America: A Prophecy. Intro. by G. E. Bentley, Jr. Ed. Comments & Check-List of Secondary Materials by Roger R. Easson. American Blake Foundation, 1974. x + 21 pp., 18 unnumbered pls.; + 9 numbered pls. in cloth ed. (1 in color). $10, paper (with Vol. II, Europe, forthcoming); $35, cloth (Vol. I, singly).
America a Prophecy. (Unbound facsimile pages.) Published by Blake Newsletter, 1975. 1 p., 41 pls. $2.50.
By a coincidence years in the making, two sets of black-and-white facsimile pages of America have been published within a few months of each other—and can conveniently be described by comparison.*↤ David Erdman has been putting several fascinating articles on Blake into BNYPL, hoping for more subscribers. Let me call them by their initials, ABF and BNL, for brevity—using TIB for the facsimile in The Illuminated Blake (pp. 137-155, 392-395), to which it may be helpful to make further comparison.
The BNL packet consists of 1 page explaining the project (it was suggested and assisted by Everett Frost), large reproductions of America i-ii, 1-16, a-d (all except i and d in two styles of reproduction, back to back), and no commentary. The paperbound ABF volume consists of full size reproductions of America i-ii, 1-16 (confusingly referred to as 1-18 despite the convention generally accepted since Keynes; but the facsimile pages are not numbered, so one may write in either set of numbers, or both), with prefatory matter and brief Introduction and Check-List. The clothbound ABF volume adds nine plates: canceled plates a-d of America (plate d, “Thiralatha,” in color); photographs of the extant copper fragment of plate a and of two impressions from it; familiar variant plates (pl. 2 from copy N, with “The stern Bard ceas’d . . . ” uncanceled; pl. 13 from copy B, with a single tail on the serpent); and two drawings, the “Chaining of Orc,” and an early sketch for the title page, about which more anon. Both the Foundation and the Newsletter publications use the uncolored Rosenwald copy, E, except that BNL substitutes for plates 9 and 13 the somewhat clearer proof pages of copy a*. There is of course only one set of the canceled plates, a-d, and all reproductions of them seem to derive from the same set of photographic negatives. For clarity of detail, copy E was a better choice than uncolored copy N, used for most of the plates in TIB—though the worst thing about the TIB plates is their reduction to two-thirds the original height.
The American Blake Foundation America, as the first sample in a series that promises to supply low-priced classroom materials, is reassuring in the quality of its plates. They are unlovely, but they are unreduced in size and are adequately printed by fine-screen offset lithography. Unhappily the typography of the book reveals no attempt to improve on the uninspired drabness of Blake: Book Illustrator, the Foundation’s earlier venture. And each plate is unnecessarily encumbered with a heavy gray panel of paper tone which creates arbitrary edges that correspond neither to the edges of Blake’s plate (often hard to locate) nor to the edges of the image or of the paper. In using a very light gray tone and avoiding such misfitting, the printers of VFD (Blake’s Visionary Forms Dramatic) and of TIB set precedents ignored by the printers and editors of ABF and BNL. The latter make a to-do about the “overall grayness” of their halftones—which are even darker than those of the ABF—but ignore the question of misfitting borders and mistakenly suppose that the only alternative is to go to the other extreme. “All but two plates of America here are reproduced twice, once on each side of the page,” they announce. “The starker, less ‘harmonious’ reproduction is taken from a high-contrast photographic print; the grayer, muddier reproduction is taken from a medium-contrast photograph and printed as a standard halftone.” The contrast of the first is increased by the process called “dropout halftone.” Both employ a halftone screen of “a good 150 lines per inch.” (Why not try the dropout with the medium-contrast, or the high-contrast without dropout? If the student of these materials is a printer, he will see the purpose of seeking the golden mean.)
A screen any coarser would hardly be good enough. The ABF halftones use a screen of 300 lines per inch. TIB uses the 150-line screen, VFD the 300-line. The finer paper in VFD makes a greater difference than the difference of screens. In the reproduction of such simple black-and-white images as the uncolored pages of America, these differences prove to be of almost no significance. The canceled plates, with some shading and pencil lettering added, are another matter. Significant differences appear when we compare the reproductions of these. For canceled plate b, the BNL halftone is the most satisfactory; it reveals the shading in the bottom left corner to have been made by unlifted pencil, a detail less evident in the ABF halftone, and scarcely to be guessed at from the TIB. The differences of screen matter less than those of photographic exposure or inking. Reduced size is a part of the trouble with the TIB image; yet the still smaller image in VFD (with better paper and 300-line screen) is more adequate. If we look now at the 150-line dropout halftone on the reverse side of the BNL sheet, we see that the penciling has lost its shading variations, while details in the human and dragon forms are blackened out. (If I were using the BNL sheets in class, I’d use the halftone pages for the illuminations and turn to the dropouts for the reading of the text.)
Carelessness can defeat any process, however. The 300-line ABF reproduction of canceled plate c ought to be best for the textual revisions penciled between lines and in the margin, though ideally the photograph should have been infra-red. But the ABF plate-maker picked up a fleck of dust begin page 124 | confusingly attached to the c of “anci[ent]” (partly cut away) beside line 11, making the ABF print inferior in clarity to all but the BNL dropout.
It is a bit disconcerting to see, when you hold them to the light, that the BNL dropouts vary in size from the halftones they are printed behind. Curious about the closeness of size to the originals, I have made a table of comparative measurements, using the Berg Collection copy L as base. (When the image edges were vague, I measured between two precise points such as the top and bottom of the text lettering.) All facsimiles, including that of the Blake Trust (of copy M), vary somewhat in size from each other and from the original. The Blake Trust plates are almost consistently a quarter of an inch shorter than the originals, with three a bit above and two a bit below that. The ABF plates are all short, with a mean of 5/16″ shorter than copy L and plate 7 as much as 11/16″ shorter. The BNL halftones are twice exactly right and most of the time slightly taller or shorter by a small fraction (a mean between ⅛″ and 3/16″ taller), only twice going as variant as ⅜″ taller. The BNL dropouts have a mean of ¼″ shorter than L, with plate 15 a half inch shorter and plate 9 shorter by 9/16″. The figures are puzzling but reassuring. One must add, however, that in BNL the canceled plate d (“Thiralatha”) is printed on its side and almost two inches too wide, obscuring the fact that it is a fragment, or rather that the picture is painted over a fragment of a printed page. In ABF the Thiralatha plate is the right size, and it is in color—but with too coarse a screen, 133 lines, which makes it blotchy. (I haven’t compared the original for accuracy of colors.)
Now let’s consider the quality and usefulness of the supplementary material and commentary in the ABF editions. It is a shame that only the costly edition lives up to the offer of “Materials for Study.” The “Editorial Comments” by Easson claim a “student-oriented” purpose yet admit that only the limited edition goes beyond serving “the general reader.” Nor is it plausible to argue that the difference of editions is one simply of “formats.” It is plainly one of contents, the reason: money. Many publishers today are putting their clothbound prices sky high (this happened to TIB) to keep the paperbound prices low—but at least without robbing the paper “format” of essential matter. May I urge the fourteen eminent Directors of ABF, whose names use up a whole leaf of paper, to reconsider this policy and to make the difference between the cheap and costly versions of the next Volume one actually of mere format, i.e., quality of paper and binding, and not of contents?
The Foundation’s Directors might also assert some authority to give better copy-editing attention to the “bibliographical essays” which they have invited “leading scholars” to contribute, if they wish them to keep in the lead. The “Bibliographical Introduction” by G. E. Bentley, Jr. for this first volume seems to have been put together and sent to press in great haste. True, even some of its flaws can be put to use by an ingenious teacher, if only as pitfalls to be escaped by learning the wariness that transforms a general reader into a scholar. Here is an example. “Blake,” we are told on page 7, “evidently often sold America and Europe together, for eight of the twelve known copies of Europe were bound with America, and America copies O and P were clearly produced as companion copies to Europe copies K and M.” The alert student will learn from the next page that copy P was produced posthumously “by Frederick Tatham” and so rule it out as evidence of Blake’s packaging. Then, to get any further, he will hunt up a copy of the Keynes and Wolf Census, where he will find listed only 11 known copies of Europe. (Has Bentley found another? He makes no comment upon the Census.) He will find that only five of these were ever bound with copies of America: B, I, K, L, M of Europe with G, N, O, Q, P of America, respectively. He will find that America Q, as well as P, was posthumously printed and bound, that America G was bound only after Blake’s death. Of Bentley’s declared eight copies, that leaves only N and O. On p. 7, however, Bentley wonders if N is a posthumous printing; on p. 8 he becomes certain of this. One copy, O, was indeed bound with a copy of Europe (K) in Blake’s lifetime—for his friend Linnell, not necessarily at Blake’s direction.
As students what do we learn from this? That G.E.B. is careless about his own declarations of fact, treating posthumous printings and bindings as acts of the living Blake? No. Bentley does not say that Blake had copies bound up together, only that their binding together is evidence that he “often sold” them together. But now suppose the copies of Europe, I, L, M, that were bound with posthumous copies N, Q, P of America, were also posthumous copies? No evidence of Blake’s selling them. Or, suppose they were copies made by Blake? Negative evidence: for if he had sold copies of America to go with them, there would have been no need for someone to make posthumous copies of America for that purpose. Perhaps we are learning something about reasoning. Even the fact that Linnell had the two works bound together doesn’t demonstrate that he acquired them at the same time or was advised by Blake to combine them. (The titles just might have suggested the idea.)
Bentley does update the Census with the information that copies D and Q, unlocated by Keynes and Wolf, are now in Princeton. Copy R he continues to list as “untraced”—but magically proceeds (p. 6) to give it a date (1797) and to describe it as colored (p. 7). Recently? The Census described it as uncolored. Also (p. 6) Bentley describes copy R as having a masked plate 4 and as the source of an Edwin Muir facsimile in color in 1887 (p. 9). Before we jump to the plausible conclusion that Bentley has been shown copy R by a private owner and been sworn to secrecy, we must note that he does not know, as he does for all other copies, what watermark the paper has or how many serpent tails there are on plate 13. It may be that he is deducing his information from the Muir facsimile and taking the precaution to refrain from assuming that the serpent tail treatment there is begin page 125 | faithful to the original. And yet, he confidently asserts that R is printed in green (p. 6), not the blue of the Muir (p. 15).
The student will also need to resist being daunted by the odd contradictions between the dating table on page 6 and the paragraphs facing it on page 7. The table shows colored copy K (in the cluster “I-L”) as dating approximately 1794. On page 7 we are told that “all eight early copies” are uncolored, giving the cluster as “C-I, L” and thus removing K from the early copies; then K is designated one of the “late copies.” But in the sentence beginning “There is evidence that the late copies were colored,” the circular hopping about of dates for A, K, O, R does indeed daunt this student. The fact that each is given a firm approximate date on the table (except that the date for A is queried) only increases the confusion.
A different kind of resistance is set up by the garbled citation, in footnote 4, page 4, of a passage from my textual notes in Poetry and Prose (p. 724) which may or may not discourage the student of Blake’s materials from looking up the reference but has certainly kept Bentley himself from calling attention to the points he obscures by misquotation. “Erdman,” he says, “finds, somewhat implausibly, that the ‘spirit and quality of drawing’ and the tone of plate 4, lines 19-20 suggest ‘1794 or 1795 as the date of etching.’” Put that way, the remark is more than merely “somewhat” implausible.
What I actually wrote was: “The final plates seem very different in spirit and quality of drawing from the canceled ones”—meaning that all plates in the complete copies of America differ markedly from the three canceled plates, a-c, which represent an earlier version we may designate version One. The final version, which we may call Two, is dated 1793 on the title page and was advertised that October. The text of One already has Blake’s idiosyncratic leftward serif on its g’s which implies a date not earlier than perhaps 1791, more probably early 1792. Differences between One and Two in iconography and in aesthetic power are evident, for instance between the flag and proclamation on plate a and the upsweeping flames on its revision as plate 3, or between the text and design of plate b and their revision in plate 4. To me these changes suggest a considerable interval of time between One and Two, and this seems the sort of thing that might profitably be studied in these materials.
My other point, conflated with this one in Bentley’s garbling, was that “the ‘harp-shattering’ lines (19-20) on Plate 2”—the lines expressing the bard’s shame in his own song, his “sick & drear lamentings,”—also suggest a considerable passage of time between version One, which presumably lacked the Preludium and these lines, and version Two. Further, since when Blake finally “published” America he masked the lines of bardic lamentation, plucking up his bardic courage again, another lapse of time, between the etching and the printing from the masked plate, ought to be posited. (It would be further to the point here to note the evidence I discuss in TIB p. 392 that version One lacked the Preludium.) What the lapses of time may have been we may only conjecture; my suggesting a date of 1794-95 as the date of publication (I should not have said of etching) was meant simply to indicate the extremes of possibility: version One ca. 1791-92; version Two with bard’s lines, 1793; version Two without bard’s lament, 1794-95.
The garbling here may well have been an accidental byproduct of condensation, but it suggests that similar misconstructions may underlie other paragraphs. The confusion produced in a two-sentence effort by Bentley (pp. 9-10) to declare how “very reliable” (or not) certain modern reproductions of America may be is a model of unreliable categorizing. It singles out as the “only” reliable facsimile besides that of the Blake Trust “a facsimile edited by Ruthven Todd (1947)” (entry 9 in the list on p. 16). It considers entry 11, on the other hand, unreliable—and rightly so. But entry 11, a facsimile prepared by George Quasha for the magazine Stony Brook, was reproduced from the Todd facsimile. Bentley does not conjecture the source for the Todd but guesses copy C or D for the Quasha; yet if he knows not the source, to what original does he judge the copy to be faithful or unfaithful? Inspection will reveal that the maker of the Todd facsimile retouched the words of Blake’s text letter by letter all through, producing at least one wrong word (“bound” for “flam’d” in 3:16), whereas that word was corrected by retouching for the Quasha text. But neither of these facsimiles should be considered reliable, literally or graphically. Both also use colors of ink and paper that are none of Blake’s. Among those listed as not very reliable (for what, we are not told) is the reduced halftone reproduction of colored copy K in VFD, of which I have spoken above. It is at least photographically reliable as to the text and the details of illumination.
Bentley’s list of “Important graphic variants in America” (pp. 12-13) is meager, and sometimes imprecise. It ignores the plates of version One altogether; it describes the rolled-up mattress in plate 12, copy 0, as “a coiled snake”; it raises the question whether the pencil inscription on plate 6 of copy F, “The Slave delivered,” is “by Blake?”—as though there were no available samples of Blake’s handwriting to test against this very different hand. (It is not Cumberland’s either.) Whether other variants are to be found, which the student may consider important, is an easy and rewarding task to be set for copy M (of the Blake Trust facsimile) or K (of VFD). He will discover that human figures are sometimes painted as leaves or flowers, dismissing their narrative or symbolic function, but are sometimes revised in function. For example, in uncolored copies the naked woman in plate 15 who stands first from the right in the cluster of three among the grapes shows her left profile, her face being turned inward. In copy M a face has been painted on what was her hair, so that she looks outward, front right. An unimportant revision, or a development of the theme of movement out from the group? begin page 126 | Bentley, who has evidently compared all copies, might have given us a fuller report on these matters. But one detail which Bentley does call attention to inclines me to forgive him all his errata.
Under “plate 2” (ii) in his list of variants Bentley notes that there is “a quite different design (in the British Museum Print Room) . . . apparently inscribed[e] ‘AMERICA’ in large, obscure letters, as if for a title-page.” In the clothbound ABF this design is reproduced as plate 27, with Bentley’s (or Easson’s) caption: “An untitled drawing often identified as a rejected design for the title page of America, courtesy The British Museum.” (The lack of cross-references either way might lead the user of the paperbound “format” to go all the way to the British Museum to look.) I have been mulling over a photograph of this page, and trying it on other people, for some time. The artist’s note to himself, on the drawing, reads: “Angels to be very small as small as the letters that they may not interfere with the subject at bottom which is to be in a stormy sea & rain seperated [sic] from the angels by Clouds.” For some time I have had a note ready for the revised Doubleday text (if that ever gets published) observing that the phrase “stormy sea & rain” suggests the title pages of Visions and America; that the bottom scene is a variant of that in the America title page; that the first word of the title is “The”; that the second and third words, roughly blocked in, could have become “MARRIAGE” and “HEAVEN,” though the latter would need much more space. Bentley’s reading of the second word as “AMERICA” provides the missing clue. I see now that the word is “AMERICAN” and the third word, so much too short for my earlier guess, is “WAR”: “The AMERICAN WAR”’. One remembers Blake’s letter to Flaxman of 12 September 1800: “The American War began. All its dark horrors passed before my face . . . . ” Probably Blake was recalling his first idea for the poem that became America a Prophecy—a version possibly different far from extant versions One and Two.
Can we date this title page? Here is a matter for students, though the suppliers of material don’t tell us about possible watermark or about what may be on the back of this drawing. My own notes say it is a watercolor design for a fan, perhaps by Stothard. And I pull down from the shelf at my elbow a copy of Blake Newsletter 27 which displays this fan design on its outside cover—and contains an article by Andrew Wilton identifying the fan design as Blake’s, of “around 1782.” I pass along the shivery thought that Blake’s planning a poem on “The American War” may have begun as soon as the war ceased’. But of course it is more likely that the verso of the fan design was sketched upon only much later, when the fan had lost interest. Too bad there isn’t a lower-case g in the title.
The only other plate that seems newsworthy is plate 26, a pencil drawing inscribed “Chaining of Orc,” offered as “an analogue for plate 3[K 1].” It is that, but again the student would be helped by some dating and bibliography. The symbolism of contrasting domed and spired churches, as in Jerusalem 46, probably belongs decades after America. In Keynes, Separate Plates (1956, p. 51), not mentioned in this book, the drawing is identified as a fairly close preliminary sketch for the plate on which Blake etched the conveniently dated caption, “Type by W Blake 1812.” Keynes does mention that the scene, Enitharmon and Los standing over Orc chained, bears some resemblance to America Preludium 1. But thrown at the student this way, dateless and undiscussed, this sort of material seems more for bewilderment than study.
The “Check-List” by Roger Easson (pp. 14-21) is perhaps properly lean, but it seems odd that the fullest recent commentary on America line for line, the notes in W. H. Stevenson’s Longman Annotated Blake (1971) is not mentioned, though his unpublished dissertation is (p. 18, footnote).