized framework: from one capable of supporting a specialized collection (of illuminated books) to one capable of supporting a far broader combination of texts and images, plus extensive information about them and robust scholarly tools to manipulate them. Obviously, it is hard to predict in advance exactly where that broadening must occur: thus the prototypes and trial runs, which have a way of turning up unexpected anomalies. We chose Blake's engraved Illustrations of the Book of Job, one of his masterpieces, because the basic configuration of image and text in the Job plates is different from that of the illuminated books—and thus presents different display problems—and because the images themselves are often very dense and crowded with elaborate details.

Non-illuminated materials present several editorial challenges—from convoluted transcriptions featuring multiple foreign-character sets to large images requiring adjustments to our scaling equations—and several technical challenges. These challenges led us to create an entirely new wing of the Archive. Technical development began in late summer 2001 and was completed in early winter 2002 with the publication of water color illustrations to the Book of Job in February 2002. Work completed during that time laid a foundation for the publication of a broad range of non-illuminated materials, including a set of Blake's engraved Illustrations of the Book of Job; manuscripts associated with Blake's illustrations to Milton's "L'Allegro" and "Il Penseroso"; commercial book illustrations designed by Blake to Robert Blair's The Grave; and water color illustrations to Milton's "On the Morning of Christ's Nativity" and Paradise Regained. These were published in our new preview mode, devised in the interest of publishing the greatest number of high-quality images in the shortest span of time. Like all other items in the Archive, works in preview are in full and accurate color, with enlargements, and with searchable transcriptions of any texts, including even the briefest of inscriptions. The only functions unavailable in preview are image search and Inote.

Meanwhile, we published additional copies of illuminated books in order to clear another hurdle: after publishing one copy of every illuminated book (in 2003 with the 100-plate Jerusalem copy E), to publish at least one copy of every printing of every book. We added America copy O; Europe copy K; The Marriage of Heaven and Hell copy G; Visions of the Daughters of Albion copy P; and The [First] Book of Urizen copies A, B, C, and F. We also added a biography, glossary, and chronology in a new wing of the Archive, About Blake. The biography, by Denise Vultee with the editors, offers extensive background information on Blake's life and works with over 100 accompanying illustrations. The glossary, by Alexander S. Gourlay, treats terms, names, and concepts in Blake that can be mystifying. The chronology covers important dates in Blake's life, including publications and places of residence. The biography and glossary were both encoded in XML to assure conformity with the rest of the Archive. Additionally, we loaded all the data from our image-production records into a new database in order to track our workflow as the Archive expanded.

By the end of 2003 the Archive had incorporated a full range of drawings, paintings, and prints from a rapidly expanding circle of 19 international collections. We had acquired for publication 4415 images, including at least one copy of every illuminated book and in many cases multiple copies, as well as a large number of Blake's paintings, drawings, engravings, manuscripts, and typographical works. Three labor-intensive years of final design work, scanning, encoding, and acquisition of new materials helped to ensure the success of the Archive as originally conceived. In 2003 the Archive received the MLA's Prize for a Distinguished Scholarly Edition, the first time the prize was awarded to an electronic edition.

Phase three, 2004-09: In expanding beyond the illuminated books, we have generally been guided by the following priorities: original works in coherent series that are closely related to the illuminated books already present in the Archive, individual items closely related to those works in turn, works in significant subcategories where we can provide a large and representative sample, the commercial prints (separate and in series, first those designed and engraved by Blake, then those designed by Blake, followed finally by those engraved by Blake), manuscripts (such as letters, unpublished drafts, and Blake's Notebook), and typographical works.

This approach—expanding coherently from the core outward—dominated the first phase of our development and has continued to dominate subsequent phases, as it must if we are to maximize the usefulness of the Archive to students and scholars. We stress, however, that this pattern of growth, useful as it is, has not been adopted out of any sense that the illuminated books are primary and the remaining work secondary. Far from it: one of our major goals is to contribute to a reassessment of such conventional assumptions. In the past, what has counted as primary or secondary among art historians and literary critics has depended far too heavily on the very restrictions (of access, of disciplinary perspective) that we want to reduce if not eliminate in the name of a fuller understanding. The new means of access we have provided, such as powerful image-searching capabilities, are most valuable to users when searches are conducted across related bodies of material. In emphasizing that kind of breadth, we are not ignoring the competing criterion of representativeness—and the tension between depth and breadth can be productive. We believe that our approach stands the best chance of serving both needs at once. That is, the Archive will steadily become more representative as it becomes more extensive but at the least possible sacrifice of utilitarian coherence. We are always aware of the dangers of mere sampling, however extensive, perhaps because the World Wide Web itself offers so many unsettling object lessons.

The third phase of development consolidated and significantly extended the earlier work, expanded the number of contributing collections to 27 and the number of digital images to nearly 6000, added contextual materials, completed a lengthy but necessary conversion of the site from SGML to XML, and set a rigorous publication schedule. We moved to a new base of operations at the University of North Carolina at Chapel Hill and established a manuscript team at the University of Rochester and a close working relationship with Rochester's Blake/An Illustrated Quarterly. In 2005 the Modern Language Association's Committee on Scholarly Editions awarded its Approved Edition seal to the Archive, the first time an electronic scholarly edition received this recognition.

The slow, painful conversion from SGML to XML permitted fundamental changes in site architecture to provide more flexible and sustainable support for the comprehensive Archive we envisioned. In 2006 we launched our XML-based Archive. The conversion involved essential upgrades in technology, an important shift to open-source software, and a conceptual shift in our organization of Blake's works. Our experiences with Blake's engravings, color prints, water colors, and manuscripts led us to reconsider our use of "non-illuminated works" as an overarching category to describe works other than illuminated books. We established a new organizational framework that gives equal weight to various genres and media within Blake's body of work. On the technical end, our new XML-based Archive replaced the SGML and DynaWeb-based architecture that had disseminated the Archive for several years. All works in the Archive (as well as the editors' bibliographic commentary and illustration descriptions) are now encoded in XML and stored in a native XML database powered by eXist. A complex set of XSLT stylesheets transforms these XML documents into the HTML that is then accessed by users, delivered by the Apache Cocoon Web development framework (see Technical Summary).

In 2007 the Archive moved its base of operations from IATH at the University of Virginia to the Carolina Digital Library and Archives (CDLA) at the University of North Carolina at Chapel Hill. This move provided far greater scope for our work. Previously we had to alternate between technical and editorial challenges, and changes registered slowly. UNC provided the resources to move forward on all fronts simultaneously.

That change in pace and scope allowed us to contemplate major changes in site design and to establish an ambitious publication schedule that could be rigorously mapped out well in advance. Publication highlights in this period included Blake's striking series of experimental color printed drawings; his water color illustrations to Thomas Gray's poems, Dante's Divine Comedy (102 drawings dispersed among seven institutions), and the poetry of John Milton; a sketchbook containing drawings for the engraved Illustrations of the Book of Job; a manuscript listing the order of the Songs of Innocence and of Experience; engraved illustrations to Edward Young's Night Thoughts; and engravings in John Gabriel Stedman's Narrative, of a Five Years' Expedition, against the Revolted Negroes of Surinam. The single most telling publication may be Blake's 20 water color drawings for Robert Blair's poem The Grave, 19 of which disappeared between their auction in 1836, their dramatic rediscovery in 2001, and their subsequent resale as individual drawings to several public and private collections. Our scholarly edition of the Grave water colors thus provides not only a means of integrated access to works that will be inaccessible as a group of originals for the foreseeable future but also access to a growing body of related works, such as the engravings (by Blake and Schiavonetti) based on the water colors.

We published numerous significant copies of the illuminated books during this period: America copies F and M; Europe copy H; The Song of Los copies A, C, D, and E; Visions of the Daughters of Albion copies A, B, and O and proof copy a; Songs of Innocence and of Experience copies A, B, T, V, and Y; The [First] Book of Urizen copy D; Milton a Poem copies A and B; The Marriage of Heaven and Hell copies K, L, and M; and The Book of Thel copies L and R.

We continued to augment and refine our contextual materials. The About Blake section expanded to include Joseph Viscomi's "Illuminated Printing," which originally appeared in The Cambridge Companion to William Blake (2003). This electronic version features additional images that enhance the analysis of Blake's innovative printing techniques. We revised Editorial Principles in About the Archive and published redesigned and updated bibliographies and collection lists in our Resources for Further Research section. These resources allow easy access to information about Blake, his works, and Blake holdings in major collections. Blake scholar Mark Crosby (Queen's University Belfast) became our official bibliographer.

In its inaugural years, the Archive's new manuscript team at the University of Rochester has updated many documents in the Archive, but the team's efforts have focused on a problematic category of Blake's work, manuscripts and typographical works, for which we have developed and tested a compact XML tagset derived from version P5 of the Text Encoding Initiative (TEI) Guidelines. This tagset allows concise but detailed markup of manuscript features such as cancellations, additions, and substitutions.

To ensure that manuscripts are fully searchable, we use additional TEI elements (<choice>, <orig>, and <reg>) to regularize abbreviations and unconventional spellings. We have also developed a highly legible color-coding system capable of displaying at a glance most features of Blake's manuscript revisions without the clutter of other symbols—reserving editors' notes for special cases.

Rochester is also home to a second initiative, the incorporation of Blake/An Illustrated Quarterly (1968-present) into the Archive, to give scholars an unprecedented fusion of cross-searchable "primary" (editorial) and "secondary" (critical) scholarship. After two years of preliminary research and consultation, we have produced a plan for publication that will draw on the open-source XML tagsets created by Digital Humanities Quarterly and such impressive e-journal prototypes as DHQ, Tate Papers, and RaVoN (formerly RoN).

Top | Significance | History | Present and Future | Dissemination

3. Present and Future

We continue to depend upon the system of close collaboration we have developed over the years: collectively, the editors make final decisions about the form and content of the Archive and control the workflow. At the University of North Carolina, Viscomi takes major responsibility for acquiring, scanning, and color-correcting new images on professional equipment designed for the purpose. Until 2002 Viscomi also transcribed and edited most of the texts in the Archive. These duties are now performed by Eaves, at the University of Rochester, and Essick, at the University of California, Riverside, who also produce XML-encoded descriptions of images and bibliographical information at various levels (plate or object level, copy level, series or work level, and so on). Day-to-day progress is coordinated by our project manager, currently Ashley Reed. For testing, all the editors collaborate with CDLA staff members—who are joined by members of the advisory board.

Since editorial decisions always have technical implications, they must be arrived at in consultation with the expert staff at CDLA and our technical editor, William Shaw. The editors participate actively in discussing all details of execution within their technical reach and test the practical outcome on the Archive's work-in-progress site. Most of those ongoing conversations are conducted daily (and vigorously, sometimes for weeks at a time) on the project's electronic forum, blake-proj, supplemented by telephone, the annual Blake camps (see above, History), and meetings in person as needed. Student assistants collaborate closely with the editors who oversee their work. The proof that this arrangement has worked well in moving the project along swiftly and efficiently is manifest, we believe, in the record of productivity to date. As we add new categories of Blake's work to the Archive, we shall continue to require significant technical support from CDLA.

The Archive's first phase was defined by three goals: to design and construct the foundations of a searchable, SGML-encoded Archive; to acquire major works, mostly illuminated books, from major collections required for the basic structure of the Archive; and to place in the Archive, fully marked up and publicly accessible free on the Web, at least one copy of each illuminated book. We achieved the last of these goals in 2003 with the publication of Jerusalem. In our second phase, we accomplished the subsequent goal of placing multiple copies of illuminated books in the Archive whenever possible, with the focus on those copies that represent different printings of each book. We continued this expansion during the third phase. Since books printed in the same session can differ significantly—with important variants in coloring, motifs, arrangements, etc.—we have included multiple copies of books from the same printing as well as those from different printings. The public's exposure to Blake—including many advanced students and not a few scholars— has been narrowly restricted to a small number of items that have been too frequently reproduced, such as Songs of Innocence copy B. We shall continue to acquire and incorporate copies never before reproduced or poorly reproduced—a category that unfortunately includes all but a very few of the books—thereby making rare and unique material widely accessible, in many cases for the first time. By mid-2010 the Archive had published electronic editions of 75 copies of Blake's 19 illuminated works. (Explanatory updates for all publications are available; see Blake Archive Updates.)

But the preoccupations of our day-to-day work shifted in the second and third phases to works in other categories: prints of various kinds, drawings, paintings, manuscripts, and typographical works. We began phase two with a large number of transparencies from these categories—about 750 ready to be digitized, corrected, marked up, and placed in the Archive at an appropriate time. To this number, we added over 2000 transparencies and digital images of illuminated and non-illuminated works from nine collections, including 1263 from the British Museum (the world's largest Blake collection), and the entire Blake collections of the Fogg Art Museum: the Museum of Fine Arts, Boston; the National Gallery of Art, Washington; the National Gallery of Victoria; the Metropolitan Museum of Art; the Birmingham Museums and Art Gallery; the Victoria and Albert Museum; and the Whitworth Art Gallery—far more than we initially anticipated. At the end of our third phase, we had published 25 scholarly editions of non-illuminated works representing each of the categories listed above. At the end of phase three, we added approximately 400 more images of works from all categories. We begin phase four with nearly 6000 images from 27 contributing collections—all the major Blake collections in the world. We hope to acquire another 500 or more images over the next five years. Archive editions are published in the context of full bibliographic information about each work, scrupulous diplomatic transcriptions of all texts, detailed descriptions of all images, and extensive bibliographies.

Phase four, 2009-13: Our fourth phase of development is defined by five ambitious goals that bring us closer to the founding vision of the Archive:

Goal 1: Design and implement an entirely redesigned user interface enriched with superlatively useful and unprecedented features, such as Related Works, by which a user can quickly identify and access other works in the Archive related to the one under consideration—finished water colors based on one preliminary drawing, for example, or an unrevised draft for a printed poem—and sophisticated, robust new digital tools, such as the Virtual Lightbox.

In 2009 the Archive added a new set of scholarly tools, known collectively as our Related Works system, that are designed to show relationships among works and individual objects in the Archive. They function at two levels. First, work index pages now include active links to related materials in the Archive. So, for example, researchers can move easily from any given series of illustrations to the Book of Job to any other series. Second, the Show Me menu on object view pages now includes "Related Works in the Archive." Like the work-level menu, this list includes active links to related designs and is meant to allow study of the related materials side by side. For example, researchers can move quickly from Blake's engraving of Behemoth and Leviathan to his drawing for it in the sketchbook and study how Blake developed his design from preliminary sketch to finished engraving. The first publication in the Archive to use this feature was Blake's illustrations to Mary Wollstonecraft's Original Stories from Real Life.

The Lightbox, developed by Matthew Kirschenbaum in collaboration with the [University of] Maryland Institute for Technology in the Humanities, is an open-source browser application that will allow our users, guided by the familiar "shopping cart" metaphor, to collect and compare images from across the site at high resolution, manipulate them for detailed study, and access image descriptions. The Lightbox will replace Inote, our original image-annotating tool.

The Lightbox allows users to zoom and crop images, making it easy to study specific details and motifs. Users can arrange images simply by dragging them around the workspace; this freedom encourages juxtaposition and comparison. Furthermore, because users can send any object in the Archive to the Lightbox, it is possible to collect items from across genres, media, and time periods. Users may add objects directly from object view pages or image search results, or they may right-click any object in the Lightbox to call up a menu of other objects in the Archive. In addition to zooming and cropping, several other features make the Lightbox a powerful tool for studying objects in the Archive. As on the object view page, users can display images at true size, read image descriptions, and view metadata about the image itself. Other toolbar buttons link to the object view page of any open image, delete the currently selected image from the workspace, undo previous actions, open help documentation, and tidy the workspace by snapping images to a grid. These images are not merely displayed at correct scale relative to one another; the Lightbox detects the user's screen resolution and, based on that information, delivers images at actual size.

Although, with minor revisions, our present interface has served our users well, a complete overhaul is past due, but is a daunting task. In 2007 we began the process by mounting on our testing site, for internal review and comment, the draft prototype of a revised interface. Since then we have refined our original ideas as we have continued to develop tools and features that can offer users a heightened experience of the Archive's ability to support advanced humanities research.

Goal 2: Move Blake's unique manuscripts and rare typographical works from early development into our normal production schedule until we are in a position to publish scholarly editions of The Four Zoas and Blake's Notebook, two extremely significant, complex, and fragile manuscripts in the British Library.

Text editing has always played a major role in the Archive's work, and a statement of the conservative, "diplomatic" or documentary, object-oriented principles that govern our practices was among the first documents we published (see Editorial Principles). We recognized that our first undertaking, the texts of the illuminated books—which Blake etched and printed for reproduction in multiple copies—would be an appropriate first step. Those mostly straightforward texts present fewer editorial challenges than his longest manuscripts, which have defied the best efforts of his most resourceful and disciplined editors. We settled on a progressive approach: we would venture out step by step from the textual rudiments of the illuminated books through the increasingly steep terrain of manuscripts arranged according to length, difficulty, and coherence with our publishing program (see History, phases two and three).

As we acquired images of visual art from major international Blake collections, we simultaneously acquired images of his manuscript pages. That process climaxed with our success in obtaining the British Library's permission to publish its Blake treasures—among them The Four Zoas, a vast "prophetic" work central to the development of Blake's ideas, which he left in draft after a decade of intensive artistic effort, and his working Notebook (with contributions from his inspiring younger brother Robert, who died of tuberculosis in 1787), a rich trove of visual and literary artistic ideas highly resistant to traditional scholarly editorial approaches. The British Library granted rare access to the highly restricted and delicate Four Zoas manuscript on site for several days and then supplied impressive new digital photography of all their Blake manuscripts. Their file collection of Zoas photographs had deteriorated badly; the new reproductions will be the finest ever published.

Textual transcription is far more complicated when dealing with manuscripts than with illuminated books. In 2006 the University of Rochester Department of English agreed to sponsor an Archive team to specialize in text editing. Grounded in the preliminary editorial work by Archive editors and assistants on Blake's manuscript satire An Island in the Moon (Fitzwilliam Museum), the Rochester team has developed a sophisticated XML tagset tailored to the needs of Blake's manuscripts and to the fundamental principles of the Archive, plus a highly legible and simple color-coding system (using XSLT and CSS) capable of displaying most of Blake's manuscript alterations and eliminating the clutter of conventional textual signs and symbols. The Rochester group has further strengthened the explanatory power of the editors' notes with zoomed images of textual cruxes.

This team is currently extending the Island system to a series of Blake's most revealing letters, another large and dispersed set of unique documents, to be published in coherent groups. We are also extending the testing to a second literary work, the so-called Pickering Manuscript (Morgan Library), which includes some of Blake's most mature lyric and narrative poems in fair copy. Soon after will follow such manuscript works as Tiriel, with its extant drawings, and Blake's marginalia (editorially difficult because they combine manuscript and print on the same physical support). As we establish a strong track record of text editing, we will start work on The Four Zoas. The Zoas, we know, will be hard and time-consuming. So we will divide the text editing into two streams, one for the Zoas and the Notebook (which includes, for example, such key works as "A Vision of the Last Judgment," "Public Address," and "The Everlasting Gospel"), the other for a host of smaller but highly significant projects, including the few conventionally printed works that Blake authored (principally The French Revolution, Poetical Sketches, and the Descriptive Catalogue for his one-man exhibition of 1809).

The Archive will continue to pursue its original mission of publishing exemplary copies from every printing of Blake's 19 illuminated works and completing the full production histories of as many books as possible. We will also continue to add non-illuminated works—engravings, drawings, water colors, paintings, and color prints—which typically have minimal texts but contain complex images that can be extraordinarily hard to describe. Currently, most non-illuminated works are published in preview mode, which makes them available to scholars and students with all the features of the Archive except for image search and Inote. But we have begun to republish these works in full mode with searchable images, and to publish new non-illuminated works in full mode, bypassing preview altogether. In late 2009 we republished three series that were previously published in preview mode: the Thomas Butts and John Linnell series of Blake's water color illustrations to the Book of Job and the sketchbook of preliminary pencil drawings for his Job engravings. In the coming months, we will publish similarly updated versions of other works presently in preview mode, including Blake's nine series of water colors illustrating the poetry of John Milton. Blake's sketches and engravings for Mary Wollstonecraft's Original Stories from Real Life have already been published in full mode, and several other new non-illuminated works—the manuscript of An Island in the Moon, A Small Book of Designs copy A, and A Large Book of Designs copy A—are scheduled for publication in full mode in 2010.

Since the Archive's second phase, we have been publishing Blake's work in media other than illuminated printing and altering the Archive's architecture and programming structure as needed. An Island in the Moon, for instance, has required a transcription markup tagset specific to Blake's manuscript works. The publication in preview mode of 20 water color drawings required the creation of a virtual group protocol that allows many individual works to appear on the same dynamically generated table of contents even when those works are not copies of the same illuminated book. Such innovations balance the integrity of the Blakean "work" and individual "object" (the fundamental building blocks of the Archive) with the site's ease of use. A Small Book of Designs has inspired changes to the Archive's Compare feature, originally designed for use with illuminated books but now modified to include works in other media. And the recent publication of sketches and engravings for Mary Wollstonecraft's Original Stories from Real Life prompted the implementation of our new Related Works feature.

Goal 3: Acquire approximately 500 new images over five years from 30 institutions (25 of which are contributing works for the first time) in line with the fundamental program of expansion that has coordinated our efforts since we first articulated it during phase two.

Our overarching aim is to incorporate, for the first time, a history of Blake's artistic production into an edition capable of representing the full range of his artistic achievement. Our starting point, Blake's greatly admired, intensively studied, but fragile and widely dispersed series of illuminated books, has served as the archival and editorial backbone for the project. It supports a twofold strategy: to evolve along lines that maximize both the broad coverage of Blake's work, across multiple decades and various media, and deep scholarly coherence. We maintain coherence by expanding the core with works that are closely related (historically, thematically, physically) and by giving priority to significant interrelated clusters (clustered by medium, such as the large color prints of 1795, or by subject or theme, such as the Job and Milton illustrations in several media). Whenever possible, we assemble these individual clusters into larger ones. For instance, we are able to represent Blake's work as a printmaker in several graphic media more coherently than previously possible by expanding outward from the core of illuminated books (typically water colored relief etchings) to the other works designed and engraved by Blake, and then to those designed by him but engraved by others, and finally to those designed by others but engraved by him. The Archive's Related Works feature, which dynamically links works in different media across the Archive, supports the twofold goal of broad coverage and scholarly coherence: as we publish new acquisitions, Related Works will make their relationship to core works clear, allowing students of Blake to consider and compare works in different media in ways that have until now been difficult if not impossible.

Goal 4: Describe and encode all Archive images to make image searches—that is, deep searches on the actual subject-content of images at a very fine level of detail—function across all genres and media, making it possible for the first time to study Blake's works through text- and image-searching tools that harness the combined power of superbly detailed, accurate reproductions; a unique keyword lexicon; meticulous image descriptions; and innovative software.

Blake scholars are intensely interested in the precise composition and the semantic content of his images, but searching for those details in books and dispersed originals has been extremely difficult. Training dedicated graduate students to deploy our keyword vocabulary has made this work increasingly efficient; recruiting and retaining these assistants is part of the Archive's long-term plan. Ongoing improvements to the Archive's image-search capabilities include an expanding visual vocabulary and technical improvements such as thesaurus-based searching.

Goal 5: Complete the work, begun in 2009, of incorporating Blake/An Illustrated Quarterly (1968-present)—for over four decades the journal of record in Blake studies and the publisher of an impressive body of articles and notes, bibliographies, discussions, and reviews—to create an unprecedented synthesis in humanities scholarship.

This amplified Archive will be, we believe, the first major electronic resource in the humanities to integrate a major scholarly journal. All images ever printed in the quarterly (typically mediocre black-and-white halftones) will be linked to their vastly superior counterparts in the images of the Archive and, of course, to all the tools, features, and information of the Archive (deep image and text searches, for example). Conversely, all works in the Archive will be linked to related publications in the quarterly. A researcher interested in Blake's Jerusalem can access our editions of Jerusalem, related works, and their bibliographies, but also the quarterly's Jerusalem scholarship, such as Aileen Ward's and V. A. De Luca's lucid essays on the complex history of its composition. An art historian tracing the fascinating history of the recent discovery and subsequent sale at auction of 19 of Blake's water color designs for Blair's The Grave will have access (through the quarterly) to Robert Essick's annual accounts of Blake sales, G. E. Bentley, Jr.'s annual checklists of research, and Martin Butlin and Robin Hamlyn's comprehensive art-historical account of the water colors—as well as to the Archive's meticulous editions of the water colors themselves. A user interested in the methods by which Blake produced his stunning series of color printed drawings of 1795 can examine the scholarly editions of those drawings in the Archive while, in the quarterly, following the unexpectedly fierce scholarly debate over Blake's color-printing methods, which would have been nearly impossible for Robert Essick and Joseph Viscomi to document without their many enhanced enlargements of details from Archive images. The "pages" of all past issues of the quarterly will appear in a streamlined, uniform electronic format designed to integrate seamlessly with the rest of the Archive, while archival PDFs of all the original printed pages of the journal will also be available.

The Archive as scholarly resource: Since its inception, evidence of the use of the Archive to sustain serious scholarship—our stated objective—has continued to accumulate. For scholars, publishers, and others, the Archive has increasingly become a standard source of images and transcriptions of Blake's works for republication. In his multimedia essay "Golgonooza Text" (2005), Nelson Hilton calls the Archive "an indispensable resource without which this presentation—for one example—could not exist" (par. 5). Thomas Pfau ties his discussion of The [First] Book of Urizen to our electronic edition of copy G in his book Romantic Moods: Paranoia, Trauma, and Melancholy, 1790-1840 (2005). And the latest Norton Critical Edition of Blake's works (2008) uses the Archive's transcriptions of the illuminated works and reproduces images provided by the Archive (see Articles about the Archive).

The Archive has created new opportunities for scholarly discoveries. Two definitive essays on Blake's "color printing"— Robert Essick and Joseph Viscomi's "An Inquiry into Blake's Method of Color Printing" and "Blake's Method of Color Printing: Some Responses and Further Observations," in Blake/An Illustrated Quarterly (winter 2002 and fall 2002, respectively)—would have been unimaginable without the array of evidence from the Archive images. In the ensuing debate about Blake's methods, Blake scholars Michael Phillips and Martin Butlin responded to these findings in the fall 2002 issue of the quarterly. The following year, Essick and Rosamund A. Paice announced, in the spring 2003 quarterly, nine previously unknown drawings discovered in the British Museum by the Archive editors when they photographed its collection. The number of articles on the Archive project itself continues to increase (see Articles about the Archive).

The Archive's technical significance is signaled by the first dissertation inspired by the Archive, Vladimir Misic's "Mixed Raster Content for Processing of Colored Engravings" (Department of Electrical and Computer Engineering, University of Rochester, 2003). The dissertation involved active collaboration with Xerox Corporation. Rob Buckley of Xerox presented a paper on the subject, "Document Imaging on the Web with MRC [Multi-Raster Content, a Xerox technology] and JPEG 2000," at the invitation-only EVA (Electronic Imaging and the Visual Arts) Harvard Symposium, October 2-3, 2002.

The Archive has also achieved recognition for its status as an online humanities archive and a scholarly edition. EDSITEment—which, under the auspices of the NEH and affiliated organizations, assesses and compiles online humanities resources—has named the Archive one of the "Best of the Humanities on the Web" in its Literature and Language Arts category. The Archive was also selected as "a groundbreaking hypermedia project" in the humanities by the Charles Babbage Institute, the foremost archive for scholarly internet sites, which has acquired the Archive's early project records (1995-2002) for future study by scholars interested in the history of information technology. When the Archive received the 2003 Prize for a Distinguished Scholarly Edition by the Modern Language Association, the selection committee's citation affirmed that "the William Blake Archive has set a high mark for future editorial practice through its clarity, user-friendliness, beauty, and erudition." In 2005, the Archive became the first electronic scholarly edition to receive the Approved Edition seal from the Modern Language Association's Committee on Scholarly Editions.

Top | Significance | History | Present and Future | Dissemination

4. Dissemination

The full resources of the Archive are available to all who can access it via the World Wide Web. We are committed to the continued development and maintenance of a free site, for reasons we have outlined above. From the start we have encouraged feedback from users via an e-mail link (we routinely respond to inquiries and comments). By filling out a form at the site itself, users can also subscribe to Blake Archive Updates and automatically receive e-mail notices with the latest information. These notices are regularly posted to several online discussion groups as well.

One of the foremost advantages of electronic publication is the ability of the medium to accommodate growth and change. In one sense, then, a final product and date of publication never arrive, at least not as they do in the world of print. That said, the Archive has expanded very nearly to its full intended shape. The intricate XML architecture and stylesheets required to integrate and display the full range of Blake's works in all media have been developed, and the Archive contains objects from each category of Blake's works.

Since moving to the University of North Carolina, we have increased the number of publications per year from four to eight, which we will continue over the next five years. Beyond the Archive itself, we disseminate our findings in lectures, conferences, and published reports (see Presentations and Demonstrations about the Archive). The Archive is a frequent subject in discussions of humanities computing, editorial theory and method, and the future of scholarly communication. We increase our outreach through participation in discipline-based organizations, such as the MLA's Committee on Scholarly Editions, and in initiatives such as NINES.

Because the Blake Archive has become a central scholarly resource, we have recognized our role as a liaison among scholars, institutions, and private collectors. Scholars needing the best reproductions and texts for their own published research now come to us frequently, and, in cooperation with our contributing collections, we supply their needs whenever possible. Incorporating the journal of record in Blake studies, Blake/An Illustrated Quarterly, into the Archive in the near future will further strengthen the circuit of scholarly communication in our field. Altogether, this augmented William Blake Archive aims to realize our original vision by setting a new standard of accessibility to a vast array of visual and textual materials that are central to a thorough grasp of the art and literature of the eighteenth and nineteenth centuries.

Top | Significance | History | Present and Future | Dissemination

About the Archive

Blake Archive Homepage