Delineation Editing of Co-Texts:
William Blake’s Illustrations

1. In his own lifetime and for the generation after his death, William Blake was best known as an illustrator of other authors. The illustrations represent the majority of his artistic output, and for his immediate posterity, his most famous works were not the illuminated books but the commercial book illustrations of Edward Young’s Night Thoughts (1797) and Robert Blair’s The Grave (1807) (Gilchrist 220).[1] Blake illustrated the works of several other authors, including Mary Wollstonecraft’s Original Stories from Real Life (1791, 1796), Erasmus Darwin’s Botanic Garden (1791), James Thomson’s Edward and Elinor (1793), G. A. Bürger’s Leonora (1796), Thomas Gray’s Poetic Works (1797-98), Chaucer’s Canterbury Tales (1810), John Bunyan’s Pilgrim’s Progress (1794, 1824), James Hervey’s Meditations and Contemplations (c. 1820), Virgil’s Pastorals (1821), Dante’s Divine Comedy (1824-27), Spenser’s Fairie Queene (c. 1825), numerous scenes from the Bible, and multiple works by William Hayley, Milton, and Shakespeare. This list does not include the many single plate illustrations he designed, and if, as Blake claimed, he helped create some of the designs he engraved, then the number of authors Blake illustrated grows tremendously.

2. As much as any of the illuminated books, Blake’s illustrations of other authors are profound examples of composite art. With them, an editor confronts a dizzying array of media and diverse relationships between images and texts. Materially, the illustrations explore the range between commercial print culture and the world of fine arts. Blake’s commercial illustrations, for example, were usually based on his own watercolor or pencil designs, some of which were displayed to the public in anticipation of publication. Some of the illustrations commissioned by patrons were printed from engraved plates and exist in multiple copies that have been colored and touched up differently, much like the illuminated books. Other commissioned illustrations were watercolor designs used by their owners as extra book illustrations.[2] Blake’s formal paintings have a complicated relationship with his own sketches, his prose or poetic descriptions, and the source texts of the designs.[3] In editions of Blake’s works by Keynes, Bentley, and Erdman, the Descriptive Catalogue, A Vision of the Last Judgment (1810), and the Public Address (1810-11) have enjoyed a largely independent existence apart from the paintings since the lack of a visual corollary effectively transforms them into more of a commentary on Blake’s poetry and mythology than the paintings. These editions largely elide the fact that these works exist in different material forms and were composed for different audiences. The Descriptive Catalogue, which was written for and sold at the 1809-10 exhibition, was the longest work Blake ever committed to print. In contrast, A Vision of the Last Judgment (1810) and the Public Address (1810-11) were notebook works never available to the original audience of the paintings, and even the ordering of the text is the result of editorial labor. Since Blake’s Last Judgment painting has been untraced since his death, no one was ever able to pair the description and aesthetic statement with the painting, and it was not until the early 1980s that the existence of the Chalcographic Society (the addressee of the Public Address) was verified.[4]

3. Editors of Blake’s illustrations of other authors must confront the function of the source text—its influence on the design, Blake’s relationship to the author and other illustrators, and even the material position of the text in relationship to the design. Much of Blake’s discussion of his Chaucer painting in the Descriptive Catalogue is a close reading of Chaucer’s text with quotations from a “bad” edition.[5] In the Epitome of Hervey’s Meditations Among the Tombs, Blake collapses the distance between a catalogue description and the painting space by identifying its figures in the painting itself. The Thomas Butts designs for Milton’s “L’Allegro” and “Il Penseroso” include a transcription of the relevant passages, descriptions, and commentary in Blake’s hand on accompanying manuscript leaves. The illustrations to Night Thoughts and the watercolor designs for Gray are literally built around the letterpress pages of other books. In the case of Night Thoughts, this book included Young’s own emendations, while in the illustrations to Gray, Blake wrote in the blank pages of the mounted edition, including an introductory poem for their intended owner, Nancy Flaxman, and descriptions or titles of the upcoming designs. Blake’s handwritten words often run out of the text box onto the space of the design, ruining the designs from one perspective but, from another, more intimately fusing word and image. As seen in The Illustrations to the Book of Job and the illuminated manuscript of Genesis, Blake continued to use texts by other authors to explore the interaction of word and image to the end of his life. But his union of word and text in these works stands as a significant departure from the segregation of text and image in commercial illustrations, where texts such as the design’s title, quotations, and the publisher’s colophon were written by a writing engraver beneath the design. While the vast majority of Blake’s commercial designs followed this format, Blake himself wrote the subscriptions for the Butts watercolor illustrations of the Bible in large, fine letters. The subscriptions are visible when the watercolors are displayed as paintings (as in the 2001 Tate exhibit), but reproductions in catalogues (print or electronic) omit the subscription in order to present the design as large as possible in the given space of a page, unfortunately eliding the relationship of the word and image.

4. Most discussions of editing Blake have ignored the innovative range presented by his illustrations of other authors to concentrate myopically on the illuminated books. In this focus, the editors of the William Blake Archive are typical: “We saw the illuminated books, once we had substantially achieved our first-phase goal of including one copy from every printing of every book, as a kind of archival and editorial backbone for the project” (“Editorial Principles”). As Justin Van Kleeck, Ali McGhee, and Rachel Lee show in their essays for this volume, the Blake Archive is now revising its editing procedures for two of Blake’s heavily revised manuscript works in the recognition that different material forms demand different sets of editing procedures. But while the Blake Archive is adapting to these forms, it has not considered how the diverse forms of Blake’s illustrations of other authors equally demand adjustments to its editorial methodology. The failure to theorize the unique editorial demands of illustrations has meant that many of their essential material elements have been elided.

5. I will argue that social-text editing provides the most appropriate editorial model for Blake’s illustrations of other authors. Social-text editing originated in the work of D. F. McKenzie, Jerome McGann, and others.[6] As McGann has frequently insisted, it is a model of editing that allows for a fusion of critical and facsimile editing on a single platform. In the codex, critical editing typically sought to create “a single, authoritative, original state of the work” that is linked to the intentions of the author, even if this eclectic copy differs from any and all pre-existing versions (Buzzetti and McGann 55). Facsimile editing had more fidelity to one historical document, which it approximated through photography or diplomatic transcription. But facsimile editions typically isolate one edition or state of the text. The advent of digital media allowed social-text editing to be put into practice (McGann, “From Text to Work,” par. 27). Social-text editing seeks to combine the functions of facsimile and critical editing and turn editorial and readerly attention to the different versions of the texts and the key contextual material. The many historic marriages of these different versions and contexts are why McGann insists “no book is one thing, it is many things, fashioned and refashioned repeatedly under different circumstances” (“From Text to Work” par. 36). As Susan Schreibman writes, “social-text editing presumes that published words are collaborative acts between writers and any number of agents: editors, family members, friends—even critics” (“Editing Electronic Editions” 24). Thus, social-text editing provides the “full range of social realities which the medium of print had to serve” (McKenzie 15).

6. The best known example of a social-text edition is McGann’s Dante Gabrielle Rossetti Archive. It brings facsimiles, transcriptions, and critical notes of all the contemporary editions of Rossetti’s writings together with all states and reproductions of his paintings, drawings, and prints. It also includes the writings and visual arts of his circle and important influences and documents regarding his reception history. As the Blake Archive does with the multiple copies of the illuminated books, the Rossetti Archive is able to present complex, multi-media projects like “The Blessed Damozel” in a new totality that precludes reifying one version of the work as the whole (“From Text to Work” pars. 27, 32). But the Rossetti Archive also reveals the social existence of the work, while the Blake Archive has largely avoided social-text editing for its innovative combination of facsimile and critical editing that is typified by its presentation of multiple copies of illuminated books.[7] This is not to downplay the importance of seeing both Blake’s commercial illustrations and his illuminated books together on the same editorial platform. As the editors of the Archive write, bringing together these different works yields “an augmented ‘Blake’ considerably larger than the one most familiar to students and scholars, especially those who approach Blake from the literary side” (Eaves et al. “Plan of the Archive”). My criticisms do not mean to disregard the immense accomplishment of the Archive, or to disparage the important work of the editors in focusing vital attention on Blake’s use of his media, which itself was truly a “radical editorial revision” (“Plan”). But I do want to suggest that the hesitancy of the Blake Archive to embrace social-text editing has its roots in traditional readings of and editorial approaches to Blake’s illustrations of other authors. These longstanding assumptions have contributed to the many contradictions found in how the Blake Archive has so far approached Blake’s illustrations of other authors.

7. I propose a specifically Blakean notion of social-text editing that I have termed delineation editing. Delineation editing uses Blake’s theory of the outline to flesh out key elements of social-text editing in Blakean terms. What I am most interested in is how Blake’s line draws an imaginative frame that accords conceptually with the edited textual body. As Blake’s theory of the outline suggests, that which exists outside of the outline—the unedited chaos of context so important to social-text editing—insists upon its own existence and calls attention to what is not enclosed within the frame. Delineation editing seeks to capitalize on the lack of autonomy in Blake’s illustrations of other authors to expand the editorial frame beyond the illustrations themselves to encompass their material and social realities. As I will show, the recent work of Joseph Viscomi and Saree Makdisi on the virtual elements in Blake’s prints has already moved scholarship in this direction, and I will apply their insights to how we can edit Blake’s illustrations.

   The Strong Blake Theory

8. In articulating his notion of the fluid text, Bryant challenges McGann by declaring that social-text editing “precludes the writer’s prepublication creative process” (Fluid 52). As concerns Blake, I will show this is precisely the field that most editors have most overvalued. Editors of Blake can be forgiven for keeping the oft-cited words from Jerusalem at the forefront of their minds: “I must Create a System, or be enslav’d by another Mans / I will not Reason & Compare: my business is to Create” (10.20-21, E 153). This passage creates the image of a Blake strong in authorial intention, one not willing to be corrupted by the influence of booksellers, publishers, patrons, and readers. While this image is accurate on several levels, what I will call “the strong Blake theory” has erected many impediments in appreciating and analyzing the complex intertextual nature of the illustrations. Frequently the illustrations are read as a kind of visual species of his witty and incisively written annotations, which are often taken as the model for how Blake read all texts.[8] As E. P. Thompson wrote:

   [Blake] would look into a book with a directness which we might find to be naïve or unbearable, challenging each one of its arguments against his own experience and his own “system.” This is at once apparent from his surviving annotations—to Lavater, Swedenborg, Berkeley, Bacon, Bishop Watson or Thornton. (Beast xvi)

   Yet if we understand the illustrations as the same kind of direct commentary on the authors Blake read, we are not only eliding the ambiguity inherent in visual media but also misreading Blake’s assertion in Jerusalem that he “will not Reason & Compare” but “Create.”

9. How to interpret Blake’s illustrations is a difficult question, and it is in many ways far more difficult to interpret the illustrations than the text. No one believes that Blake thought of an illustration as a “pure picture” untranslatable into words (Elkins 55). As he wrote his would-be patron, Dr. Trusler, “I hope that none of my Designs will be destitute of Infinite Particulars which will present themselves to the Contemplator” (E 701). But how Blake meant his “Contemplator” to interpret “Infinite Particulars” is unclear. Can “Infinite Particulars” even be bound in articulation? The relationship of the illustrations to their source texts only adds additional ambiguity. By the 1970s, the interpretative solution settled on by Blake critics was to integrate Blake’s illustrations fully into the mythic system of the illuminated books, which, in this period, was believed to be fairly fixed. Writing on the Night Thoughts designs, Morton D. Paley, for example, acknowledged Young’s immense popularity and the literalism in the designs that was apparent to early nineteenth-century readers like Henry Crabb Robinson and Edward Bulwer-Lytton. But he insisted that Blake’s mythology provided the crucial context for interpreting the designs:

   It is true that at times Blake does merely turn a trope into a picture, but frequently he only appears to be doing this. What neither Robinson nor Bulwer-Lytton nor the others realize is that for Blake the pictorialized trope is often a means of making a symbolic statement which depends for its meaning not on Young’s text but on the myth developed in the Lambeth books and in Vala. In pictures such as these, Blake adapts, ignores, or even subverts Young’s meaning in order to develop his own. (137, emphasis added)

   Paley raises an important point about the intertextual nature of Blake’s designs that I will return to later. But the knowledge that Paley suggests is necessary for the interpretation of Blake’s published designs is a steep burden for the contemporary reader who would be forced to ignore the seeming literalism of the designs for knowledge of Blake’s early illuminated books and an unpublished manuscript poem that was, at best, just begun in 1797. The strong Blake theory forces us to imagine Blake laughing in his sleeve at his naïve readers, and it grossly simplifies the complex relationships Blake had with the authors he illustrated.

10. One can argue that the reliance on the illuminated books to interpret the illustrations already exemplifies the idea of social texts, and to some degree this is correct. But the strong Blake theory errs by always imagining the same über-reader, which, following M. H. Abrams, ultimately only finds its true manifestation in Blake himself (Eaves, “Expressive” 784). While Blake as both creator and ideal audience is an important context, it also eclipses other interpretative scenes that have concrete historical existences. These scenes are worth exploring both for their own sakes and for their influence on how Blake may have conceived of the reception of his works.

11. By rejecting the strong Blake theory, I do not mean to suggest that Blake did not have disagreements with the authors he illustrated. But social-text editing would demand that Blake’s source texts be considered “co-texts” instead of merely one of many equal contexts that construct signification for the designs. Co-texts have been conceived in several ways that are useful for thinking about Blake’s illustrations. In linguistic terms, a co-text is a previous utterance that constrains interpretation (Brown and Yule 49), and in literary studies, the idea of co-texts has been used to analyze how texts function together on the page or in the same volume (Mühlethaler 39). Blake’s source texts function in their own socio-historical network, which existed before Blake’s illustrations and which the illustrations themselves join. The importance of the editorial apparatus to co-texts is suggested by research into electronic chatting and the difficulty in maintaining conversation strands across two or more conversation threads. Mariano Gomes Pimentel and his co-authors have designated this difficulty “co-text loss,” which they define as occurring “each time the reader is unable to identify which of the previous messages provide the elements that are necessary to understand the message that is being read” (484). Co-text loss in this sense is an apt description, for example, of the confusion faced by readers of Blake’s prophetic books where one is uncertain which preceding utterance can provide meaning. But co-text loss also has material and sociological implications for editing the illustrations because it suggests how much is lost by neglecting the importance of the source-text on Blake’s design. By downplaying the contributions of the source author for those of the illuminated book, editors are facilitating co-text loss.

    Electronic Editions and Blake’s Co-Texts

12. The chief editorial accomplishments involving Blake’s illustrations of other authors have been facsimile editions. The best of these (Essick and Paley’s edition of The Grave; Essick and LaBelle’s edition of the Night Thoughts engravings; Grant, Rose, Tolley, and Erdman’s Night Thoughts; and Hamlyn’s Night Thoughts) have been limited in the sense that they were forced to use the medium of print to represent print. As McGann has suggested, digital media has created an exterior field space that allows the bibliographic codes (or the mark-up language) of print to be revealed in new ways (“Rationale” 20-22 and “Marking” 205-06). However, the same editorial philosophy of the Archive that has done a wonderful job of presenting the variations of the illuminated books has produced a strangely abrogated and distorted view of the illustrations. The distortion of these illustrations originates in the diplomatic editorial philosophy of the Blake Archive, which emphasizes “the physical object—the plate, page, or canvas—over the logical textual unit—the poem or other work abstracted from its physical medium” (“Editorial Principles”). Missing from the list of physical objects is, of course, the codex, be it stitched pages or the bound print edition, which has been the chief “physical medium” of poetry, prose, and illustrations since at least the sixteenth century. While the editors of The Blake Archive have produced beautiful print facsimiles of Blake’s illustrated books (to say nothing of their own visually stunning scholarly works), the Archive does not present editions but simply the illustrations themselves removed from their co-text and context. Not recognizing “the bounding line” (E 550) of the print book means that the Archive excludes much, if not all, of the original works that spurred the illustrations, omits the physical context in which the design exists, and leaves out paratextual features such as prefaces, epigraphs, and even blank leaves that Blake may have considered as part of the field for constructing his meaning. It is important to recognize that this material is not merely one context but an essential co-text for the illustrations.

13. The Archive’s variation of the strong Blake theory leads to a host of contradictions about Blake’s works and their relationship to wider contexts and suggests that, for the illustrations at least, the best physical analogue to the Archive remains the traditional artistic catalogue, which gathers works from their places in specific contexts and collates them according to some rubric, such as author, period, medium, or owner. In terms of its hierarchical structure, the Archive proper (i.e., Blake’s Works) is organized first by its focus on Blake and secondly by its focus on the media he used. The contradictions in these rubrics emerge in the commercial engraving for The Grave. The Archive includes the portrait of Blake engraved by Louis Schiavonetti after a painting by Thomas Phillips. In its time, both the painting and the engraving were greatly admired, but Blake is only the subject. Given the inclusion of a work by two other artists in the Archive’s category of “Commercial Illustrations,” it is noteworthy that Blake’s poem “To the Queen,” which appeared in the print edition of The Grave, has yet to be included. What gives the portrait of Blake precedence over a poem by Blake is not clear, other than its immediate accessibility when the other prints were scanned. While the poem will certainly be added at some date, it will be divided from its co-texts and presumably will be classified with “Manuscripts and Typographic Works.”

14. Yet the decision to include the portrait of Blake is not a bad one because it follows the logic of the printed book in which the other designs appeared, and if, as Eaves suggests, electronic editing provides a model of textuality that is relatively “unstable” (“Crafting” par. 27), then Blake’s illustrated books provide a residual stability that editors must acknowledge. By not recognizing print editions in its hierarchy of objects, the Archive selectively bowdlerizes the material most associated with Blake from a volume and leaves the rest as dross. This may have been necessary for a host of reasons, but it bifurcates the co-textual nature of the illustrations, creating a false distinction between Blake’s own work and his source author as well as obscuring Blake’s place in the collaborative world of publishing.

15. While the critical resources on Blake’s life, mythology, and illuminated printing found on the Archive make up for this bifurcation to some degree, the sticky problem of co-texts is endemic. Once an object is selected to be in the Archive, all of the text printed on it is transcribed, including the publisher’s colophon, even if there is no indication that the text originates with Blake.[9] A major exception to this is the text of other authors, as seen in the elision of Young’s and Gray’s poetry, despite the fact that their poetry forms the literal center of the designs. In the case of the illustrations of The Grave, inconsistencies regarding what should be represented in the textual transcriptions abound. The Archive transcribes the subscription and colophon of each plate. These subscriptions include many of the titles that were first recorded in the November 1805 prospectus issued by Robert Cromek (Bentley, Records, 212). It is unclear whether these titles originated with Blake, Cromek, or were a collaborative effort. Who controlled the subsequent changes to the titles is equally unclear, but we know two facts for sure. Blake did not engrave them, and they are all in dialogue with Blair’s poem.[10] To take one example, the design often referred to as “Christ descending into the Grave” in the first prospectus was titled “Christ descending into the Grave, with the Keys of Death and Hell.” (I’ve italicized the lines from Blair.) In the published design, "Christ descending into the Grave” appears in large letters, and beneath this title an excerpt from Blair’s appears in smaller letters: “Eternal King! whose potent Arm sustains / The Keys of Hell and Death.” While Blake did not engrave the subscription himself, the design, title, and excerpt work together to provide an important gloss of Blair’s lines because of its identification of Christ with the “Eternal King,” a connection which, given the tension between revealed and natural religion in mid-eighteenth-century poetry, was not automatic and may have been contested by some readers. Without the accompanying lines from The Grave in the subscription to the design, the significance of the title would be lost.

16. Despite Blake’s questionable relationship to the subscription, from the perspective of delineation editing, the Archive properly transcribes it. The editorial omission of the facing page of text[11] (see page 1) by the Archive, however, radically curtails the most basic correlation between Blake’s design and Blair’s text in both theme and materiality. The facing page of text was the essential co-text. These excised lines describe the utter darkness of the grave:

       The Grave, dread thing!
    Men shiver when thou’rt nam’d: nature appall’d
    Shakes off her wonted firmness. Ah! how dark
    Thy long-extended realms, and rueful wastes,
    Where naught but silence reigns, and night, dark night,
    Dark as was chaos ere the infant sun
    Was roll’d together, or had tried his beams
    Athwart the gloom profound! (1)

    The design illustrates these lines as well since the light coming from Christ illuminates the darkness of the grave, aligning Christ with “the infant sun” whose “beams / Athwart the gloom profound.” As Christ descends, he approaches a line of fire running up the steps of the grave. Given that it is Christ and not the fire that dispels the darkness of the grave, Blake is likely making an allusion to the “dismal Situation waste and wild” that is lit by “darkness visible” in Paradise Lost (Milton I.60, 63). The comparison between Christ and the sun also suggests a host of allusions not only to Milton but to the typological tradition at work in Blair’s poem. Even if one argues that it was Cromek, and not Blake, who positioned the designs in relationship to the poem, every reader of the book (including Blake) experienced the design in this position.

17. My larger point here is that the meanings of Blake’s illustrations have a complex interdependent relationship with their verbal and visual co-texts that leads us to a reconsideration of the wider contexts in which they circulated. Removing Blake’s design from its place in the edition of The Grave facilitates co-text loss. In making this point, I echo with significant variation Robert N. Essick’s assertion that “The ever-present and generally unwelcomed demands of booksellers, partners, print dealers, and connoisseurs all influenced Blake in complex ways” (Printmaker 80). Certainly Blake’s disagreements with these figures must be acknowledged, but so must his real-world assumption that engagement with the public sphere in either print or painting depended upon them.

18. Most of my examples have come from the Night Thoughts and the Grave projects, but the same questions of co-text and context arise for Blake’s other illustrations. Blake’s three illustrations for Bürger’s gothic tale Leonora (1796), engraved by Perry, have never warranted an independent edition, either in print or electronic media. The two text illustrations are of soldiers departing for war and of reunited lovers, but these are anticipated by the frontispiece, which illustrates the heroine’s ride with the ghost who has been impersonating her lover. By anticipating the ending of the story itself, Blake’s frontispiece casts an uncanny pallor over the seemingly blasé domestic scenes of departure and reunion. While eighteenth-century critics condemned the wild design as “ludicrous, instead of terrific” (Bentley, Records, 75), the frontispiece essentially markets the shock-ending of Bürger’s narrative, which is what no doubt elicited the four editions of the work that were published in 1796. The co-textual relationship of Blake’s illustrations to Bürger’s text, however, is complicated by the presence of eight lines altered from Young on the frontispiece, making the verses a third co-text to consider. Blake had begun the Night Thoughts designs by this date, so it is worth asking whether Blake himself selected and adapted Young’s lines or whether this was done by Blake’s publisher William Miller, who would also sell copies of The Grave. Notably, Miller’s edition was also competing with one by the Night Thoughts publisher Richard Edwards, which contained designs by Lady Diana Beauclerc. Was Blake challenged to better these designs after seeing them in Edwards’s shop, or did Miller want to bank on what he may have seen as the fame that Blake would win with his Night Thoughts designs? And whatever answers are made to these questions, what role did Young’s own popular association with the gothic tradition play in the use of his lines?

19. Exploring these possibilities editorially would necessitate positioning Blake, Leonora, and Young in a wider network of literary and commercial relations. This approach to editing is an answer to the strong Blake theory and catalogue logic that removes Blake’s works from their various public contexts. Physical editions would be a necessary part of these networks, and rather than being gutted, Blake’s illustrated books could be represented in their entirety by a multitude of electronic formats—each with their own benefit. Future editors of Blake should begin thinking of how to convert the biographical and bibliographic heritage of Blake scholarship, especially as exemplified by Bentley’s Blake Records and Blake Books and the catalogues of Bindman, Butlin and Essick, into digital forms, as well as exploiting the exploding host of electronic scholarly editions, museum and library collections, and projects like Gale’s Eighteenth-Century Collections Online, 18thConnect, Google Books, and the Internet Archive.

    Delineation Editing

20. If a Blakean warrant for delineation editing is necessary, it can be found in A Vision of the Last Judgment, where Blake gave detailed instructions on how to view the now lost painting:

    If the Spectator could Enter into these Images in his Imagination approaching them on the Fiery Chariot of his Contemplative Thought if he could Enter into Noahs Rainbow or into his bosom or could make a Friend & Companion of one of these Images of wonder which always intreats him to leave mortal things as he must know then would he arise from the Grave then would he meet the Lord in the Air & then he would be happy (E 560)

    The viewer must become “a Friend & Companion” with Blake’s images, and this friendship becomes the basis for an enthusiastic melding with the divine. Blake describes this process as a perpetual resurrection, which suggests, recalling the letter to Trusler, how Blake envisioned engagement with “Infinite Particulars.” The alternative model this passage suggests is the cursory viewing, where the viewer treats Blake’s images as isolated, dead objects. If Blake’s ideal viewers are resurrected through their active relationship to the image, then those who view the painting incorrectly are themselves corpses, unable to arise from their graves. McKenzie, it is worth noting, describes editing in these same terms: “[B]ibliography as a sociology of texts has an unrivalled power to resurrect authors in their own time, and their readers at any time” (28-29).

21. The fact that the Last Judgment painting that Vision describes is lost creates an ideal situation for delineation editing. Spectators must not rely on their senses, part of “mortal things,” but weave together Blake’s description with their own biblical knowledge to create the lost co-text imaginatively. This method of viewing the lost Last Judgment echoes Blake’s description of how he viewed “those wonderful originals” of ancient monuments “in vision” when creating his painting The Spiritual Form of Pitt (E 531). Blake physically recreates these original monuments from what he understood to be imitations by the Persians, Hindus, Egyptians, and Greeks of lost Jewish masterpieces. Likewise, the lost “Imaginative Image” of the Last Judgment painting is recovered by “the seed of Contemplative Thought” as aided materially by Blake’s own works and those analogous to it (E 555). In this sense, a lost object has more reality than one that is visible and viewed wrongly.

22. By describing contemplation as a chariot and a seed, Blake suggests that contemplation both conveys the mind to a physically distant image and nurtures a present but undeveloped image. It is in this way that contemplation corresponds to the line that creates but also uncovers or recovers preexisting forms. As Blake writes in the Descriptive Catalogue, “Leave out this l[i]ne and you leave out life itself; all is chaos again, and the line of the almighty must be drawn out upon it before man or beast can exist” (E 550). Delineation editing of co-texts would frame a proper “picture” of elements that allows for contemplative exploration and self-conscious identification of the imaginative frame that the editorial “bounding line” and the “Spectator” create around the illustration and its source text(s). James Elkins’s discussion of Wittgenstein’s idea of a picture illuminates how the infinite particulars in Blake’s designs may be bound by a conceptual outline. Elkin cites Wittgenstein’s statement 2.14 that “‘the elements of a picture are related to one another in a determinate way,’ otherwise the object is not a picture” (62). Wittgenstein’s commentator, Max Black, glosses this statement using terms that are very similar to Blake’s condemnation of “blundering blurs” that destroy art and vision (E 572): “A smudged or blurred picture is not a picture at all” (qtd. in Elkins 62). Rather, “the blobs of paint of which a picture is made must be organized in a single, definite way, out of the many that are possible . . . in order to constitute a determinate picture” (qtd. in Elkins 62). As Robert N. Essick has pointed out, Blake did not always follow through on his theory of the outline in artistic practice (“Production of Meaning” 20). The outline is less artistic method, then, than it is a cognitive and interpretative model.

23. Removing the co-texts from the illustrations is in essence removing part of the infinite particulars constitutive of the design. Admittedly, the distinction between co-text and context can be a slippery one, but the immediate relevance of the source text to Blake’s designs should be apparent. Delineation editing of co-texts insists that the requisite material and social realities of Blake’s picture be included in the interpretative horizon.

24. Some of the problems and potentials of delineation editing are suggested by the recent work of Joseph Viscomi and Saree Makdisi, which has examined Blake’s designs in terms of electronic and imaginative forms of virtual reality. The analogy they draw between the electronic virtual image and imaginative image (the object of Blake’s contemplation) can be seen as a significant departure from the Archive’s emphasis on the starkly individualized physical object. Viscomi has shown that many of the illuminated books of 1795 have designs that fit together because they originated on one plate as a kind of painting before Blake decided to make them into books (“1795”). Viscomi uses digital media not only to reunite the divided physical images but also to reconstruct how the designs existed on the same plate. His digital images help readers to visualize the lost, imagined object of the print and the initial publication format Blake envisioned:

    Through such digital creations, we reify the experiences of memory and imagination, comparison and contrast, that we employ when reading / seeing Blake’s works. Moreover, by doing so, we engage in creative processes involving memory and imagination similar to Blake’s own when inventing new designs from elements of others. [. . .] In short, our [electronically generated] virtual reality is ideally suited for realizing Blake’s virtual designs.

    By recovering lost originals, Viscomi’s electronic virtual reality replicates the cognitive framing demanded by Blake’s outline.

25. How porous the outlines of this interpretative space can be is a topic explored by Makdisi. Makdisi stresses the “open logic of the illuminated books” (112), and as he contends, “if we try to read one of the illuminated books as a self-contained object, we will almost inevitably be frustrated. We will have greater success if we try to read it as a part of a virtual network of relations that opens away from itself and undermines its own autonomy” (130). Makdisi calls attention to Blake’s “graphemes,” such as the design for “Death’s Door,” which are replicated again and again throughout Blake’s works. The replication of graphemes (and even, one may add, of the exact verbal phrases that Blake would repeat in his poetry and prose) offers the tantalizing promise of a visual and verbal grammar. But since different contexts change the significance of graphemes, their replication also disrupts and frustrates the efforts to decode them. Given these semiotic contradictions, Makdisi suggests:

    [T]he stable self-containment of a single illuminated book is superseded by the wide virtual network of traces among different plates, different copies, different illuminated books; virtual because it is not always necessarily activated, and, even when it is, not always activated in the same way. (114)

    Makdisi avoids the pitfalls of the strong Blake theory because he recognizes that the connections among the illuminated books are virtual in that they exist in potential and are not always realized. He is not assuming Blake as his ideal reader. Instead, the meaning an individual reader creates by following “the wide virtual network of traces” is only one possibility or one imaginative frame. At the same time, this frame is also determined by the individual context of the grapheme (or whatever one takes as the individual unit in the network).

26. Blake’s illustrations of other authors exemplify Makdisi’s argument even more than the illuminated books. Paley recognized in his early essay on the Night Thoughts designs that graphemes do not respect the differences between the illuminated books and Blake’s illustrations of other authors, but the strong Blake theory limited Paley’s conception of the “virtual network of relations” to the illuminated books. These books certainly provide one set of relations, but there is a multitude of others, particularly if we are concerned with how Blake’s original readers would have interpreted the texts. Blake’s illustrations undermine their autonomy in ways far more complex than their referential relationship with the illuminated books because of their status as co-texts. As Viscomi and Makdisi suggest, editors and readers must be attuned to when, how, and whether a virtual network is activated. Editors and scholars are in a unique position to frame the elements for the reader. This crucial role has been underscored by Kathryn Sutherland (“Being Critical” 24), and it is important to keep in mind when performing social-text editing. Social-text editors must be acutely aware of historical readers and scenes of reading to be resurrected by the co-textual and contextual information they provide, and they must recognize that these readers lack the composite view they enjoy. Most of all, social-text editors should be wary lest they recreate a Platonic ideal of a text from the aggregate of their historicized particulars.

    Praxis

27. Delineation editing of co-texts outlines the fields in which Blake’s designs and their co-texts can be read and interpreted. The Night Thoughts designs, for example, embody a wide range of material forms and raise a host of questions regarding Blake’s relationship to his source author, his publisher, competing projects, visual and literary influences, and audiences. As detailed by G. E. Bentley, the Night Thoughts project was Blake’s chance to compete with the great book illustration projects of the 1780s and ’90s, such as Josiah Boydell’s Shakespeare, Thomas Macklin’s Poets Gallery and Bible, Robert Bowyers’s History of England, and Henry Fuseli’s Milton Gallery (“Great” 59). Like these projects, Blake’s Night Thoughts went well beyond being one book. Blake drew 537 watercolor designs—one for every page of the poem (including blank ones), some of which were displayed at various shops before being bound in two volumes. The print editions have several significant variations. Young’s text often runs onto the design due to printing errors that resulted from the experimental nature of the project. Some copies have the third state of the title page of Night the Second, while others have the fourth state. More than twenty of the engraved editions have been painted in watercolor, and some of these were done by artists other than Blake. Blake also found other uses for the proofs, employing one as a backing sheet for Hayley’s Ballads (Bentley, “Date,” 488) before using it and others in The Four Zoas manuscript. If McGann can describe the objects and contexts that constitute “The Blessed Damozel” as interstellar bodies that pull on one another (“From Text” par. 32), then Blake’s Night Thoughts is truly “when the stars threw down their spears” (E 25). In its efforts to capture all these aspects of the project, the 1980 Clarendon “Complete” edition contains black and white reproductions of all the watercolors, a facsimile of the engraved volume, all known proofs, selected colored reproductions of the watercolors, sketches and earlier designs by Blake, and a catalogue of the painted and engraved editions. The recent edition of the poem by Robin Hamlyn sought to represent the actual size and color of the watercolors, but reviewers have noted how the marginal designs were not reproduced and how the colors of the designs were still not fully captured (Snart, “Young”). Despite its enormous cost and craftsmanship, the Hamlyn edition could represent only a small portion of the Night Thoughts project, and like the Oxford edition, it does not even try to represent the painted print editions.

28. Far from being failures, of course, these two editions are forced delineations of certain elements within the wider Night Thoughts project, a delineation demanded by their status as print books. Digital media provide editors the opportunity to truly edit this project, but the editorial philosophy of the Blake Archive threatens to fracture the immensity of the project and to obscure junctures where different material aspects of the project might have intersected. In the two copies of the print edition of Night Thoughts currently available (one colored and one uncolored), the Archive provides the illustrations, and the textual notes include the descriptions found in an “Explanation of the Engravings,” which was included in some of the print editions, though there is no evidence the descriptions were by Blake or whether these individual copies contained them. Missing from the Archive are the title page, Richard Edwards’s advertisement, the pages of Young’s poems that have no engraved designs, and blank pages. While the relevancy of Young’s poem as a co-text should be apparent, as noted earlier, none of the text is transcribed even if the passage is marked by an asterisk.

29. I would suggest, however, that even the blank leaves have editorial value when considered in terms of the production history of the project. Not only did Blake illustrate the blank leaves in the Night Thoughts watercolors, but when the Night Thoughts watercolor designs were displayed to the public, the exhibit was a large part of how the original audience thought about the book. To this audience, the unillustrated pages in the print edition were not simply empty space, but they alluded to a more complete project that the original audience might have seen. Blake’s Night Thoughts watercolors were displayed at the shops of Richard Edwards, James Edwards, and Robert Bowyer, who, at the same time, was displaying prints and paintings from his illustrated History of England (Bentley, “Publishers,” 81). The relationship of these two grand illustration projects would have been obvious to all viewers, and together they would have served as individual works of fine arts, advertisements for their respective print editions, and as commentaries on the status of the British Arts. Even the readers who never saw the watercolors still would have read the spring 1797 prospectus that announced “forty very spirited engravings from original drawings by BLAKE” (Bentley, Records, 78). The engravings, for these viewers, replicated the more original watercolors, a fact that Blake was almost assuredly banking on for the future work of coloring the engraved editions.

30. The Library of Congress CD facsimile edition of colored copies J and B of Night Thoughts provides a good contrast to the methodology of the Archive.[12] It provides the open leaf images of both books, including all the paratextual information and the exterior and interior covers, a comparison of the differently colored designs, and a transcription of Young’s text. These features reveal much about the book as an artifact and the evolution of its production and province as an object. Whereas the omission of the non-Blakean pages by the Archive suggests their expendability in constructing Blake’s meaning, even the blank pages reveal foxing, plate impressions, and irregular leaf edges. While the CD edition does not explicitly exploit any features of social-text editing, its emphasis on the materiality of the books suggests several avenues of investigation. The anonymous inscription in copy J that it was “Coloured by Mrs. Blake” (Baker 6) points to Catherine Blake’s own activities as an artist and her role in creating the massive Night Thoughts project. Although it may seem tedious to transcribe each copy of Night Thoughts, by doing so, the Octavo team discovered two different spellings for “wrapped” on page nine. In copy J, it is spelled “Wrapt,” the same spelling found in the edition used in the watercolor designs, while in copy B and in the uncolored copy owned by Essick and reproduced by Essick and LaBelle, Grant et al., and the Archive, it is spelled “Wrapp’d.” This small difference underscores the erratic production of the edition, which may have necessitated the printer Robert Noble compositing page nine or at least the line twice. But because the page lacks a design, it is missing from the Archive. Aided by OCR scans, transcriptions of all copies could illuminate the production history of the volume by revealing unrecognized collations. Viewed in this way, the Night Thoughts project becomes less about, in Paley’s terms, deciphering “a symbolic statement which depends for its meaning not on Young’s text but on the myth developed in the Lambeth books and in Vala” and more about Blake’s collaboration with Young, Catherine, Richard Edwards, and a host of other figures.

31. If delineation editing of co-texts creates an editorial “form divine,” we can also think of the host of textual bodies outside of the delineated edited space as “visionary limbs,” a phrase I use to indicate the conspicuous absence of a particular design or image in a discreet unit that is readily available in another editorial platform. Visionary limbs are the necessary casualties of editorial delineation, but they also await their own resurrection since all belong to other potential editorial bodies. These visionary limbs are equivalent to Makdisi’s unactivated networks of meaning, and Viscomi’s work suggests that these limbs can be recreated through editorial work, be they actual historical texts and social networks or speculative theories of the editor. An important visionary limb of the Night Thoughts project would be its relationship to the other illustrated book projects. As Eaves has shown, Josiah Boydell’s Shakespeare gallery provides a window to a range of political, aesthetic, religious, and technological networks. A grand electronic edition of the great book illustrations of the late eighteenth century could incorporate the texts, paintings, prints, and exhibitions these projects generated. The edition would reveal how each project provided a context for the others. Such an edition would complete Alderman Boydell’s vision for a multi-media edition that collapsed partly under the pressure of bridging letterpress, engraving, and painting (Eaves, Counter-Arts, 35).

32. One of the most important features of many print facsimiles and scholarly commentaries on the illustrations, which has not been replicated in any electronic edition to date, is the inclusion of Blake’s vast number of visual influences. J. M. Q. Davies’s Blake’s Milton Designs, for example, is not an edition, but it presents the Milton designs alongside a rich genealogy of Blake’s graphemes and projects an entirely different model of Blake’s work than the Archive’s catalogue model. By putting Blake’s designs into a wider context, the designs lose their autonomy, but they gain a place in set iconographic traditions. This again has Blakean warrant since it forces us to look at his own graphemes as pre-existing visual types. Viscomi anticipates how a creative editor could exploit Blake’s graphemes as types when he replaces Death with Nebuchadnezzar in a virtual recreation of Death’s Door. This replacement highlights the typological and homomorphic connection between Death and Nebuchadnezzar that is found not only in Blake’s mythology and designs but also in wider Christian thought. Adam Komisaruk’s effort to map Blake’s mythology in three-dimensional space in his Blake Model is an important step in this direction, but such models also need to acknowledge and represent Blake’s co-texts and not simply his own universe.[13]

33. There are many tools that could help editors do this. Geographic Information Systems (GIS) offer a range of possibilities in visually organizing new editions of Blake and in analyzing the data in these editions. In her contribution to this volume, Mary Lynn Johnson describes her experience in updating the now colored maps of London, England, and the Holy Land, which are available in the second edition of Blake: Poetry and Designs. GIS models can be used to bring together information about Blake’s friends, patrons, publishers, and customers in ways that would reveal much about the milieu in which Blake worked and the lived reality he experienced. Sally-Beth MacLean and Alan Somerset’s work in mapping the provincial routes of Shakespeare’s companies using the Records of Early English Drama provide a useful model for mapping Blake’s London, as do several projects that utilize Historical GIS, such as William G. Thomas III and Edward L. Ayer’s Valley of the Shadow.[14] Susan Schreibman’s Thomas MacGreevy Archive, particularly her bibliography, provides a useful model for editing Blake’s letters and account books, and utilizing GIS and linking useful information about the persons and places mentioned could further illuminate Blake’s life and work. Following Komisaruk, GIS could model Blake’s spiritual London as well. Such ways of organizing knowledge about Blake need not flatten Blake’s “Visionary forms dramatic” (E 257) to the Newtonian space of Ulro. Acknowledging the universalizing, disembodied episteme found in typical uses of GIS, Mei-Po Kwan has argued that GIS has untapped potential for feminist inquiry of the type that accords well with the efforts of social-text editing to illustrate the complex agency and materiality of textual production (“Feminist Visualization” 652). Likewise, projects in Speculative Computing and Temporal Modeling have provided important challenges to the hierarchical organization demanded by markup language,[15] and if McGann is right, we may be moving to a future where nonhierarchical models of markup will become the norm (“Prologue” 21).

34. Social-text editing remains an underdeveloped means of exploring Blake’s works, their co-texts, and their contexts. Editors must create the borders of a space (physical or electronic) in which texts and / or images exist, and this is where careful delineation of the editorial field is key. Matt Kirschenbaum has described at length both the difficulties of redesigning the graphical user interface (GUI) of the Blake Archive and the potential future of the computer interface in general.[16] Social-text editing complicates interface design since it displaces the author and work from an isolated center. Social-text editing also demands that editors visibly represent more ephemeral concepts, such as how Blake’s works circulated in a range of contexts, his relationships to the authors he illustrated, the enthusiastic religious “underground” of the Romantic age, the visual and print cultures in which his works circulated, and the multitude of reading situations in which they participated.

35. By invoking Blake’s theory of the outline, delineation editing would require editors draw a “firm and determinate outline” (E 549), which would reveal the presuppositions of their editorial and textual theory and the remediation effected by the electronic media. Kathryn Sutherland has recently critiqued how many electronic editions with their wealth of versions and documentary witnesses overwhelm most readers:

    Can we really go forward into an age of digital editing with a model that suggests that each user is (or wants to be) her own editor? And if we do not (if, that is, we accept that electronic editions enact further controlling interpretations and theories about what text is), how will we equip the user to understand (and critique) those theories and interpretations? [. . . .] How will we make electronic editions worth desiring by more than a few developers? (19)

    A wealth of versions and social texts can quickly become chaos to those not versed in the textual and interpretative issues at stake. But to recognize this point is to arrive at the difference between “chaos” and “infinite particulars” for Blake. Delineated editions of Blake will be successful if they serve, as other editions have done, as the contemplative chariots that aid readers in imaginatively approximating Blake’s works. By exposing the social realities of these works, future editions would strive to help readers understand the wider material, public, and discursive worlds in which these works were created and functioned. In his illustrations to other authors, Blake shows how these worlds were essential portions of his creative acts.