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. "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 studied, exhibited, and collected of British artists: in April 1999, one copy of his illuminated Book of Urizen (24 plates) sold at auction for $2.5 million, widely noted to have been the most ever paid for a work of British "literature," and in May 2006 some of his watercolors for Blair's Grave fetched high prices at auction. In the last ten years alone large shows have been mounted in Australia, England, Germany, Israel, Japan, Scotland, Spain, and the United States, including major exhibitions in Melbourne (1989), Tokyo (1990), Barcelona (1996), Madrid (1996), and Melbourne (1999). 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.

"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 the latest books on the meaning of science, technology, history, 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 in the 20th century--more often than any British writer other than Shakespeare. There are significant Blake ballets (Ralph Vaughan Williams again), 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. 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 uses now extend 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. Representative works include illustrations 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.

In the broadest terms, the William Blake Archive, an archive of electronic editions available free as a site on the World Wide Web (http://www.blakearchive.org), 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 will soon 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' detailed image descriptions and innovative 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 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. 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 Complete Poetry and Prose of William Blake, the standard printed edition for reference.

Meanwhile, the roster of major contributors has grown: the Library of Congress (now a sponsor) has been joined by the Huntington Library and Art Galleries; the Essick Collection; the New York Public Library; the Pierpont Morgan Library, New York; the Houghton Library and Fogg Museum of Art, 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 New York; the Museum of Fine Arts, Boston; the Birmingham Art Museum; the Whitworth Art Gallery, Manchester; and the Victoria and Albert Museum. Discussions are in progress with the only two major Blake collections not in the Archive. Our success to date in obtaining the confidence and good will 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, and from the Getty Grant Program, which provided major funding for the first phase of design and construction, with additional grants from the Paul Mellon Centre for the Study of British Art (London), and the University of North Carolina, Chapel Hill. Hardware and software contributions have come from Sun Microsystems and Inso Corporation.

The 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 a technological formation sufficiently revolutionary to alter some fundamental assumptions in scholarly editing; the emergence of sound 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 an organization specifically devoted to giving technological form to the ideas of humanists. Together these four events 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 this 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" (the 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 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. The stage was thereby set for a new phase of radical editorial revision--but revision of a kind that would be almost unimaginable in the medium of print.

The 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.

The 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 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.

The 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 then-new (and still unique) 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 useable 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
• 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. Portability makes disks a popular medium in classrooms and on desktop computers. But for scholars doing sustained original research, there is 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 have now severely limited even scholars' access to the fragile originals. This includes at least two of our contributors, the Huntington and the Fitzwilliam. Even registered scholarly readers at the Huntington, for example, must get special permission from the Associate Curator of Early Printed Books to see a single illuminated book. Both institutions recently 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 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, 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 have heretofore been imperceptible without close first-hand scrutiny of the dispersed 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, 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 have never been reproduced before. This constituted, for the first time, an archive of reproductions suitable for serious research. Only in the past five years, 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 has made possible for the first time a sound scholarly archive, including numerous copies of the illuminated books that have 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 is yielding 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 has already begun to reveal the painter-printmaker whose illuminated books emerged from the materials, work routines, and imagery of eighteenth-century history painting, watercolor 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 is deepened by the inclusion of Blake's extensive group of literary illustrations, such as those to Milton's poems, which should be seen in connection with Blake's 50-plate illuminated book Milton.

We are hoping, of course, that the Archive, once extended to encompass the full range of Blake's work, including his many illustrations of the work of other writers, will ultimately set a new standard of accessibility to a vast collection of visual and textual materials that are central to an adequate grasp of the art and literature of the eighteenth and nineteenth centuries. But we have also 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 unique 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.

The 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. 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 have begun to develop, which we hope will become useful prototypes of "distance editing," depend upon the intensive day-to-day teamwork among the three editors and the staff of IATH 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 project will leave future archivists, editors, and cataloguers the benefits of new tools. The Document Type Definition (DTD) developed by the Text Encoding Initiative (TEI) has become the default standard for encoded electronic texts. We believe that the Blake Archive DTD has the potential to become the standard for encoded electronic images that the TEI is 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 readily adapted 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, will be 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, 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, texts), 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, 1995-2000. 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, "Collaboration Takes More Than E-Mail: Behind the Scenes at the Blake Archive". Journal of Electronic Publishing 3.2 [Dec. 1997]: http://www.press.umich.edu/jep/03-02/blake.html).

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, 4x5 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, our 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. "Machine 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, 1973-1985.

Machine 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 its disadvantages are major. It is little known among art historians, and its status and stature even in the image-resource community (of librarians and curators) are insecure. People who mention it seldom, if pressed, seem to understand it. Users complain that the codes are hard to remember--they have to be tracked down laboriously in the printed volumes or in the electronic browser--and that new codes are too difficult to add, making the system less flexible than it needs to be. Very few projects use it intensively--and none as intensively as we would be using it. Some collections (of slides or art objects) use Iconclass codes--usually only a few--to classify the pictures in their collections, but such classifications are very different from our intricate descriptions of individual images and their numerous components. Despite the aspiration of Iconclass to become a universal system of image description, its origins are still evident in its disproportionate emphasis upon a highly conventional Renaissance iconography.

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 interested in and open to those original alternatives, however. We plan to look into the possibility of integrating Iconclass codes with Archive codes once our vocabulary is set; and we recently reviewed pattern-recognition searching with a Kodak engineer, who suggested that it could be useful in searching for colors, textures, and techniques. We intend to remain in the vanguard of efforts to create working solutions to these problems. (The HyperIconics Project lists the Blake Archive as one of its "exemplary sites"--a special tribute, we feel, considering that the project backers, at the University of Leiden, are among the heirs of the Iconclass legacy.)

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 in late 1999.)

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 editor's 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 special 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 4x5 transparencies as we approached other major repositories.

Years 3-4, 1997-99: Whatever remained unresolved in our usual forums--the test 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 1997-1999 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" makes available a large body of documentation and supplementary material: The Archive at a Glance (for a quick overview), Editorial Principles and Methodology, 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. 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, if we are not mistaken, ever given by the Mellon Centre in support of an electronic project. We gave public presentations of the project on several occasions; a well-attended session on Romanticism at the annual MLA meeting (San Francisco, Dec. 1998) eventuated in the publication of five papers on the Archive (The Wordsworth Circle, 1999, see Articles about the Archive Articles about the Archive).

Year 5, 1999-2000: We spent the year consolidating our gains and preparing to initiate, in 2000-2003, 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 Blake's Complete Poetry and Prose, 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 is creating JPEG 2000, a new version of the JPEG standard by which the images on our site are generated. (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, including Dublin (Ireland), Charlottesville (VA), and Halifax, Nova Scotia.

Blake Archive, Phase Two, 2000-2003. As our initial 3-year funding from the Getty Grant Program allowed us to initiate the first (5-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. As we extend the reach of the Archive to accommodate such works, we shall also begin to build a statistical database of information about the uses of the Archive and develop a method of assessing users' responses.

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 the ultimate 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 must deal with. The task at hand has been to anticipate changes in the DTD and stylesheets, interface, image descriptions, art-historical and textual information, and arrive at an initial set of blueprints that allowed us to add new subcategories of Blake's work to the Archive.

Along with the redesign and extension of the DTD come problems of rendering and description--how to render images across various new categories in ways that will 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 will require 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 shall have 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 present controlled vocabulary has been developed.

We began the shift from illuminated works to the other categories with two large and significant bodies of material: Blake's illustrations of the Book of Job and of John Milton's poems, in several different media (almost 200 items). (Users regularly write us asking for Blake's designs to Job, Dante, and Milton.) 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 constitute a logical, large, but manageable first extension of what we have already done. The Huntington, one of our contributors, holds the world's largest collection of Blake's illustrations to the works of Milton. In addition, we shall include all of Blake's illustrations to Edward Young's Night Thoughts (537), Thomas Gray's Poems (117), and Dante's Divine Comedy (101).

From these clusters we shall move to others, generally 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), and typographical works. (See Present and Future.)

This approach--expanding coherently from the core outward--has dominated the first phase of our development and has continued to dominate the second, 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 we started with 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 relax 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 terrifying object lessons.

By the end of 2003, the Archive contained approximately 4200 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. A significant proportion of these will be fully tested and publicly available. Phase two will continue well beyond these three years, certainly. But three labor-intensive years of final design work, scanning, encoding, and acquisition of new materials will have helped to ensure the success of the Archive as originally conceived.

Top | Significance | History | Present and Future | Dissemination

4. Present and Future

We continue to depend upon the system of close collaboration we have developed during the past several 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 SGML-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 Justin Van Kleeck. For testing, all the editors collaborate with all IATH 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 IATH and our Technical Editor Andrea Laue (University of Virginia). 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 of course by telephone, the annual Blake camps (see above, History), and meetings in person as needed. Any 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 IATH, whose configuration of expertise, like its mission in the humanities, is unique.

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. We are now very far along toward 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. But since books printed in the same session can differ significantly--with important variants in coloring, motifs, arrangements, etc.--we are including multiple copies of books from the same printing as well as those from different printings. The public's exposure to Blake--and this includes 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.

But the preoccupations of our day-to-day work have shifted 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 have in the past three years added over 2000 transparencies and digital images from nine collections, including 1263 from the British Museum (the world's largest Blake collection), and the entire Blake collections of the Fogg Museum of Art, the Museum of Fine Arts, Boston, the National Gallery, Washington, the National Gallery of Victoria, the Metropolitan Museum, the Birmingham Museum and Art Gallery, the Victoria and Albert Museum, and the Whitworth Art Museum, Manchester--far more than we initially anticipated.

The process of acquiring and editing the non-illuminated works has been generally guided by these priorities:

I.a: Large clusters of significant works, such as Blake's Job and Milton illustrations, that are most closely related to the illuminated books and to one another. This step maintains the core coherence in the Archive as it grows and at the same time provides scholars and students with the range of materials they have found most valuable for study.

I.b: Individual works that are also related to the illuminated books, such as proofs, separate colored impressions, commercial engravings, etc.

II.a: Works in subcategories of which we can provide a large and representative sample, such as the famous group of twelve large "color prints" of 1795.

II.b: The entire group of original and reproductive prints, first those both designed and engraved by Blake (such as the engravings for Young's poem Night Thoughts), followed first by those designed by Blake and engraved by others (such as the illustrations to Blair's poem The Grave), and finally by those engraved by Blake after other artists.

III: Manuscripts and typographic works, all either unique or very rare.

On the public side, our main accomplishment of the second phase has been the publication of our first two "non-illuminated" works including The Book of Job watercolors (Butts copy), watercolors and manuscripts associated with Blake's illustrations to Milton's L'Allegro and Il Penseroso, and illustrations designed by Blake to Robert Blair's The Grave. 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.

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: publication of non-illuminated materials required 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 the Job watercolors 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 engravings, Illustrations of the Book of Job, published in December 2002.

Along with the wing of non-illuminated works, we also added a biography, glossary, and chronology in the "About Blake" wing. The biography offers extensive background on Blake's life and works, and even on his acquaintances, all with over 100 accompanying illustrations ranging from works by Blake and others to pictures of Blake's printing methods. The glossary treats terms, names, and concepts in Blake that can be dizzying for his audience. The chronology covers important dates in Blake's life, including publications and places of residence. The biography and glossary are both encoded in XML.

During this same period we have continued to add illuminated works to the Archive. We now have 50 copies of all of Blake's 19 illuminated works. In addition to America copy O and Europe copy K, which we published in March 2001, we have published the following illuminated works: The Marriage of Heaven and Hell copy G, Visions of the Daughters of Albion copy P, and The Book of Urizen copies A, C and F, and Jerusalem copy E. (Explanatory updates for all publications are available; see Blake Archive Updates.)

Since its inception, evidence of the use of the Archive "to sustain serious scholarship"--our stated objective--has continued to accumulate. Three recent examples are Thomas Pfau's decision to tie the discussion of The Book of Urizen in his new book on German and British Romanticism to our electronic edition of copy G instead of to a printed reproduction; and two definitive essays on Blake's "color printing" that would be unimaginable without the array of evidence from Archive images (Robert N. Essick and Joseph Viscomi, "An Inquiry into Blake's Method of Color Printing" and "Blake's Method of Color Printing: Some Responses and Further Observations," Blake/An Illustrated Quarterly, Winter 2001/02 and Fall 2002, respectively)--see www.blakequarterly.org. An article announcing 9 previously unknown drawings discovered in the British Museum by the Archive editors when they photographed its collection is forthcoming in Blake/An Illustrated Quarterly. John E. Grant and Mary Lynn Johnson have recently requested permission to use Blake Archive texts as the basis of their revised version of the printed Norton Critical Edition of Blake. For scholars, publishers, and others, the Archive has increasingly become a standard source of images and transcriptions of Blake's works for republication. Over the past year, the number of requests for use of Archive materials has increased dramatically.

The number of articles, reviews, and conference papers on the Archive project itself continues to increase. Recent examples include J. Hillis Miller's "Digital Blake," presented at the fall 2000 conference of the Digital Cultures Project of the University of California, a multi-campus research group, and first published on the Project's web site, Stuart Peterfreund's essay-review in the European Romantic Review, and papers in Sheila Spector's session on "Blake Scholarship and the New Technology" at the annual MLA meeting (December 2002). Recent interviews and articles on the Archive by the editors on the Archive include:

"'Once Only Imagined': An Interview with Morris Eaves, Robert N. Essick, and Joseph Viscomi on the Past, Present, and Future of Blake Studies." Conducted by Kari Kraus. Studies in Romanticism. (Fall 2002), with an online version at Romantic Circles http://www.rc.umd.edu/indexjava.html.

Eaves, Essick, Viscomi. "The William Blake Archive: The Medium When the Millennium is the Message." Romanticism and Millenarianism. Ed. Timothy Fulford. London: Palgrave, 2002. 219-33.

Viscomi. "Digital Facsimiles: Reading the William Blake Archive," Computers in the Humanities 36.1 (February 2002). 27-48 + 17 illus.

The Archive's technical significance is signaled by the first dissertation inspired by the Archive, Vladimir Misic's "The Segmentation and Compression of Images in the Blake Archive" (in progress, Dept. of Electrical and Computer Engineering, University of Rochester, due for completion in spring 2003). Most recently, the dissertation has 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, 2-3 October 2002. This collaboration shows promise of leading to further developments, including a working demonstration of Misic's innovative image-compression system--tailored to the problems of reproducing Blake's images--on the Blake Archive site.

The Archive has also achieved recognition for its status as an online humanities archive. EDSITEment (http://edsitement.neh.gov) chose the Archive as one of the "Best of the Humanities on the Web" in its Literature & Language Arts category. EDSITEment brings together leading humanities resources, as selected by the National Endowment for the Humanities in partnership with the National Trust for the Humanities and the MarcoPolo Education Foundation. The Archive was also selected as "a groundbreaking hypermedia project" in the humanities by the Charles Babbage Institute. The Babbage Institute, the foremost archive for scholarly internet sites headquartered at the Center of Information Technology, University of Minnesota (http://www.cbi.umn.edu), has acquired the Archive's project records and will be its official archivist. Most recently, the Archive received the 2003 Prize for a Distinguished Scholarly Edition awarded every two years by the Modern Language Assocation. This is the first time that the award has been given to an electronic edition. 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."

Meanwhile, one of our consultants, Matthew Kirschenbaum, is collaborating with the [University of] Maryland Institute for Technology in the Humanities on the development of an exciting new image viewer, the Virtual Lightbox, which will allow our users, guided by the familiar "shopping cart" metaphor, to collect images from across the site and import them into a display area where they can be freely arranged for comparison and manipulation. Virtual Lightbox 1.0 is ready for implementation, with additional features and functionality under development; we hope to incorporate it into the Archive as a Java applet in the near future.

Much of the work of the second phase has been behind the scenes, making editorial changes, creating prototypes, and running tests. These are precursors to basic revisions of the elements in the Blake Archive DTD (document text definition)--the SGML (standard generalized markup language) recipe, as it were, on which the Archive rests. 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 are also using new transcription standards in the display of several of Blake's works. These new standards attempt 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 have noticeably improved access times, which will be increasingly important as the Archive continues to expand the number of works on the site.

We have now entered a new phase of active production and publication, as evidenced in our publication of three more copies of Urizen, the Job watercolors, and the many illuminated and non-illuminated works we have planned for production this spring, including Urizen copies B and D, the Songs of Innocence and of Experience copies A, B, T, V, and Y. We also plan to publish the Book of Job pencil sketches, engraved illustrations to Edward Young's Night Thoughts, the watercolor illustrations to two series of Milton's Paradise Lost and Comus, miscellaneous watercolor drawings and color prints, and the manuscript of Blake's Island in the Moon.

As always, there are areas of technological development. Our work-in-progress site is a veritable laboratory of ongoing experiments--no sooner is one completed than another is added. Additionally, we have just loaded all the data from our image production records into a new database, which will shortly be the means by which we track our workflow as the Archive continues to expand in new directions. We have also begun moving away from our DynaWeb interface and taking preliminary steps towards conversion of the markup of our texts from SGML to XML and/or other emerging text-encoding technologies that promise to provide a base on which to build a successor to DynaWeb.

In these and other areas, we have been making 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 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.

As we revise the infrastructure to accommodate a full range of materials, it becomes increasingly apparent that the expansion of the Archive will alter the environment for users. Though we have no intention of changing our base objective--to provide materials and tools for scholarship--we see that a broader Blake Archive will be able to address scholars and students together. Additional materials and improved tools provide what students, like scholars, need: more starting points for research and more useful connections among them.

After focusing on redesign and acquisition in recent years, we felt that we had reached a point where we could redirect our energy to publication. Even as we move ahead on other fronts--publishing full lists of the Blake collections of our contributors, additional illuminated books and non-illuminated materials, restyling the interface, etc.--we are aiming to make significant progress toward our ultimate goal of publishing the full range of Blake's work. In addition, we are preparing a study guide for the works of Blake as well as an article explaining illuminated printing. These are helpful, we believe, to nonspecialist users of the Archive, including the increasing number of students and teachers who use it in all levels of education, from grade school to graduate courses.

Top | Significance | History | Present and Future | Dissemination

5. Dissemination

The full resources of the Archive are already 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.

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, by the end of 2003, the Archive will have expanded very nearly to its full intended shape. The intricate SGML architecture and stylesheets required to integrate and display the full range of Blake's works in all media will have been developed, and each category will include a meaningful selection of significant works. By the end of 2003 we hope the Blake Archive will have taken its place among the preeminent models of humanistic scholarship in the medium.

From the start we have encouraged feedback from users via an e-mail link (we routinely respond to inquiries and comments from users). Recently we began collecting statistics on the use of the Archive, using software that counts the number and duration of hits for the site as a whole and for individual works. We shall design a questionnaire to survey our users' purposes and their opinions about the site, which we shall evaluate with the help of our advisory board.

As the Archive has evolved from a demonstration site showing three short and unsearchable illuminated books to a fully functioning archive that incorporates (at this writing) complete editions of 50 copies of all of Blake's 19 illuminated books, its reputation has rapidly spread far beyond the community of Blake studies. The Archive is now linked to all the major search engines, scholarly reference sites (such as Voice of the Shuttle), compendia (such as Romantic Circles), and journals (such as Romanticism on the Net). By filling out a form at the site itself, users can now subscribe to Blake Archive Updates and automatically receive e-mail notices with the latest information; these notices are also regularly posted to several online discussion groups. As word of the Archive continues to spread, the number of articles, reviews, and papers-and of requests for our texts and images-- continues to increase in both the mainstream and scholarly press (see Articles about the Archive).

For the time being, the chief obstacle to ideal access is the familiar one: speed. We are designing the Archive with an eye to the technical improvements of the future. For now, however, we supply several partial remedies for the bandwidth restrictions of the present: (1) We give users options: higher-resolution images that are slower to load than our inline images are made available by links; users with less adequate equipment or outdated browsers may bypass our Java applets entirely at the click of a mouse. (2) We continue to update and refine our software for optimal efficiency with current browsers (Inote is frequently revised to keep up with changes in Java).

We are especially interested in establishing multiple or "mirror" sites to improve access time for users with less than optimal connections to the Internet. Since December 2001, the public Archive has been mirrored on a Sun server at Oxford University and maintained by Oxford University Computing Services.

Eventually we may publish selected resources from the Archive in a second form. Those who need speed and portability--teachers in their classrooms, for example--will eventually be able to buy, at a price that individuals can pay, major works on portable disks. These amount to an extension, for the sake of convenience, of parts of the Archive. Although we have discussed disk publication with interested academic and commercial publishers, we must determine what publication arrangements are best for the future of the project as a whole.

Top | Significance | History | Present and Future | Dissemination