review
begin page 31 | ↑ back to topJoseph Viscomi. Blake and the Idea of the Book. Princeton: Princeton University Press, 1993. xxvii + 453 pp. $49.50.
When Robert N. Essick’s William Blake, Printmaker came out 14 years ago, it seemed to me the last possible word on most of Blake’s workshop techniques, and it certainly was a vast improvement on all that had been said about them before. Joseph Viscomi’s new account of the processes used in the illuminated books, the first of two planned studies about Blake and printing, carries the discussion beyond the methods themselves into their consequences in the entire illuminated canon and their theoretical implications for understanding Blake’s art. It constitutes a Grand Unified Theory of illuminated printing and publishing, revealing important new patterns in the vast sea of data about Blake’s books that Keynes, Wolf, Erdman, Bentley and others have compiled so lovingly. A few points in this ambitious book will no doubt be further refined, and it will not make Bentley’s Blake Books or the forthcoming supplement to it obsolete, but for the foreseeable future Blake and the Idea of the Book will be just as indispensable for everyone who writes about Blake’s illuminated books and his other graphic works. Even those disputing Viscomi will do so in the terms of his arguments.
Viscomi writes very clearly and carefully, though his emphatic, enthusiastic style can be trying, even when one agrees with him—sometimes it’s like being shouted at . . . or prophesied to. But most readers will appreciate his exhilarating distrust of all received opinion and everyone will profit from his extraordinary ability to synthesize complex information in new ways. His most unusual asset in this undertaking is his extensive practical experience with processes virtually identical to Blake’s, some of which he acquired in co-publishing the Manchester Etching Workshop facsimile edition of the Songs: impressions from reproductions of some of Blake’s relief printing plates were hand-colored, producing pages that look even more like Blake’s than the collotype and stencil facsimiles by the Trianon Press. His participation in this venture afforded essential knowledge about the presswork involved in printing relief plates and the process of finishing them. Viscomi has also thoroughly researched the history of graphic techniques and conducted careful experiments with transfer methods, resists, etching and engraving begin page 32 | ↑ back to top techniques, and then thought deeply about the combined implications of this information. He included with the Manchester Etching Workshop facsimile a much shorter version of a few of the present arguments, but the searching thoroughness in Blake and the Idea of the Book distinguishes them here.
The first part of the book is organized according to the steps of the process of illuminated publishing, with excursions into variant methods and collateral questions as they arise. Several themes run through these discussions, but three seem especially important. First, Viscomi sees Blake’s methods in general as variations upon conventional techniques in etching and engraving rather than completely unprecedented inventions. Second, the chief distinguishing characteristic of illuminated printing is that it minimizes divisions between idea and execution, between text and image—divisions that are enforced by conventional technology and that continue to be reflected in critical thinking about Blake’s books. And third, although Blake’s publishing enterprise was always inefficient and its products more varied than those of conventional printers, much of the process was mechanical and determined by the need for efficiency in small-scale mass production; only later in Blake’s career were illuminated books deliberately differentiated as unique artifacts.
Viscomi begins by arguing that Blake usually created the text and design directly on the plate with only preliminary drafting of both, rather than completely settling the text and designs and laying out the pages in advance. Thus Blake usually did not employ the techniques for transferring designs that he must have used in much of his other graphic work. Viscomi shows how these transfer processes work, especially in the complex process of configuring the Night Thoughts drawings, and explains why such methods weren’t used in most illuminated pages. I found the digression as useful as the main argument, for the description of transfer methods cleared up a question that has troubled me for years. Blake occasionally made mysterious errors in his conventional etching/engravings which suggest that he somehow forgot the function of a given line within a drawing, as if myopia, a mechanical process, or a singularly obtuse assistant had intervened between a preliminary drawing and the preparation of the plate. For instance, in the stipple engraving of the Laocoön group for Ree’s Cyclopedia (reproduced in Essick, Commercial fig. 278), Blake mistook a pair of lines delineating one fold of the serpent and the inside of Laocoön’s right leg for an extra teardrop-shaped loop of serpent next to the scrotum, an error that he had not made in the related pencil drawing or in his later annotated engraving of the subject.
If Blake used squaring to reduce the images and counter-proofing to transfer them to the Cyclopedia plate, the lines that appeared on the copper should have been expressive enough to prevent gross errors. At one point I hypothesized that Blake might have used a pantograph or some similar device to reduce the image in his drawing to fit the layout of the plate, thereby creating lines so inexpressive that they could lead to such a mistake. (Even if a pantograph makes a line in exactly the right place, which it usually doesn’t, the line often lacks the critical variations in tone and thickness that allow us to understand what it represents.) But I found no evidence that Blake or other engravers used pantographs regularly, perhaps because these devices weren’t sophisticated enough to do such delicate copying. If his eyesight was worse than the eyeglasses preserved at the Fitzwilliam suggest, I thought next, Blake might have resorted to a loupe, which would have restricted his field of vision so severely that he couldn’t see what he was working on in context.
After reading Blake and the Idea of the Book I understand that Blake must have used squaring to make reductions, but that he then used some combination of calking (which is a little like using carbon paper), tracing, reversed tracing, and/or counterproofing to transfer designs to the plate. Both calking and tracing tend to produce somewhat inexpressive lines—calked lines are particularly likely to mislead, in that they are inexact in both position and strength. Blake probably lost track of the lines of Laocoön’s leg at the calking stage, just as a calked image of a mournful angel must have suggested an owl at an early phase in engraving page 23 of Night Thoughts (In Grant, et al., the drawing is NT44, the finished plate NT13E; the owl proof is reproduced as page 13E in the Proofs section of Volume 2).
One reason why Viscomi goes to so much trouble to show that Blake did not (or, in some cases could not) use transfer methods to create most relief plates for illuminated printing is in order to debunk once again the persistent fallacy that Blake wrote his texts forward and then mechanically reversed them onto the plate so that they would print forward. This is not the occasion for reiterating the debate on this point—it has been over for 14 years, despite Bo Ossian Lindberg’s attempt to revive it in his review of Essick’s Printmaker—but Viscomi goes beyond beating the dead horse to argue that the continued presence of the aromatic carcass of the transfer theory has kept us from understanding why as well as how Blake made illuminated books. Transfer theories, in Viscomi’s view, reflect the widespread eighteenth-century assumption that all printmaking processes are essentially reproductive—that is, mere mechanisms for replicating art that “really” exists in another medium—and they misrepresent illuminated printing by separating composition from production, illumination from text, visual art from verbal, inspiration from execution. Viscomi argues that relief etching offered an unpremeditated, autographic and organic alternative to existing reproductive and mechanistic media. It is essentially drawing and writing (backward) on copper—a medium in which the artist composes text, lays out the page, and illustrates it, working directly on the copper without a final draft of texts or designs on paper (obviously the Songs were mostly or completely written in advance, but Viscomi thinks they are exceptional rather than typical).
begin page 33 | ↑ back to topIn Printmaker, Essick suggested that Blake used a fine brush and a combination of oil and tallow to write his texts and draw his designs on the copper. Viscomi argues persuasively that this mixture would be difficult to use for writing text and unsatisfactory as a resist. He instead follows Linnell, who reported that Blake used a version of ordinary stopping solution, probably thickened asphaltum varnish, ground with linseed oil and diluted with turpentine. Viscomi also contends that Blake used a quill pen rather than a brush to write his texts, but I am skeptical on this point. I can’t make a quill write anything like a Blakean script, whether it is cut with a sharp spreading point or a square or oblique one, whether writing backward or forward; judging from the quill-written facsimile texts that Viscomi offers, he can’t quite do it either. Blake’s characteristic fluent cursive script and even the larger examples of his fussier upright Roman display an immensely complex pattern of thicks and thins that seems most consistent with Essick’s pointed brush; the texts written by Viscomi often duplicate the basic letterforms of Blake’s texts but the strokes appear to be uniform in thickness. I will not be surprised if it turns out that Blake used more than one tool to write with, just as he used several different scripts. A quill might well have produced the tiny letters (without noticeable thicks and thins) of some of the Songs of Innocence. But until someone demonstrates true Blakean scripts that have been etched in copper and printed, we won’t know for certain how Blake did them.
In addition to writing and drawing in resist, Blake worked on illuminated plates in several other ways, some of which involved transferring designs (not texts). Even in the early illuminated books Blake employed a variety of means for modifying the designs drawn in resist upon the plate. In most cases he simply treated the applied resist as if it were a conventional etching ground and scratched through it before the plate was etched, which usually printed as narrow white lines in the midst of the black areas created by the resist. Elsewhere he worked with a burin on the plate surface after etching the design; the effect is similar, but it may produce distinctive white lines running through the border left by the etching dam.
Blake also used two processes that Viscomi identifies with the confusing memoranda on the techniques of “woodcut on pewter” and “woodcut on copper” (Erdman 694). Both are means of producing white lines or white spaces on a full layer of conventional etching ground, and are called “woodcuts” because they are printed in relief as woodcuts are. “Woodcut on pewter,” as Viscomi sees it, is a technique in which the artist transfers a design by calking or counterproofing and then simply scrapes away both ground and underlying metal to produce a design composed of white areas surrounded by black (the ground here is not a resist but rather facilitates transferring the design and heightening the contrast between scraped and unscraped metal). Similarly, Viscomi argues that Blake’s “woodcut on copper” is the equivalent of modern white-line etching, and involves scraping or scratching away at a resist on copper and then etching the exposed metal to print whites (rather than blacks as in conventional intaglio etching).
These two processes are mechanically similar to Blake’s techniques for modifying designs drawn in resist, described above, but (and I think I’m getting this right) they are distinguishable because they employ an area of full ground, they necessarily involve a preliminary design (though it can be sketched on the plate itself rather than transferred), they require slightly different tools, and their white lines and spaces delineate primary forms. Although Viscomi’s case is fairly persuasive that these techniques are those called “woodcut on pewter” and “woodcut on copper” in Blake’s elliptical memoranda, I hope this terminology does not catch on outside of this book: here it helps to make important distinctions between the various white-line techniques, and wrenches some sense out of the memoranda, but the terms would be needlessly confusing in most other contexts, and even Viscomi sometimes uses them in ways that invite misunderstanding.
Once the copper printing plate was etched and otherwise satisfactory Blake could make major changes in the appearance of the image it generated by employing color-printing. For relief-printed illuminated books, this involved painting viscous colored inks on the raised and/or etched surfaces of the plate with a stump brush and printing all colors at once. Viscomi thus treats Blake’s color-printing as a relief-printing variant upon a common eighteenth-century mode of color-printing, the single-plate intaglio process called à la poupée, in which colored inks were applied (often rather approximately) to different areas of the intaglio plate.
In his discussion of the various printing processes used to create illuminated books from etched plates, Viscomi ingeniously reconstructs the presswork involved in printing specific books. He shows that in many cases Blake must have printed several copies of each book at a time. Most Blake scholars (including me) have underestimated the amount of work involved in setting up the press to print, and therefore assumed that Blake could have done so casually, whenever he felt like it, or when he got an order for a book. Viscomi argues persuasively, citing both practical experience and the evidence of the books themselves, that in most cases Blake must have printed the books in small editions of 10 or so, rather than one at a time. This a major discovery, but in his enthusiastic pursuit of the demon Error, Viscomi is a little too hard on others who have written about the subject. Several times, he cites someone as another dastardly perpetuator of the canard that each copy was printed separately, when the unfortunate citee merely implied this in arguing that each copy was unique in an editorially significant sense, that copies were finished differently, or that they were finished at different times.
Viscomi is so zealously attentive to his argument that he doesn’t always seem to recall something he clearly knows very well: in Blake’s case, printing a copy of an illuminated begin page 34 | ↑ back to top book is not the same as producing a copy of it, which for colored copies must include the entire process of assembling and finishing with pen, brush, ink and watercolors. Collective finishing as well as printing appears to have occurred, but unlike printing, finishing and assembling could take place at any time, and revisions, reassembling, and refinishing were always possible. Strictly speaking, editors must be prepared to weigh the authority of each instance in which Blake was apparently satisfied that his work on any feature, even a “blur or mark,” was complete or complete enough, and that may include hundreds of distinguishable instances in the production of a single page of a single copy of a single book. We will have to decide which differences are important and which are not. But even without dwelling on the infinite regression of “completeness,” editors of illuminated books can’t afford to put undue emphasis on “editions” when they are as small and as various as these. Even uncolored copies printed at the same time in the same ink on the same paper and bound in the same order will manifest differences that may be significant from an editorial point of view: the printed punctuation (at least) always varies, and textual editors and those discussing illuminations will have to continue to think about as many copies as possible even if many of them were printed at the same time. Overemphasizing the uniqueness of copies encourages certain kinds of critical foolishness, as Viscomi shows, but he also shows that critics have been wrong in other ways when they thought about too few copies.
Even if this prophet isn’t always gracious with Error, he seems to be on good terms with Truth: the evidence for small “editions” is abundant and various, now that he has pointed it out. Viscomi’s account of the ways in which the work of printing, finishing, and assembling books must have proceeded or could have proceeded is detailed and plausible. Everyone who writes about the illuminated books will have to reckon with this theory. Most of the time Blake, working with Catherine as his “printer’s devil,” must have printed multiple copies of each page (often in pairs); then the two of them colored the pages together en masse (often perfunctorily) and only later assembled them into books. Casually finished books might subsequently be touched up or even elaborately refinished at another time, as some late copies of Songs were, but most pages of most books were never the object of sustained individual attention and their collective effect in a particular book may or may not have been in the mind of the colorist (who may or may not have been Blake himself). The account of Blake’s publishing process here has profound consequences for all those who think about editing illuminated books; what was already one of the most difficult of editorial problems is made even more richly problematic. Although Viscomi’s instinct for polemic in these chapters is sometimes distracting, he has provided the groundwork and much of the framework for the next, more sophisticated phases of editorial theory about Blake.
In the second half of the book Viscomi discusses recoverable printing history for each of the illuminated books in detail; in every instance his account makes new sense out of old puzzles, and often the puzzles melt away entirely. Some of the information in these chapters has already been summarized in the new Blake Trust facsimiles that Viscomi helped to edit, but much of it is entirely new. He leaves few applecarts unturned, redating almost the entire illuminated canon in various ways, and he indicates the consequences of his theories for editors and even biographers. All of this is important and some of it is even exciting, but as one might expect, several of the arguments make heavy reading if you aren’t immediately engaged in the questions they address, as in the latter part of Chapter 32, “The Production and Evolution of Milton: 1804-1818,” which uncovers this complicated process in exhausting (but not exhaustive) detail. A very useful five-page Appendix charts the hypothetical publishing history of all 21 illuminated books from 1788 to the posthumous copies, identifying the approximate date of printing (not finishing) for each copy. Most readers will find themselves referring to this chart regularly whenever they are working on illuminated books.
Blake and the Idea of the Book is handsomely laid out and sturdily constructed in an oblong format, perhaps in deliberate imitation of Erdman’s Illuminated Blake. It is generously illustrated with 312 good-quality monochrome halftones that convey as much information as such illustrations can, which is to say barely enough, and includes 13 color plates that do only a little better. It would have been impossible to publish this book for less than $50 if more or better color had been used, and I doubt that Princeton will make much money even as things are; the problem is that many of Viscomi’s arguments necessarily depend on attending to phenomena that aren’t easily seen even in good offset reproductions. Perhaps when all the pages of all copies of the illuminated books are available in color-corrected, high-resolution digital images, as envisioned by the scholars working on the electronic Blake archive, Viscomi will assemble an elegant illustrated hypertext version of the publication chart in Appendix A that will make everything as clear to us as it is to him.
Works Cited
Bentley, G. E., Jr. Blake Books. Oxford: Clarendon P, 1977.
Essick, Robert N. William Blake, Printmaker. Princeton: Princeton UP, 1980.
—. William Blake’s Commercial Book Illustrations. Oxford: Clarendon P, 1991.
Erdman, David V., ed. The Complete Poetry and Prose of William Blake. Newly rev. ed. Garden City, NY: Doubleday, 1982.
Grant, John E., et al., eds. William Blake’s Designs for Edward Young’s Night Thoughts. 2 vols. Oxford: Clarendon P, 1980.
begin page 35 | ↑ back to topLindberg, Bo Ossian. Review of William Blake, Printmaker, by Robert N. Essick. Blake: An Illustrated Quarterly 15 (Winter 1981-82): 140-48.
Viscomi, Joseph. The Art of William Blake’s Illuminated Prints. Manchester: Manchester Etching Workshop, 1983.