William Blake, S. W. Hayter and Color Printing
By Martin Butlin

“Opposition is true friendship”[1]

I should be flattered that my article on Blake's color-printing techniques should have provoked an answer from Robert N. Essick and Joseph Viscomi no less than two and a half times as long; even more surprising is the speed at which they must have prepared their answer (Butlin “Private War,” and Essick and Viscomi “Method,” both in Blake/An Illustrated Quarterly 36.2 [fall 2002]). Unfortunately this has resulted in a number of inaccuracies and distortions that come strangely from those who claim to “Labour well the minute particulars” ( “Method” 49).

Above all it is the authors' failure to recognize fundamental differences between Blake's color-printing technique and that of S. W. Hayter that casts doubt on their ability to see beyond their own commitment to the one-pull cause. But before discussing this central issue, there are a number of minor points that need attention. An inevitable but annoying result of the speed with which they prepared their article is the lack of any page references to my own piece; this is compounded by their idiosyncratic approach to quotations and quotation marks. An example is the quotation attributed to me, presumably jokingly, on page 49: “looks like two pulls to me and my friends”—no doubt they can read Blake's mind as well! A real quotation, but one equally strange in its implications, occurs on page 56 where the authors put quotation marks around my description of Richard Lloyd's position at Christie's; if they really question this they need only look in Christie's sale catalogues. A third case is rather more serious. On page 57, note 10, one reads that “Butlin states that 'creating precise registration . . . was indeed anything but a mechanical process.'” After a worried period wondering what exactly I was supposed to have meant by this I went looking for it in my article; finally, right towards the end ( “Private War” 48), I discovered the words “It [the two-pull process] was indeed anything but a mechanical process creating precise registration; rather, it was the reverse of the precision that, it is claimed [by Essick and Viscomi], could have been produced by a one-pull process.” The authors have managed to reverse what I actually said, making nonsense of it in the process.

Our authors also demonstrate some strange uses of the English language. That of the word “replicas” (59) is misleading, particularly when the relevant illustrations (fig. 13, 15 and 16) are captioned with the far more accurate description, “based on Blake's design.” Worse, on their first page (49) they indulge in a bit of quibbling over the word “prominent” and then go on (50) to apologize for failing to mention my review of Viscomi's Blake and the Idea of the Book in which I expressed my reservations on the one-pull process, in “a paragraph of 206 words.” I would be more flattered if they had gone on to read my next paragraph; this also deals with Blake's color-printing techniques and totals approximately another 200 words, thus doubling the total to about 400 in a review of 800 words—this proportion seems fairly “prominent” to me, particularly as it was the only critical element in my review (Burlington Magazine 137 [1995]: 123).

This lack of attention to an opponent's case is also seen in their reaction to my pointing out the inaccuracy of their denial (repeated six times in all!—see “Private War” 46) that Phillips had found other “such failures of registration” in copy E of Songs beyond that in the “Nurses Song” (Phillips, Creation of the Songs 103). Far from apologizing, they exclaim “This is the sort of general comment in which the advocates of the two-pull theory specialize” (51). They then go on to ask rhetorically, “Which plates? Where in each plate? What is the evidence supporting this opinion? Where is the 'overwhelming evidence' . . . ” (51). Well, for those who care to read, Phillips listed five examples on the page I referred to, at the end of the long paragraph on page 103 that concludes just before the end of the right-hand column. Even if our authors do not agree with Phillips's conclusions, they really should not deny that he does give these examples. These and similar inaccuracies and distortions may be regarded as trivial, as failures of method rather than substance, but they cast doubt on the authors' scholarly approach as it applies to their own researches, whether into “minute particulars” or on more general matters.

To move on to concerns of more substance, much of the authors' new article goes over old ground, restating their arguments on, for instance, the “Nurses Song” in copy E of the Songs (50), the 1794 date on Songs of Experience copy T1 (52-55), and even pin holes (60n12)! They fail to appreciate my distinction between color printing to reinforce lines and color printing to produce areas of dense color and texture, and the implications this has for the need for “precise registration” in the first case and its relative lack of importance in the second (50, 51, etc.). I would also question their statement (51), apropos Dörrbecker's claim that the full-page illustrations in The Song of Los were color printed “from almost unetched plates, occasionally in multiple layers of paint . . . ” (Continental Prophecies 310), that this does not imply multiple pulls; surely the mixture of different colors while they were still soluble (!—see their page 61) would produce a murky effect totally unlike the brilliance achieved by Blake. The authors also reverse my views on the direction of Blake's development (61; “Private War” 48, where I do not even mention the “small relief etchings” —perhaps I should have done so, but I was only covering the years 1794-96), but it seems futile to go over these points yet again if they are only to be misread or ignored.

What is good about their article is that they have now ventured onto the difficult ground of intaglio printing and the large color prints of “1795.” They are still not prepared to address directly the two fundamental questions, firstly the differences in pressure needed for printing in intaglio and in the thick color medium used by Blake, and secondly the respective drying times of ink and Blake's color medium. Instead they refer to experiments by S. W. Hayter, stating rather coyly that “It may come as a surprise to Butlin, as it did to Essick, that Hayter believed that simultaneous printing in intaglio and surface colors . . . 'will seem so obvious [to the layman] that he will be surprised to hear that it was not carried out successfully long before'” (60). Essick may well have been surprised (cf. his William Blake, Printmaker 130), but Butlin was suspicious. It is indeed strange that Essick and Viscomi should be so happy to rely on Hayter's example when, as Morris Eaves has remarked, they claim to have demonstrated that the efforts by Hayter, together with Joan Miró and Ruthven Todd, to recreate Blake's color-printing processes were “dead wrong.” [2]

A look at Hayter's own description of how he achieved the combination of intaglio printing and color printing in the key work in this development, Cinq personages, 1946, which takes up a whole chapter in his New Ways of Gravure (1949, 155-61; revised edition 1981, 132-35), together with a look at the good color reproduction in the book from which they do quote by Peter Black and Désirée Moorhead (15-16, 23, 168-69) or even better at an original copy of the print, [3] will show how irrelevant the comparison is. Even in reproduction it can be seen that the areas of color in Hayter's print are translucent if not fully transparent (although interestingly they do show some of the reticulation found in Blake's work); this is a result of their having been printed in colored inks rather than in the thick medium used by Blake. Both the visual effect and the textural effect are therefore totally different from that achieved by Blake in his color-printed books.

Hayter's written description makes this clear: “On the print, color would appear as offset from the plate surface and, over it, would be the intaglio impression in black, appearing slightly in relief above the surface of the paper” (159/134); this can be compared with the results of Viscomi's own experiments in printing from an intaglio version of The Marriage of Heaven and Hell plate 10, when, although the illustration was printed at the same time as the text, “The ink did not print though the colors” ( “Method” 59, fig. 16). Hayter continues: “Registration with the black [in applying colors] was simplified by the fact that the plate was visible through the [silk] screen while the work was being done” (160/134). Earlier in his book Hayter states, in a section on “Simultaneous printing from surface and intaglio” (chapter 7, 132-33/115), that “Inks must be used and color deposited on the surface in such a way that it does not mix with the ink already on the plate. Then, as it is clear that an opaque oily ink between intaglio ink and damp paper would prevent its adhering to the print, the film of color on the surface must remain so thin that it is microscopically porous. Also, it is seldom possible to superimpose two [or more] colors . . . so practicable transparency will be confined to a bitten texture in intaglio over a simple surface color in each part. Of course, the surface color being above the intaglio in the plate will appear underneath it in the print” (132-33/115; cf. “Method” 54). This would appear to describe a process just about as far as it could be from that attributed to Blake by Essick and Viscomi.

Interestingly, when Hayter's textures did become thicker, as in Poisson rouge of 1957, he forsook the fine lines of intaglio printing and replaced them with a broad line created by a felt pen and spotting resin, producing a relatively thick textured outline rather than the thin line of Blake's intaglio work (Black and Moorhead 16, 222-23, no. 237, illustrated in color). Hayter's line in these later works is in fact closer to the brush-drawn outlines found in some of Blake's large color prints (as demonstrated by Piers Townshend in the course of a talk to the Blake Society in London on 19 November 2002 ). Incidentally, although Hayter says that it took him fourteen years to perfect his new one-pull technique, this is not strictly accurate: his first experiments, with Anthony Gross, did take place in 1932 (Black and Moorhead 15) but he did little further work on this aspect of his technique until the years 1940 to 1946, from Maternity, 1940 (Black and Moorhead 15, 22, 145, no.132, illustrated in color), to the small Centauresse of 1943-44 (Black and Moorehead 15, 22-23, 161, no. 157, illustrated in color) and, the subject of Hayter's main account, Cinq personages of 1946, a print for which Hayter needed no fewer than four assistants.

The authors raise the question of why The Book of Los and The Book of Ahania only exist in one copy each (57). There are plenty of reasons other than technical ones why this should be so. Of the various factors involved, the most important seems to have been a crisis in Blake's output of poetry. Both books, plus The Song of Los, of which the known copies were done in what, following Viscomi, one can call a single edition, are significantly shorter than their predecessors, The Song of Los covering the two continents Africa and Asia in four pages as compared with America's 16 plates and Europe's 14 plates, while the two other books contain only seven text pages between them, albeit rather more densely filled with writing. No other books were even nominally published until 1804. Instead, Blake began the long unfinished epic Vala or the Four Zoas, commencing this in a finished, copperplate hand suggesting that he was planning a single deluxe manuscript rather than another illuminated book. At the same time the technique of The Song of Los, still relief-etched, though perhaps the most magnificently and richly colored of all Blake's books, seems to have led to technical difficulties (see my “Evolution of Blake's Large Color Prints,” 115-16) that may have suggested the separate issuing of illustrations without text in the small and large Books of Design and the large color prints of “1795,” a development set out by me ad nauseam elsewhere. With the best will in the world, The Book of Ahania and The Book of Los cannot be said to further Blake's attempts to integrate text and illustration, and he may well have been worried by the disparity between the sharply defined lines of the text and the broad, relatively undefined forms of most of the illustrations; only the head-piece to chapter 1 in The Book of Los points to a possible solution. As Essick points out in William Blake, Printmaker (149), “In The Book of Los and The Book of Ahania, both with surface-printed designs but intaglio lettering, the technical gap between word and picture widens. At that point, the text-design synthesis made possible by relief etching broke down completely, and Blake's color printing rapidly developed independent of his work as a poet.” So why did Blake even experiment with intaglio printing, if not to achieve a possible easing of the admitted difficulties of the two-pull process?

“To Generalise is to be an Idiot.” [4] On page 58 I am accused of saying “if two pulls here, then two pulls in everything leading up to it [the intaglio books].” The authors go on to stress “how experimental Blake was as a printer and printmaker,” but rather spoil their argument by concluding that his “consistency resides in his continued use of one-pull printing,” “In this matter . . . Blake was as consistent as Butlin supposes.” Given the universally accepted double printing in the “Nurses Song” from copy E of Songs there is already a built-in inconsistency here, and if “the exception proves the rule” it also breaks it—clearly Blake did use two pulls in this particular work, as even Essick and Viscomi allow.

I am not convinced by Essick's change of mind which now attributes the spreading beyond the platemark of the printed color on The Book of Los title page to the presence of bevelled edges on the plate (58); the Book of Job plates, which do have bevelled edges, were done as the result of a relatively generous commission, whereas in the Lambeth Books bevelling would have been a luxury. Viscomi himself suggests that the plates for The Book of Los and The Book of Ahania were cut from a single sheet of copper which was only bevelled on the original outer edge ( Idea 287, 414n26, where he states that plate 3 of Ahania “was the recto side of the sheet, for its right side was bevelled, whereas this edge is on the left side in plate 4 and is sharp from being scored and broken” ; he does not mention the sections used for The Book of Los but would, presumably, have pointed out if there were any signs that Blake's treatment was different).

Essick and Viscomi state that Blake's color-printing medium could be revived by being reprinted onto damp paper under pressure (61). This again begs the question of how long Blake's medium took to dry, and the term “water-based colors” fails to distinguish between thin watercolor washes, which might well be so revived, and the thick glue- or gum-based medium used by Blake in the color-printed books. (In fact even thin watercolor becomes insoluble in time; Blake's Dante illustrations in the Tate, as well as a number of Turner watercolors, have been carefully washed in recent years, and the discolored paper of certain early Blake watercolors such as the Tate's Oberon, Titania and Puck with Fairies Dancing has also been bleached.) Essick himself wrote in 1980, “Repainting and printing from a millboard caked with old, dried pigments would be at best a terribly awkward business” (William Blake, Printmaker 132). But, as Essick also points out, the gum medium in his own copy of Lamech and Two Wives “is about as insoluble in water as animal glues, and like such glues must generally be heated and agitated before it can be used as a fixative for watercolors. Unlike gum arabic, it is most receptive to subsequent applications of watercolors or pen and ink work without disturbing lower layers” (131). He goes on to say that the paint used in Blake's color-printed works must have required “a high level of insolubility” (131). Presumably Essick has changed his mind, but first thoughts are, it would seem, the best.

Our authors' arguments get even more desperate when they suggest that, when some of the 1795 large color prints were reprinted in 1804-05, “The outline and composition, the latter in thin layers of dried colors on the support after printing, could be returned to years later and the process repeated” (61); it is hardly likely that Viscomi, despite all his experiments, has been able to leave his thick color medium for a long enough period to see if it would remain insoluble—certainly not for the nine or ten years necessitated for the later pulls of the large color prints.

Essick and Viscomi admit that at least one of the large color prints, God Judging Adam, was printed from a metal plate, and Rebecca Donnan has confirmed this by pointing out that the version in the Philadelphia Museum is embossed on the back from the etched plate. [5] Whether the other prints were done on metal or on millboard is irrelevant to our authors' suggestion that “The outline of the design was probably drawn . . . in India ink, which would adhere permanently to the support and not transfer to the paper” (61). Piers Townshend, however, who restored the pull of Satan Exulting over Eve in the Tate, has shown that such outlines, drawn fairly broadly with a brush, were in fact transferred to the final print (lecture of November 2002). By now, particularly if he was using millboard, Blake, like Hayter in his Poisson rouge, did not need to print the outline as a separate process, though God Judging Adam, the one print definitely from a metal plate, does seem to show relief-etched outlines. This seems to be the exception to the rule; even within the series of twelve prints Blake's technique was inconsistent.

Essick and Viscomi make heavy weather over maculatures, stating that they are incompatible with a two-pull process (63). Other things being equal (which admittedly Essick and Viscomi do not accept) there seems no reason why maculatures cannot appear on successive pulls made either on blank paper or on paper already printed with outlines. The edition could be printed either way, and if Tatham could do without damping his paper (52), why not Blake? (Incidentally, Ruthven Todd suggests that “it may be doubted whether he [Blake] used his paper in the soaked condition common among modern copper-plate printers” ; rather it would be removed “from its wrappings at least two weeks before starting on the job” and left “lying beside the press to absorb the natural dampness of the printing-shop” [ “Techniques” (1973 ed.) 30].)

One possible line of research that leads back from the large color prints to the later color-printed books is perhaps worth reviving. Ruthven Todd, in his William Blake the Artist (1971) suggests that some of the full-page illustrations in the 1795 books could have been printed not on a rolling press but merely by lowering a sheet of paper onto the plate and applying a light pressure: “Any press would not only flatten the colors, but would also spread them messily” (37). Such a technique could have been used at least for the full-page illustrations in The Song of Los and The Book of Ahania, and perhaps the clearly defined designs on text pages such as plate 6 of Ahania and plate 5 of The Book of Los. The same would apply to the color-printed versions of Albion Rose and The Avengers (Essick, William Blake, Printmaker pl. 66-67, 69, 70). Without the confines of the printing press, registration, not necessarily too precise, would have been easier, as would have been the task of printing from the relatively large area of the large color prints. This would explain the absence of platemarks on all but God Judging Adam. This possibility has been rejected by Viscomi for the illuminated books but not for the separate color prints (Idea 126), but could be a possible development in the process of technical simplification in the cases mentioned above.

There is clearly much work still to be done on Blake's color-printing techniques. Unfortunately the two camps, for one-pull printing and for double-pull printing, seem to have become irretrievably bogged down. Blake, as we have seen, was inconsistent even within such a series as that of the twelve large color prints. More flexibility, and considerably more time, is needed to write a full account of his color-printing techniques in the 1790s. Essick and Viscomi postulate such a book (49), and one is sincerely to be hoped for, but they should give themselves time, months or years rather than days or weeks, to produce the study of which they are surely capable, one worthy of the subject, of William Blake, and of themselves.


[Editors' note: Click here for the response from Robert Essick and Joseph Viscomi.]


Works Cited

Black, Peter, and Désirée Moorhead. The Prints of Stanley William Hayter. London: Phaidon, 1992.

Butlin, Martin.“The Evolution of Blake's Large Color Prints of 1795.” William Blake: Essays for S. Foster Damon. Ed. A. H. Rosenfeld. Providence: Brown UP, 1969. 109-16.

Butlin. “The Physicality of William Blake: The Large Color Prints of '1795.'” Huntington Library Quarterly 52 (1989): 1-17.

Butlin. Review of Viscomi, Blake and the Idea of the Book. Burlington Magazine 137 no. 1103 (Feb. 1995): 123.

Butlin. “'Is this a Private War or Can Anyone Join In?': A Plea for a Broader Look at Blake's Color-Printing Techniques.” Blake/An Illustrated Quarterly 36.2 (fall 2002): 45-49.

Dörrbecker, D. W. Commentary in William Blake, The Continental Prophecies. Blake's Illuminated Books vol. 4, General Editor David Bindman. London: William Blake Trust/Tate Gallery, 1995.

Erdman, David V., ed. The Complete Poetry and Prose of William Blake. Newly Revised Edition. Berkeley and Los Angeles, 1982. Cited as “E” followed by page number.

Essick, Robert N. William Blake, Printmaker. Princeton: Princeton UP, 1980.

Essick, and Joseph Viscomi. “An Inquiry into William Blake's Method of Color Printing.” Blake/An Illustrated Quarterly 35.3 (winter 2001-02): 74-103. Online version.

Essick and Viscomi. “Blake's Method of Color Printing: Some Responses and Further Observations.” Blake/An Illustrated Quarterly 36.2 (fall 2002): 49-64. Online version.

Hayter, Stanley William. New Ways of Gravure. New York: Pantheon Books, 1949; revised edition, New York: Dover, 1981.

Phillips, Michael. William Blake: The Creation of the Songs from Manuscript to Illuminated Printing. London: British Library, 2000.

Todd, Ruthven. “The Techniques of William Blake's Illuminated Printing.” The Art Collector's Quarterly 29 (Nov. 1948): 25-37. Reprinted in The Visionary Hand. Ed. Robert Essick. Los Angeles: Hennessey & Ingalls, 1973. 19-44.

Todd. William Blake the Artist. London: Studio Vista, 1971.

Viscomi, Joseph. Blake and the Idea of the Book. Princeton: Princeton UP, 1993.