University of Virginia Library

The Proof-Reading of the First Folio Text of Romeo and Juliet
by
Charlton Hinman

One cannot yet speak of the printing-house history of the First Folio of Shakespeare with any approximation of the finality that may soon be possible.[1] When a great many of the surviving copies of the Folio have been carefully collated throughout, so that we can be quite sure that at least most, if not all, of the variants produced in the course of the printing have been discovered and recorded—only then can we hope to reconstruct the full story of the proof-reading and stop-press correction that took place in Jaggard's shop during the printing of the volume. And only then can all of the editorial implications of the facts thus brought to light be properly understood.

Few, therefore, save the most tentative of conclusions can at present be advanced as to the ultimate bibliographical and textual import of the variants in the First Folio. But the detailed collation of a mere twenty copies through about a third of the volume has already turned up variants in great abundance; and certain observations may here be made, if only to suggest some of the possible implications latent in one small group of variant readings that are evidently characteristic of some parts of the First Folio—and as evidently not characteristic of other parts. A convenient point of departure for these observations is offered by a newly found proof-sheet for one page in Romeo and Juliet.

Three First Folio proof-sheets have hitherto been known, and may be briefly reviewed as follows.

I. Proof-readers' marks calling for some 19 changes in one page of Antony and Cleopatra appear on the verso of a now disjunct single leaf from what was once the full sheet XXI:6. An examination of the 38 First Folios presently available for study at the Folger Shakespeare Library reveals: (a) that two of these 38 copies, some 5%, show the uncorrected state of XX6v (page 352); and (b) that the conjugate forme-page, XXIv (page 341), is invariant.

II. Folger First Folio No. 47 shows marks on VV3r (p. 333, in Othello) calling for nine corrections. One of the errors here marked must have required the proof-reader to consult his "copy": correction called for the substitution of one whole line for a very different one in the earlier state. Four of the 79 Folger copies, again about 5%, show the uncorrected state of both pages of the forme (VV3r:4v), though neither Copy No. 47 nor any other shows proof-reader's marks on VV4v (page 336).

III. Folger No. 48 has marks on qq6v (p. 292, in Lear) requiring five corrections. The uncorrected state appears in about 7% of the Folger copies. The other forme-page (qq1r, p. 381, in Hamlet) is variant, but no Folger copy showing the earlier state of this page is marked for correction.[2]

A fourth set of proof-reader's marks has now come to light. Forme ff1v:6r (pages 62 and 71) in Romeo and Juliet exists in two states. Nos. 24, 26, 38, 50, and 79 of the thirty-eight Folger copies at present in the Folger Library show the uncorrected state; and the two errors that characterize this state of ff6r (p. 71) are marked for correction in Folger No. 50 (see Cut I). The variant readings of the forme are as follows (uncorrected states first; positions identified by page-number, column, and line in the Folio and within parentheses by act, scene, and line in the Globe edition):

• p. 62, col. b, line 5 (II, iv, 124) [qua-/] th aGentleman:
• [qua-/] t ha:Gentlemen,
• p. 71, col. a, line 8 (IV, i, 37) own
• now
• p. 71, col. b, line 38 (IV, ii, 2) Coekes
• Cookes

On page 62, three kinds of alterations are made. "Gentleman" is made plural and both the spacing and the punctuation are changed. Yet we have in essence only one variant reading here: the good Angelica, having first reacted to the previous speech by Romeo, is made by the correction to begin with a properly deferential salutation the question that she now addresses to all three of the gentlemen before her. Although the corrected reading is substantially that of the copy from which the Folio text was set, the Quarto of 1609, the change almost certainly reflects only the independent authority of the proof-reader. (Had copy been consulted, the errors in the line above would doubtless have been mended as well.) The corruption of the first setting is only a little less simple and obvious than that in the two readings that were corrected in the other forme-page.

As in the two other instances in which the marked page appears in an extant Folio, the Romeo and Juliet leaf bearing the proof-reader's marks has its normal conjugate and the whole sheet was evidently bound into the volume in a perfectly regular manner, like any unmarked sheet. And as in both the other instances the conjugate forme-page is characterized by a single reading that was corrected, although the error was not marked for correction in the copy that served as proof for the other page of the forme. Thus, it will be observed, we now have further confirmation for the hypothesis first advanced in 1942:[3] that, in effect, the proof-reading unit for at least some parts of the First Folio was normally the single page, though the correction unit was of course the forme, both pages of which were ordinarily corrected in a single unlocking operation.

But the present proof-sheet is chiefly of interest for other reasons and in conjunction with other evidence of the kind of proof-reading to which Romeo and Juliet was subject. The factors that differentiate the press-correction of Romeo and Juliet from what we find (or fail to find) elsewhere in the volume may prove, indeed, to be far more significant than the common factors.

It may be remarked, first of all, that the evidence for identifying ff6r in Folger No. 50 as in fact a proof-sheet is anything but abundant. The errors marked, and those alone, were the ones corrected, to be sure; and that the marks on the Romeo and Juliet page (see Cut I) were the standard symbols for indicating transpositions and obvious single-letter errors is made perfectly clear by the repeated uses of the same marks for such errors in the other proof-sheets. (See Cut III for examples in the Antony and Cleopatra proof-sheet of both kinds of marks that appear in the Romeo sheet. Cut II shows the corrected state of p. 71; Cuts IV and V

show the uncorrected and corrected states of p. 62.) Yet the ink of the Romeo proof-marks is now darker than that found in all the other proof-sheets; the transposition mark is not, as elsewhere, marginal; and there is some slight difference in the form of this symbol here and on p. 352 in Antony.[4] That these marks in Romeo and Juliet and those found in the other proof-sheets were made by different hands obviously cannot be maintained on the basis of such evidence; but the suspicion here bred is fostered by various other considerations.

Certainly it is rather surprising, to begin with, that so many copies show the uncorrected state of ff1v:6r—five out of 38, nearly one copy in every seven: something over 13%. This phenomenon might, of course, be supposed the result of mere chance, except that it is encountered elsewhere in Romeo and Juliet. Variant forme ee3r:4v is found uncorrected in five copies out of 37; ff2r:5v in four; and uncorrected ee2v:5r in 21 out of 36 copies, almost 60%. Certainly, it would seem, there was no very eager concern in Jaggard's shop over the errors that might mar large numbers of the copies of the finished book in Romeo and Juliet; for it is manifest that the proof-reading for this play was frequently not undertaken until long after the first pulls on a new forme had begun to come off the press. And this lack of concern is reflected also by the small number and inferior quality of the corrections that at length were made—and by the hosts of errors, many of them apparent at a glance, that yet remain in the "corrected" pages of this play, not to mention new error sometimes introduced in the very process of correction.

One of the most striking things about the Romeo and Juliet proof-sheet, of course, is that it marks only two small errors for correction; and we have seen that only one reading was corrected in the other page of the same forme. This kind of thing is not invariably characteristic of the variant formes in Romeo and Juliet, but it is certainly the general rule. One forme (ee3v:4r) shows fourteen variants in one and eleven variants on the other of its two pages. Throughout the rest of the play, however, the largest number of corrections in any single page now known to be variant is a mere three; and four of these pages have but one correction each. Combined, moreover, with this paucity of individual variants is a startling abundance of variant pages. A collation of only twenty copies has turned up 12 variant pages out of the 25 pages of the play; and it is not impossible that we may eventually discover that every single page of the Folio text of Romeo and Juliet was subjected to stop-press correction.[5]

The treatment accorded Romeo and Juliet during the printing of the First Folio seems to have been characterized, therefore, by a great deal of rather tardy proof-reading that resulted in very few actual corrections.[6]

The remarkable contrast between this kind of treatment and the kind accorded the plays immediately before and after Romeo and Juliet in the Folio (Coriolanus and then Titus Andronicus precede; Timon, Julius Caesar, and Macbeth follow) is perhaps best pointed up by the following three facts:

• (1) The first two and a half leaves of the ee quire of the First Folio contain the last five pages of Titus Andronicus, Romeo and Juliet beginning on ee3r. Hence each of four formes of the ee quire contains one page of Titus and one of Romeo. Now every variant forme throughout Romeo in which both pages contain Romeo text is variant in both pages of the pair. But two of the twelve variant pages of Romeo, ee5r and ee6r, are in formes shared by Romeo and Titus. In each of these cases the Titus text is invariant, correction having been confined—at least in the copies so far collated—to Romeo and Juliet.[7]
• (2) The collation of twenty First Folios, though revealing numerous variant pages in Romeo and Juliet, discovers not a single variant in Coriolanus, only four variant pages in Titus, no variants whatever in Timon, four variant pages in Julius Caesar, and one variant page in Macbeth: a total of nine variant pages in these five whole plays as compared with twelve in Romeo and Juliet alone (though there are a few more individual variants in the nine pages than in the twelve).[8]
• (3) Perhaps most clearly indicative of the kind of printing-house treatment accorded Romeo and Juliet in contrast with that given most of the plays just before and just after it in the Folio is the inferior quality of the

  ---

  68

  
  
  
  Romeo text notwithstanding the relatively large number of its pages now known to have been proof-read and corrected.

On this last point the present writer can speak only with "a scattering and unsure observance," for he is by no means expert on the texts of the first six tragedies in the folio. Yet he could not fail to notice, while collating these plays, that Romeo and Juliet is peculiar in the abundance of obvious error that both its corrected and its uncorrected states show—and which, indeed, seems sometimes to infect the very process of correction itself. The proof-page, for example, shows the correction of two small mistakes but leaves unmarked and uncorrected at least a dozen others where corruption is just as evident. Column a alone has, among others:

• Godſheild: I for Godſhield I
• ſtreames " ſtraines
• heart, and " heart and
• ſtay " ſlay
• thinglike " think like
• Fro of " From of
• chappels " chapless

And most of these errors originated in the Folio: only two of the seven are present in the copy, the Quarto of 1609.

The other page of the forme is about equally corrupt. Even its one variant speech illustrates the carelessness and incompletenes of the proof-reading to which it was subject; for although the second line of the speech is corrected—except that the spacing is never made quite right (see Cuts IV and V)—the comma after "to" in the line above should patently come after "mar"; and it is almost as apparent that a word has been omitted before "ſaid" (Q 1609 has "well ſaid"). Probably, too, the final letter of this line should be an "o" rather than an "a", though here the copy is accurately represented. Again, the single correction made in p. 70 (ff5v) is of a minor error in spacing—"boſom ehenceforth" for "boſome henceforth", as in the Quarto—yet numerous equally obvious and scarcely more trifling mistakes remain:

• bowles for bowle
• ride " tide
• ſtraue " ſtarue
• Hlacke " Alacke
• talke for talk'd [probably; though here F=Q]
• Benig " Being

—to mention only a few, and only one that is derived from the Quarto. Even the page that shows fourteen variants (ee3v, p. 54) still shows in its corrected state a good many obvious errors not found in the copy:

• others for other
• Love,is " Love is
• Be. n " Ben.
• henec " hence

among several others. And beside such failures to correct at all may be set examples of failure to correct correctly. In variant page 74 (gg1v) "Tortyrs" is changed to "Tortoyrs"—the word "Tortoys" [ tortoise], as in the Quarto, having been intended. An "o" was rightly added after the second "t" but the absurd second "r" of the original setting was not removed. Whether the marker of the proof-sheet for this page was as much at fault as the "correcting" compositor we can only guess. Finally, in page 55 (ee4r) the line becomes where what was intended was evidently (This is the reading of the Quarto, where, however, part of the spacequad between "ſhew" and "well" is slightly inked. This inked spacequad does not look much like a comma, but the Folio compositor seems to have been scarcely more alert when he set this line than when he corrected it.) Our proof-reader doubtless indicated that the word "ſhew" was to be substituted for the "ſhell," of the first setting; but it was substituted for the similar "ſhall" two words earlier instead. Here, surely the compositor nodded badly; though again it is possible that the over-casual marks of the proof-reader contributed to the ultimate muddle. In any event it is clear that the proof-reading of this line only produced, at last, a reading even worse than the one originally set.

We have been considering, of course, only errors of the simplest kinds, those due as a rule to the minor lapses of a compositor. Obscure and complex errors, and corruptions resulting either from the peculiarities of the copy or from the more serious failures of the compositor, have not been in question (nor were, evidently, to the Romeo proof-reader); and it should be remembered that a text quite free of obvious errors may yet be very corrupt—though it seems but reasonable to suppose that type-setting that is full of small errors is likely to contain large ones as well.[9] Hence

the phenomena described above do not of themselves provide any sound basis for final conclusions as to the essential qualities of the texts of the six plays mentioned. But the phenomena are nevertheless suggestive. The plays just before and just after Romeo and Juliet in the First Folio exhibit no such abundance of obvious error as we find here—nor so many variant formes; nor such large numbers of copies showing the uncorrected states of these variant formes; nor so much evidence that the proof-reading was characteristically as careless as it was tardy; nor such signs of badly executed correction. To some extent the typesetting, but above all the proof-reading that went into the Folio text of Romeo and Juliet, is of a special order, clearly differentiating this play from most of the plays nearby. A particular attitude toward the Romeo text would seem to have prevailed in the printing-house during the preparation of this play for the Folio—as if the compositor and the proof-reader alike, though having their jobs to do, were somehow unable to discharge their appointed tasks with their usual conscientiousness—as if they were but too well aware that they were here perpetuating only the reprint of a reprint of the play.

As I insisted at the beginning of this paper, it would be imprudent to infer from the variant readings of the First Folio so far known any but very tentative conclusions. The evidence that has been presented will serve at least to indicate that different plays were treated with very different degrees of care in Jaggard's shop during the printing of the First Folio, the relative attentiveness of the proof-reader[10] being one of the significant variables. Perhaps it will indicate also that a study of all of the Folio's variants, when these have at length been found, may have a good deal to teach us, not alone as bibliographers and students of printing-house practices and techniques, but also as editors of the various plays of Shakespeare. And, finally, a very specific suggestion may be made. The conjecture was hazarded above (see note 8) that at least some of the plays printed for the Folio from manuscript copy were given much more careful treatment than were certain other plays that were set, like Romeo and Juliet, from printed copy. The basis for this inference is admittedly not yet very solid: it consists in part in the mere scarcity of variants so far discovered in such plays as Coriolanus, Timon, and Macbeth. But let no prospective editor of Coriolanus, for example, forthwith conclude that the Folio text of this play is not only remarkably good but also invariant. For if variants turn up, as they probably will, they may prove more interesting than those minor tidyings-up so characteristic of Romeo and Juliet. As in other things, the rarer the more likely to be valuable.

Notes