University of Virginia Library

II

In January of 1601, the publisher Thomas Bushell entered in the Stationers' Register for his copy of "A booke called the plaie of Dcor ffaustus."[7] The first known edition of Faustus is a black-letter quarto, known as A1, printed by Valentine Simmes for Bushell in 1604. The A1 quarto collates

A-F4, but the unique copy in the Bodleian Library, classified Malone 233(3), lacks F4, which was presumably blank. In his study of the printing of A1, Robert Ford Welsh notes that identifiable headlines recur in groups and maintain their relative positions (except on sheet E).[8] Apparently, the quarto was printed with two skeleton-formes throughout, the skeletons switching formes in sheet E. Welsh suggests that the switch of the skeletons at E may have been due to some break in the printing process between sheets D and E ("Printing," p. 111). This may well be, for there is another anomalous change between sheets D and E that has not been previously noticed: in sheets A-D each page has 37 lines of type, but sheets E-F have only 36 lines per page. Such a change in the number of lines per page might be taken as a sign that Simmes had shared the printing of the quarto with another printer.[9] Moreover, sheet F of the only surviving copy has a different watermark from the previous sheets. Such a change in paper might also be suggestive of a change in printer. However, the watermark on sheet F is found in some copies of Q3 of I Henry IV (STC 22282), which was also printed by Simmes in 1604; so it seems likely that Simmes simply ran out of the paper that Bushell had supplied for Faustus, and so had to use another stock that he had on hand for the final sheet F. Such a hypothesis supports Bowers's contention that sheet A was not cast-off and set last, but was instead the first to be composed (Complete Works, II.146). Moreover, the 82 mm. black-letter font used in sheets E-F is the same as that used in A-D, and the fact that several identifiable types from the earlier sheets recur in the later effectively rules out the possibility of shared printing.[10]

The A1 quarto was set in type by two compositors, first distinguished by Welsh who labeled them X and Y. Welsh's Compositor X abbreviates speech headings and punctuates them with a period, uses an upper-case "E" in Exit directions, prefers -ea- spellings in words like year, dear, and chear, prefers bloud over blood, and uses -ll spellings in words like will, shall, and hell. Welsh's Compositor Y frequently uses unabbreviated and unstopped speech headings or abbreviated speech headings punctuated with a colon. In contrast to X, Y uses a lower-case "e" in exit directions, prefers -ee- spellings in words like yeer, deer, and cheer, prefers blood over bloud, and uses single -l spellings in words like wil, shal, and hel. Although Y's habits may reflect simple spelling preferences, they should also be seen to represent

work-saving expedients. The black-letter font (probably of French origin) included -oo- and -ée- ligatures. Y apparently had no desire to set two pieces of metal if he could get away with setting only one: thus his preference for blood over bloud, -ee- over -ea-, and -l over -ll.

Welsh surveyed all of the dramatic quartos Simmes printed from 1603 to 1605 and concluded that "we have no relevant knowledge of either Compositor X or Compositor Y outside of the Faustus quarto itself" ("Printing," p. 126). However, the habits of Welsh's Compositor Y resemble those that W. Craig Ferguson had previously noticed in Q1 of 2 Henry IV, printed by Simmes in 1600: unabbreviated speech headings and exits with lower-case "e". Moreover, the compositor of 2 Henry IV, designated Compositor A by Ferguson, resembles Compositor Y of Faustus in that he does not distinguish names by setting them in a contrasting type font.[11] Ferguson's essay has been supplemented by a series of articles by Alan Craven which, taken together, claim that Compositor A set the type for quite a bit of Renaissance drama as we know it: all or part of Q1 Richard II (1597), Richard III (1597), Q2 Richard II (1598), A Warning for Fair Women (1599), An Humorous Day's Mirth (1599), Much Ado about Nothing (1600), The Shoemakers' Holiday (1600), Q2 The First Part of the Contention (1600), Q1 2 Henry IV (1600), Q1 Hamlet (1603), Q3 I Henry IV (1604), and Q1 and Q3 The Malcontent (both 1604).[12]

Ferguson and Craven attach a great deal of importance to the unusual (although by no means unique) compositorial habit of setting unabbreviated and unstopped speech headings; any dramatic quarto printed by Simmes between 1597 and 1604 in which such speech headings appear is automatically assigned to Compositor A. The assumption that only one of Simmes's compositors would ever have set an unabbreviated speech heading without adding a mark of punctuation is so strong that little significance is attached to some of the manifest differences between the texts attributed to Compositor A. Q1 Hamlet, for instance, is assigned to A despite the fact that "proper names (characters and places) in the dialogue are often set in contrasting italic type, a practice never used in 2 Henry IV and Much Ado" (Craven, 1973; p. 40). And even though one of A's hallmarks, setting lower-case exits, is nowhere to be found in Q3 of I Henry IV, Craven asserts that this evidence should not "raise doubts" about the identity of the compositor (1979; p. 188).

As the dramatic texts for which Compositor A was assigned responsibility began to multiply like Falstaff's men in buckram, A1 Faustus was added to the list by Craven, who asserts (in a brief footnote) that Welsh's Compositor Y

and Compositor A are one and the same (1973; p. 49n). The identification was accepted by Bowers, who, in the textual introduction to his Cambridge edition of Faustus, substitutes A for Welsh's Y. There are indeed a few points of similarity between Compositor Y in Faustus and Compositor A: both show a preference for eie over eye and the -nesse suffix over -nes. But these minor similarities are outweighed, it seems to me, by major differences. Welsh's Y has a nearly absolute preference for the spelling blood (8 occurrences; the spelling bloud appears only once in Y's share of Faustus). However, Craven has shown that A has a nearly absolute preference for bloud (37 occurrences in Q1 Richard II and 30 occurrences in Richard III; the spelling blood does not occur in A's supposed share of either of these two texts). Moreover, whereas Compositor Y has a marked preference for singe -l spellings, Ferguson characterizes A by his preference for the longer -ll form (1960; p. 23). Craven, intent upon proving that A and Y are the same workman, admits that single -l spellings are indeed "rare" in A's work, and is forced to dig back to Q2 of Richard II (1598) to find a single example of this habit among the multitude of texts he attributes to A (1973; p. 49n).

Although the characteristics of Compositor Y in Faustus do resemble, to a certain extent, those of Compositor A in Q1 2 Henry IV (1600), we ought to bear in mind Peter Blayney's caveat that "similar habits found a year or so apart in the same printing house do not prove identity."[13] Of the dramatic quartos that Simmes printed in 1604, the same year as A1 Faustus, Craven has assigned three to Compositor A. These three texts do contain the unabbreviated and unstopped speech headings necessary for attribution to A, but the other distinguishing characteristics of Compositor Y appear not at all (see table 1).

Table 1: Compositor Y's Characteristics in A1 Faustus Compared with Other Play Quartos Printed by Simmes in 1604

               

Text	unabbrv.		Names	Single -l
	Unstopped	Lower-	not set in	Spellings
	Speech	case	contrasting
	Headings	exits	font
A1 Faustus	Yes	Yes	Yes	Yes
Q1 Malcontent	Yes	No	No	No
Q3 Malcontent	Yes	No	No	No
Q3 I Henry IV	Yes	No	No	No

I suspect (although I cannot prove) that, over a period of eight years, Simmes may have employed more than one compositor who did not abbreviate and did not punctuate his speech headings when he set a play quarto, and that "Compositor A" as he is currently constructed was probably a number of different workmen. In any case, I can find little evidence to support Craven's identification as it pertains to Faustus, for it seems clear that Welsh's Compositor Y is not Compositor A. Moreover, the habits of X and Y in

Faustus do not resemble the recurring patterns of the two other identified compositors (B and S) who worked for Simmes.[14]

Although Compositors X and Y have not been found at work on any other Simmes book, they can be readily distinguished within the text of A1 Faustus where their individual habits form clear and regular patterns. The alternating groups of speech heading punctuation on sheet C may serve as an example (see table 2).

This sort of patterned alternation might in itself be sufficient evidence from which to deduce that compositor shifts have taken place at C2r:25, C3r:28, C3v:31, and C4v:16. When the punctuation data are combined with the spelling evidence, the compositors' stints can be charted throughout the quarto with some accuracy (see table 3).[15]

Curiously, the compositors' stints generally do not begin (or end) at the beginning of a new page or sheet. Instead, the compositors frequently change in mid-page, often at the beginning of a new scene or at an entrance direction for a character within a scene. The shift on signature C2r:24 comes at Enter with a diuell drest like a woman, with fier workes; on C3v:31 at Enter Lucifer, Belsabub, and Mephostophilus; on D1r:28 at Enter Faustus and Mephastophilus; on D3r:9 at Enter Robin the Ostler with a booke in his hand; on E3r:6 at Enter to them the Duke, and the Dutchess, the Duke speakes.

As Bowers observes, "this practice is so consistent and so odd as to call for explanation" (Complete Works, II.147). Welsh found evidence of type-shortages indicating that the compositors were setting seriatim, rather than by formes, and Bowers notes that this conclusion is supported by the nature of the stints which are not confined to formes. Welsh assumes that the two compositors worked chiefly in turn rather than simultaneously, each setting a scene of the text ("Printing," pp. 114-115). D. F. McKenzie claims that such a procedure for shared setting would have been normal in a seventeenth-century print shop: "normally, even when two or more compositors worked on a book, they did not work together setting sheet and sheet about. What usually happened was that one took over where the other left off."[16] So strong is McKenzie's insistence upon "a fairly accurate definition of 'normality'" that there can be little doubt that this was the way that compositors usually worked. However, the irregularity of the stints in Faustus is so unusual that we may be justified in inquiring as to the possibly abnormal circumstances in which they were produced.[17]

Welsh concludes that only one type-case was in use because both compositors suffered shortages of black-letter W, and he imagines that two type-cases could not both contain an inadequate supply of this letter ("Printing," pp. 115-119). But Bowers points out that such W shortages were a congenital

difficulty in French fonts and rightly observes that the question of whether the compositors were setting from one case or from two, in turns or simultaneously, can only be proven by a detailed analysis of identifiable broken types. However, "in the absence of more precise findings from the distribution and setting of identified types," Bowers assumes that the compositors were setting simultaneously. He then conjectures that the irregularity of the stints If, as Bowers suggests, the printer's copy was a manuscript that was particularly difficult to cast off, it might well have been easier for the two compositors each to set individual scenes in their galleys, which could then be combined into pages during imposition.[18]

Bowers's attractive and sensible conjecture has important ramifications for our understanding of the nature of the underlying copy that have not been previously realized. Compositors' stints are necessarily limited, and sometimes defined, by separable sections of the printer's copy.[19] If the compositors were setting simultaneously from separate cases, the copy itself would have to have been physically separated and divided between them. In order for each compositor to be given a successive scene from the manuscript copy, we would have to suppose that every fresh scene began on a new manuscript page. And Bowers himself elsewhere surmises that the only type of manuscript in which this might be true would be "the foul papers of a collaboration" ("The Text of Marlowe's Faustus," p. 199). The issue of whether or not the compositors were working simultaneously on the A1 quarto can now be seen to have considerable implications. Simultaneous setting might point to the use of foul paper copy.