DECEMBER 18, 2016
BRIAN VICKERS’S RESPONSE to my review of his The One King Lear runs to well over 13,000 words, exceeding the length of my already unusually substantial review by more than 4,000 words. Much of his reply is taken up by an exhaustive summary of his book. I have discussed the arguments of that book in detail in my review, and he adds virtually no new evidence or arguments to address the problems I identified. As a piece of writing, his response exemplifies the qualities of The One King Lear: it, too, resorts to invective and ad hominem attacks; it, too, misrepresents and misreads the text it discusses; and it, too, confuses bibliographical and literary-critical modes of argumentation. In these senses, his piece speaks for (and against) itself, and may not merit a detailed rebuttal. However, I think it affords a helpful opportunity to discuss some fundamental fallacies and to suggest how this kind of debate might proceed more productively.
The One King Lear, as I previously wrote, attempts to resurrect the notion that there was once a version of the play that contained all the lines printed, variously, in the 1608 Quarto and the 1623 Folio. For earlier generations of scholars, such a belief rested on the conviction that the 1608 volume prints a seriously flawed version of the play transmitted by memorial reconstruction — that is why lines only present in the Folio are missing from it. That assumption always had its critics, and in the 1980s, its credibility was dealt a series of decisive blows, most influentially by Peter Blayney’s in-depth bibliographical analysis of the Quarto. Blayney showed that the many flaws of that text could be explained as problems arising in the printing house of Nicholas Okes. No longer could those flaws be used as evidence that Lear was a memorially reconstructed text. Once that hypothesis became untenable, though, what collapsed along with it was the notion that lines only present in the Folio were omitted from the Quarto (rather than added to the play at a later date): there no longer was any reason to believe that the 1608 text was essentially incomplete. Along with the status of the Quarto, the status of the Folio changed: the later volume, it seemed, was printed from a later manuscript, one reflecting changes, cuts, and revisions that took place during the 18 or so years between the play’s first performance and its appearance in the 1623 volume.
How to treat the Folio text still remained an open question. Some argued that the later version represents Shakespeare’s revision of King Lear; this case was made most extensively in a collection of essays so influential that no matter whether one agrees with its conclusions or not, it certainly needs to be considered a “landmark” publication. (Vickers’s thinks my use of this term makes me a “disciple” of Gary Taylor and his collaborators. I have spoken to Taylor twice in my life and am by no means persuaded that the 1623 Lear is primarily Shakespeare’s revision. I am, however, persuaded that it is a revision, as are many other scholars who otherwise strongly disagree with Taylor et al.) I have serious reservations about Vickers’s discussion of the Folio text and its genesis, and I discuss some of them in my review. But he is right that I do not dwell on that aspect of his book extensively.
The reason for this is straightforward, and I give it in my original essay: what the Folio text represents is open to interpretation, and Vickers’s case, grounded in his own reading of King Lear, is one such interpretation. His reading, however, relies on the conviction that there was a unified original text from which the Folio version departs; and that conviction in turn depends on a theory of how the Quarto text came into being. In other words, Vickers’s account of the Folio Lear is not a reading of that text on its own terms, but a reading that compares it to the (non-extant) unified One Lear. Its logic, essentially, is to examine each Folio cut in turn, assess its merits vis-à-vis a specific interpretation of the play, and declare for or against its authority on that basis. The question the book insistently asks is “does this difference improve the play?” — and if the answer is “no” (as it mostly is), the difference cannot have originated with Shakespeare, and thus ought to be dismissed. Such a procedure, author-centric though it is, and characterized as it may be by an unshakeable faith in Shakespeare’s ability to always write the play Vickers wants him to write, is not inherently inconsistent. But it relies on a number of fundamental assumptions the reader has to share: that King Lear has the preoccupations and meanings Vickers suggests it does; that Shakespeare would not have departed from that particular vision, once realized; that Shakespeare did not tinker; that Shakespeare was incapable of aesthetically weakening his own plays in revision; and, most importantly perhaps, that all lines only found in the Folio are not revisions or additions, but original parts of the text merely omitted from the Quarto.
That is to say, the account of the Folio text offered in The One King Lear cannot be critiqued independently of the book’s hypothesis about the Quarto text: the basic contention that there are no additions in the 1623 version is little more than a fiction unless it can be established that the Folio-only lines were originally in the manuscript from which the Quarto was printed. If that hypothesis does not stand, the Folio reading Vickers offers is incomplete at best.
I am agnostic when it comes to the authorship of the 1623 text, and certainly open to persuasion. I might as well state, for my interlocutor’s sake, that I am not at all convinced that Shakespeare’s was the main hand behind the differences between the two texts — and if that is a “post-truth” attitude rather than an honest expression of uncertainty, so be it. I am happy to side with Richard Knowles in this regard:
the question is not whether there was revision — of course there was — but who did it, and when, and why. […] Of course Shakespeare may have made all or most of the F alterations himself: just as possibly someone else or several other people at different times may have made them.
In a revealing passage in his response, Vickers justifies his argumentative modus operandi thus:
Of course, I am aware of the revisionists’ argument, that these are all lines subsequently added by Shakespeare (and not, as Syme allows, by someone else). This issue can only be settled by examining each omission in its typographical context, as I have done, and by using arguments from probability.
He then illustrates what he means by “probability” with reference to a specific instance of Quarto-Folio difference: a brief exchange between Lear and his Fool in 3.6, in which the Fool asks Lear to answer his riddle whether a madman is a yeoman or a gentleman. In the Quarto, Lear refuses the terms of the question, answers “a king” and fantasizes about devils punishing his daughters; the Fool drops his riddle. In the Folio, Lear’s answer remains the same, but the Fool supplies the actual, not especially clever, resolution — and it is only at that point that Lear starts fantasizing about the devils and his daughters. To Vickers, it seems
very probable that this deliberate juxtaposition of two mental states was in Shakespeare’s original text and unlikely that he added it at a later date. I cannot prove this and thus emphasize that my arguments are from probability. In the humanities, we can seldom claim certainty, but some arguments can be more persuasive than others.
Knowles argues that “conclusions about the authenticity of the Q and F versions must rest mainly upon critical impression.” Vickers’s claim is much stronger: he asserts that he is arguing “from probability.” But what does “probability” mean? By what standard is one aesthetic judgment more “probable” than another? Is it more probable to say that the Quarto text removes the conclusion to the Fool’s witticism? (Is it even accurate to say that without that conclusion the “juxtaposition of two mental states” Vickers describes is not achieved?) Is it less probable to suggest that perhaps the King’s Men’s primary clown, Robert Armin, wanted the Fool’s question to have an answer — and that either he or Shakespeare supplied it at some point? Is it less probable to surmise that Shakespeare thought the Fool’s riddle should remain incomplete, more baffling, in 1605 or 1606, but changed his mind later? Is it less probable to suppose that Shakespeare wrote the complete riddle, question and answer, in 1605 or 1606 and deleted the answer by marking it with a line in the margin (a common marker of deletions), but that although Okes’s compositor correctly left it out, a later scribe, making a copy of the text from a different authorial copy of the manuscript, accidentally reinstated it? Is it less probable to think that the draft of the manuscript from which the Quarto was set had that line deleted — but that Shakespeare, when making a clean copy of that manuscript for his acting company, changed his mind and kept the Fool’s answer in the text? Is it less probable that Shakespeare initially thought it would be more effective to let the Fool stop riddling with Lear immediately when he realizes that Lear can no longer engage with the game (a possible reading of the dynamics of the Quarto exchange), but reconceived the moment later and opted instead for a more callous Fool in the Folio, a joker who cares more about finishing his witticism than about Lear’s ability to play along?
In the absence of tangible evidence, none of these rough hypotheses can be adjudged inherently more or less probable. We cannot measure how likely or not any one of them is to be true. On the basis of an examination of the text, all we can establish is whether there is a persuasive case to be made for the idea that instances such as the Fool’s riddle represent cuts rather than later additions (or any of the alternative scenarios I have sketched). But persuasiveness is a much lower, and much less uniform, standard than probability. On a case by case basis, some of Vickers’s arguments may or may not persuade individual readers. However, the grounds on which those acts of persuasion would take place are those of literary criticism, not those of bibliography (let alone “typographical context”). They rest on an agreement or a disagreement about aesthetic criteria and narratives of composition. Despite this, Vickers raises the stakes further by insisting, with regard to the Quarto, that all the lines missing from that text are absent for the same reason, and owing to the actions of the same agent: Nicholas Okes. That argument, though, requires a rather different basis than the arguments about whether individual textual instances are cuts or additions. It has to take into account how stationers operated and how books were made.
In general, I personally find Vickers’s assertions about specific textual cases unpersuasive. They too often rely on claims about what he himself finds credible or plausible. But whether one accepts that some, or even many, of the Folio-only lines were in fact cut from Shakespeare’s original draft before the Quarto was printed has no bearing on whether Vickers’s broader narrative of how the 1608 volume was produced is credible.
Why? Because the two kinds of accounts rely on two different kinds of support. A critical argument about which version of a text is to be preferred can be credible even if it cannot offer a persuasive narrative of how the text was generated in the first place. For instance, I may agree that the Fool’s answer is not the kind of thing a writer would add in a revision, that it probably was in the text all along — and that it should have remained there. I may even convince myself that in holding that opinion, I am following Shakespeare’s own intentions. But to arrive at that conclusion, I do not need to feel certain that only one of the many scenarios I sketched above corresponds to what really happened. Any number of different hypotheses about composition and transmission can serve to support the same conclusion about what “should” be in the text. Conversely, even if I were to concede that it is more reasonable to think that the line was written together with the rest of the dialogue rather than as a later revision, it does not follow that I also need to concede that it was not meant to be deleted and that the inclusion in the Folio wasn’t a mistake. Again, a variety of hypotheses can support that position. Each of the differences between Quarto and Folio gives rise to a similar range of scenarios of composition and transmission that are inherently equally plausible, and we simply have no basis on which to decide that all of those cases follow the same logic, or are part of the same history of transmission. Unless we learn more about how playwrights wrote plays and how printers turned them into books, it is difficult to determine what is more or less “probable” — because we don’t quite know enough about what playwrights and printers “probably” or “normally” did.
What The One King Lear offers, in effect, is a series of literary critical statements about the differences between the two texts. It does not offer textual evidence for the editorial interventions of a printer. At most it makes a case, based on a loosely stated theory of how playwrights write, that some of the changes from the Quarto to the Folio look like restorations of lines or phrases previously cut rather than like newly added material. If one disagrees with Vickers’s assumptions about how early modern dramatists worked, there is little basis for accepting most of those assertions. But even if one shares his view, it is not possible to jump from these individual instances to either a conclusion about the specific features of the manuscript from which the Quarto was set or the place of that particular manuscript in the sequence of documents that make up the history of the transmission of King Lear. And it is equally impossible to draw conclusions from these individual textual instances about the role Nicholas Okes may have played in altering the text as he turned it into a printed book.
Instead, to present Okes as the decisive agent of textual change, Vickers needs a larger narrative explaining why this particular stationer should have done something most other printers or publishers, to the best of our knowledge, didn’t do: edit, cut, and occasionally rewrite an author’s text. If Vickers can show that it is reasonable to think that Okes (or another printer) would act in such a manner, then it becomes possible to at least consider him as the single agent responsible for the absence of all those Folio lines from the Quarto. (It would still not be enough to make the case that Okes actually cut them all; but such a scenario would then at least appear plausible.)
Are there examples of such interfering stationers in the period? Certainly. Richard Jones, unmentioned in The One King Lear, comes to mind. Jones, when publishing Christopher Marlowe’s two-part Tamburlaine in 1590, noted in a preface “to the Gentlemen Readers and others” that he had “purposely omitted and left out some fond and frivolous jestures, digressing and, in my poor opinion, far unmeet for the matter, which I thought might seem more tedious unto the wise than any way else to be regarded.” No matter how effective the comic material he chose to leave out may have been in the theater, Jones thought that “it would prove a great disgrace to so honourable and stately a history” to have it “mixtured in print” with frivolous episodes. But this is an exceptional case: Jones was one of the first publishers of plays from the commercial theater; Tamburlaine was Marlowe’s first dramatic work to be printed; and later printed drama more commonly advertises itself as more complete, as offering more text than was performed on stage. Even as early as 1590, though, and when dealing with the work of a relatively obscure author, this stationer felt a compulsion to explain why he chose to edit the text. Okes, needless to say, did nothing of the sort — despite publishing King Lear with a title page utterly dominated by the author’s name (unlike Jones, who, despite his apologetic note, does not name Marlowe anywhere in his book).
Without this kind of explicit statement from the stationer, to be convincing Vickers’s narrative would need to be grounded in something else outside the text of King Lear: information about trade practices, evidence about Okes’s treatments of other manuscripts, research into how other stationers handled printed drama, parallel cases of printers running out of paper and cutting their copy text. The One King Lear does not provide anything like such an alternative foundation.
The evidence Vickers offers is insignificant, even if he discusses it in exhaustive detail. It consists of a tabulation of “space saving” devices in three other Jacobean Shakespeare quartos and of descriptions of a handful of volumes in which it seems that some text was cut or reformatted in response to space constraints. To these, he adds one further example (of sorts) in his response.
Let me start with the latter set of references. Did printers economize space in Shakespeare’s time? Yes, unequivocally. Especially in reprints, we often find compositors reducing blank space to save paper; many instances of this practice could be cited from W. W. Greg’s Bibliography of the English Printed Drama. The example of 1 Henry IV that Vickers discusses at length in the book is unexceptional in this regard. In his response, he seems to concede that the volume is in fact exceptionally well printed — and that it was produced by Peter Short, not Andrew Wise (its publisher). Odd, then, that in the book he writes of Wise’s “compositor” (even though Wise did not have a printing press) and criticizes the volume for omitting too many spaces; using an excessive number of ampersands (“ungainly though they appear in verse”); “tolerating [an] ugly fusion of letters and punctuation marks”; resorting to “undignified” forms of compression (such as using “continuous printing,” the setting of two short speeches on the same line); and breaking with “Elizabethan printing conventions” that allegedly considered it “especially unpleasing to have entrances crammed into a line ending.” The 1 Henry IV Quarto he describes in The One King Lear is marred by all sorts of blunders — some of them “compositorial failures” more “glaring” even than Okes’s. In his response, Vickers now claims that it is not surprising that this same book is “highly regarded for its quality,” since, unlike Okes, “the publisher or printer had already taken care of the space problem” competently enough that aesthetic damage could be avoided. But if Short’s Quarto is in fact a high-quality piece of work, then “saving space” and the compositorial strategies associated with it are common rather than exceptional; if they yield “ugly,” “ungainly,” or “undignified” results, then their aesthetic shortcomings are those of early modern typesetting as a whole.
I dwell on this moment in part to illustrate the degree to which Vickers’s response misrepresents his own arguments in The One King Lear. More importantly, though, this instance also shows his idiosyncratic understanding of early modern printing practices. Adding or omitting spaces before or after punctuation marks, using ampersands and similar symbols and contractions, spelling out numbers or setting them as digits, continuous printing (i.e., running two short verse or prose speeches into a single line) — these were all perfectly standard means of justifying individual lines. At times, they might also be employed to reduce the text by a line or two, to make sure everything would fit onto a page if there was a reason the text could not simply spill over onto the next page. There is no evidence from the period under discussion that anyone considered those strategies dubious or harmful to the work being printed.
Vickers, however, has such objections, and finds support for some of them in Joseph Moxon’s 1680s treatise on printing. Irrespective of what he writes in his response, in the book, he primarily uses Moxon as an aesthetic guide to what one should expect from a well-printed book, and what constituted blunders (such as “botches,” “pigeonholes,” and other supposedly undesirable type formations). But there is no reason to assume that those standards applied in the early Jacobean period, and in the case of a genre as relatively ephemeral as printed drama.
To stick with his use of Moxon’s description of the printing process for a moment, though, I feel compelled to note that when Vickers does write about the technical methods detailed in the Mechanick Exercises, his summary does not reflect an especially clear understanding of the text:
in Shakespeare’s age a printer […] had to count every letter and every space between words[.] Once he had counted ‘the number of letters in the whole written copy,’ he divided the total by the number of lines he planned to print on each page, thus settling the number of pages and the quantity of paper he would need.
Vickers claims to be drawing on Moxon’s description of the casting-off process here. But Moxon says nothing of the sort. In fact, he offers a method of guesstimating “the number of letters in the whole written copy,” based on counting the number of letters “contained in an ordinary Written Line of Copy, such a Line as I guess is likely to Run Line for Line with the generality of the rest of the Copy,” i.e., Moxon picks a single line that he judges to be about the average length of a full line in the manuscript under examination and counts the number of letters in it (ignoring spaces “between words,” as Vickers might be surprised to learn); then he counts the number of lines the scribe uses per page (again, in general); then he multiplies those numbers; and then he multiplies the result by the number of pages in the manuscript, yielding a rough figure of total letters in the document. Then he decides which typeface to use, determines how many letters of that will go into a line and how many lines he wants to set per page, and divides the (guesstimated) letter count of the manuscript by the letter count per printed page to arrive at an estimate of how many sheets of paper he will need. That is to say, the method Moxon proposes bears little resemblance to Vickers’s idea that compositors counted every single letter and space in their copy text. In fact, it is a method intended to minimize the effort and allow the compositor to do as little actual counting as possible.
Let me return to the 1 Henry IV Quarto. Vickers’s use of this text is a paradigmatic example of how he treats evidence: it is an object that illustrates a set of practical responses to practical problems, but it neither fits Okes’s case, nor does it demonstrate the cavalier attitude to the text that Vickers thinks it does. The Quarto is a reprint from printed copy — and as such, unsurprisingly, it was printed by formes. (I detail the differences between setting seriatim and by formes in my original review, and thus will not go over them again here.) And as in any volume printed this way, the compositor frequently was forced to observe hard page breaks — since pages 1, 4, 5, and 8 would already be in press by the time he set pages 2, 3, 6, and 7 (or possibly vice versa). The potential problem such a compositor faced was not a shortage of “paper” in any meaningful sense, but a shortage (or an overabundance) of space. But space in these cases is not defined by paper. It is defined by printed type. The space of the page is not the same as the paper on which it printed: rather, it is constituted by a myriad of typesetting decisions: the line count, the size of the margins, the width of the measure, the kind of typeface used, the font size, how much room to leave between lines and between the elements of the page, whether to include running titles and/or page numbers, and so on. Without paper, nothing can be printed. But just as in most paintings, it is the paint and its application, not the canvas, that establish the space of the picture, in a printed book, space is created by how a compositor shapes printed matter, not by the sheet which takes the imprint.
Consequently, if “space problems” arose in typesetting, those problems normally had to do with earlier compositorial decisions that defined the space in a way that later had unanticipated consequences. The process of “casting off” when setting by formes was the common method of avoiding such nasty surprises. Pace Vickers, a printer setting by formes was not primarily concerned with “calculat[ing] how much paper he would need,” but with where page breaks within the sheet he was working on would need to fall. And if he made mistakes in that calculation, those had nothing to do with paper either, but with allowing an inappropriate amount of space for the type required to print a fixed amount of manuscript copy text. Or, more precisely, errors in casting off made a compositor shape the space on the pages printed first in such a way that the spatial conditions he had thereby created for the four pages he still needed to set did not match those required by the remaining text.
All the examples Vickers cites are books either printed by formes or involving sheets inserted after subsequent text had already been printed, creating hard breaks akin to those associated with by-formes-setting. Such is the case with the Folio printing of 2 Henry IV, as with Eastward Ho and Sejanus in 1605, with the 1609 Masque of Queens, and so on. These are all examples of printers compressing text to hit a hard target. The Folio 2 Henry IV may indeed be an instance of a stationer making small cuts to make the remaining text fit. But none of those cases at all correspond to Okes’s situation; as Peter Blayney notes, “at no time was the setting [of the Lear Quarto] constricted by the fact that a subsequent part of the text had already been printed.”
The “particularly telling” additional example Vickers gives in his response, William Stansby’s 1616 Folio printing of Jonson’s Every Man in His Humour, is inappropriate for the same reason. The argument that Jonson rewrote the last act of his play because there was not enough space left for the text as printed in the 1601 Quarto entirely depends not on the stationer misjudging the amount of paper needed, but on the fact that he had not allowed enough space for the entire play. There was a hard break created by the already printed signature G and by the continuous pagination. Stansby could have inserted an extra sheet or two — but he would have needed to give that sheet an irregular signature and he would have had to disrupt the orderly sequence of page numbers in order to do that. In other words, what he and Jonson may have been dealing with was not a shortage of paper — the 1616 Folio is an extraordinarily uneconomically printed volume, and an extra sheet or two would hardly have made much of a difference — but a shortage of available space. And that space was defined not by an arbitrary limit of how much paper the printer could use, but by the text already printed.
Vickers consistently ignores this basic but essential distinction between space and paper. Conflating the two, he writes in his response that “Okes did not have enough print space, or paper, for the text space. Like other printers in this predicament, Okes had to cut the text, removing longer and shorter passages that fortunately survive in the Folio.” But “paper” is not “space.” None of the printers he refers to suffered from a shortage of paper: they suffered from a shortage of space arising from specific earlier printing decisions. Even when faced with this kind of problem, though, none of them “cut the text” — with the possible exception of Jaggard in the Folio 2 Henry IV (if we accept Eleanor Prosser’s argument). Stansby, notably, did not “cut”: he asked Jonson to revise his play.  Other printers didn’t leave text out at all: they compressed it or altered verse to prose, using all available means to squeeze as many words in as possible, and working very hard to avoid omitting anything. But I don’t want to be more confident than the evidence warrants. Although there is very little observable data that supports the idea that stationers regularly dropped text when they ran out of space, claims about cuts or their absence are necessarily speculative. It is normally impossible to prove, as a matter of bibliographical criticism, whether a cut occurred at all, and if so, whether it was made in the printing house or, at an earlier point, in the manuscript copy text. Arguments about compression, on the other hand, can actually be grounded in bibliographical scrutiny: an unusual amount of type per page is an observable fact — if we can establish what “usual” means in any given context.
One fact such observation bears out, to return to the distinction between seriatim and by-formes setting, is that the two methods create very different kinds of spatial conditions for compositors. When setting seriatim, a printer simply has none of the constraints and hard limits unavoidable when setting by formes. That doesn’t mean that a book set seriatim can’t also be produced with an eye toward using as much of the page as possible — both methods can accommodate a need to economize. But there is far less actual, immediate, ineluctable pressure: sheet by sheet, the space remains far more fluid and flexible when setting seriatim. The method is therefore best suited for challenging manuscripts, copy texts that make it difficult for compositors to commit to firm page breaks and reliable predictions of a how many lines of type a given number of handwritten lines correspond to. It would negate the benefits of the approach to shackle its practitioner to a non-negotiable sheet count, and there is no evidence — certainly Vickers provides none — that printers were regularly constrained in such a way by the stationers that hired them.
King Lear would have been a challenging text to set by formes, no matter what the condition of the manuscript, because of its mix of verse and prose, and it is unsurprising that Okes would have chosen to set it seriatim. As D. F. McKenzie argued decades ago, in an essay that remains a key text of modern bibliography, it is not at all clear that casting-off would have been anyone’s preferred method when setting a regular quarto from manuscript. To justify Okes’s decision, it is not necessary, as Vickers seems to think, to posit that the copy text for the Lear Quarto was particularly “messy.” It is certainly true, though, that a copy text of poor quality would have been a very good reason for setting the Quarto seriatim. Vickers claims that Peter Blayney “explicitly states that Okes chose this method due to Shakespeare’s messy manuscript.” As usual when reporting Blayney’s arguments, Vickers exaggerates and removes qualification. What Blayney actually says, noting that both W. W. Greg and Fredson Bowers suggested that the copy text must have been of very poor quality (“miserable,” wrote Bowers), is that this was “probably what caused Okes to print Lear in the way he did” (my emphasis). As a bibliographer, Blayney is typically careful to distinguish between certainty and informed speculation.
This may seem like a petty response, but I stand accused of “inadequate knowledge” by a critic who “doubts that [Syme] has ever read Blayney.” It therefore matters to me to point out that Vickers, both in his book and in his response, repeatedly mischaracterizes the arguments of Peter Blayney, who happens to be affiliated with the same “major Canadian university” as me, and whose work has had a deep influence on my own. Nowhere is this dishonesty more apparent than in Vickers’s charge that I “suppres[s]” Blayney’s “observation” (later “the decisive point”) that “Okes probably guessed (badly) that he could fit Lear into 10 ½ sheets.” It may come as a surprise to learn that this crucial determination cannot be found anywhere in the over 700 pages of Blayney’s The Texts of ‘King Lear’ and Their Origins. Instead, one has to turn to a footnote in The One King Lear, where the “decisive point” is reproduced from a “personal communication.” Contrary to what Vickers claims in his response, though, the passage from that “communication” reproduced in the footnote reveals that Blayney does not at all find it
puzzling that Okes didn’t return to the publisher, Nathaniel Butter, who supplied the paper and paid for the printing, especially since “it would have been to Okes’s advantage to confess that he’d miscalculated and ask to renegotiate. He’d earn more for printing twelve sheets.
Rather, what puzzles Blayney is that Vickers believes a printer would take extraordinary and destructive measures merely to stick to a poorly guessed estimate, instead of asking the publisher for additional paper.
Vickers is just as incorrect in writing of “Blayney’s demonstration that Okes differed from other printers in leaving the outside pages blank.” Blayney demonstrates (or claims) no such thing. It is true that he notes that Okes seems to have had a general preference for leaving “a book’s outside pages blank, in order to protect the print,” but he does not characterize this approach as “idiosyncratic” — that is all Vickers. As a matter of fact, Okes’s habit was nothing of the sort.
How can I say that? I checked. It isn’t a reviewer’s job to do an author’s research for him, so I did not go very far, but even a small sample of 25 non-Shakespearean plays printed between 1606 and 1608, plus the three Shakespeare quartos Vickers himself discusses in detail, proved revealing. Of those 28 volumes, six feature a blank leaf before the title page and a final blank page or leaf; one has a blank leaf before the title page; four more have a final blank page; and an additional seven have a final blank leaf. In other words, of the 28 books, only 11 have type on the final page. Only one of those books was printed by Nicholas Okes. He was following the majority of playbook producers in this sample in “protecting the print” at the end of the book, and his Lear is one of seven quartos that protects the title page with a blank leaf. Nor is it the case that stationers religiously adhered to such habits. Ten of the books in the sample were printed by George Eld, and he probably liked to end his books with a blank — eight of them have this feature. But if the text needed to continue onto the last verso, Eld let it (in The Revenger’s Tragedy and The Travels of the Three English Brothers). So if Okes had seen any need to continue the text of King Lear past the final recto, he would almost certainly have done so. Habits are not hard and fast rules.
This example may give an indication of the problematic nature of Vickers’s habit of generalizing about the differences or similarities between Okes and other stationers without grounding his claims in a broad selection of evidence. As I noted in my original review, his most extended comparison, between the 1608 Lear and three other Jacobean Shakespeare quartos, draws on far too limited a sample of books to yield meaningful results. It also focuses on aspects of type setting that have little to do with efforts to reduce the number of pages: the “space saving” strategies Vickers analyzes in this comparison are mostly methods used by printers to justify lines — to shorten or lengthen a line of type in order to make it fit its measure. He successfully shows that Okes’s compositors had greater trouble than three of their colleagues in three instances to justify lines. Whether Okes’s workers were facing greater challenges than most printers of plays, or most compositors is impossible to say based on the extremely small dataset Vickers provides — a dataset further weakened by the fact that one of the volumes (the 1619 Quarto of Lear) is a reprint set from printed copy (a much easier operation than setting from manuscript), whereas the two volumes printed from manuscript were both set by formes — unlike the 1608 Lear Quarto. What Vickers shows is that Okes crammed more words onto each page than the other three stationers (the 1605 Hamlet, the 1609 Troilus and Cressida, and the 1619 Lear reprint all have fewer words per page). What he does not show is whether what those other stationers did was normative — or how unusual Okes’s practices were in the larger context of early modern book production.
This is especially disappointing as it would not have been difficult to establish and analyze such a broader context, drawing on the archive of facsimile and digitized text now so readily available online. As it happens, I am currently working with that archive on a completely unrelated research project, and so I thought it might be helpful to give at least a sliver of the kind of context that The One King Lear neglects to provide. It’s no more than a sliver, since I’m not prepared to do all the work I think Vickers should have done. Nor is it an especially well-matched dataset for answering the question “how many words did early modern printers of plays fit on a page,” as my own data seeks to answer theatre historical questions. But, for what it’s worth, here is a sample of 28 plays: Okes’s Lear, the two other non-reprints analyzed by Vickers, and 25 other plays printed around 1608 — the very books Okes may have seen in the months leading up to the printing of the Quarto and as the volume was in press. 
A few observations. Vickers is certainly right that the 1608 Lear crowds its pages more than many other playbooks. But in a richer context, it also becomes evident that the Quarto is less crowded than some. A particularly interesting case is The Isle of Gulls. This book, like Okes’s Quarto, prints verse as prose — frequently enough that Martin Wiggins, when counting its lines for his British Drama, 1533–1642: A Catalogue, “only attempted to disentangle the verse where rhyme makes it unmistakable.” The (unidentified) printer of that volume squeezed an average of 25 more words onto each of its pages than Okes. Would readers have found those more text-heavy volumes “uncomfortably crowded,” too, as Vickers judges the 1608 Lear to be? Or would the degree of crowdedness not have registered at all? More to the point, perhaps, given the sheer range of words-per-page averages, would there have been any kind of expectation of how much or how little text a page of dramatic literature should hold?
Most striking to me, though, is the realization that even with a bigger dataset than Vickers’s, it remains difficult to draw general conclusions. We might assume that the more verse a play has, the fewer words it can fit onto each page (because prose can use the space more efficiently). But the data complicates that assumption. The most verse-heavy play in the sample, Volpone, is more space-efficient than one of the most prose-heavy titles, The Phoenix. Bussy D’Ambois, although it is 90 percent verse, is more or less at the sample’s median for words per page — almost identical, in almost every way, to The Family of Love, which is 80 percent prose. The Fleer and Northward Ho are almost equally prose-heavy (94 percent versus 96 percent), but the latter prints over 60 more words per page. My sample, small though it may be, suggests that there was no “normal” playbook: how much text a page could be expected to hold differed from printer to printer, and from book to book. All we can say is that Okes’s Lear probably did not look oddly congested to a reader who saw it side by side with Northward Ho and The Isle of Gulls.
Why does that matter? Because in order to make the kind of argument Vickers wants to make, it is necessary to establish a number of things: a) that the 1608 Lear Quarto is unusually crowded to a degree that suggests a serious shortage of space; b) that the volume uses available space in a way that confirms that Okes was running out of room; c) that the appearance of the book makes it likely that the printer had no other choice but to cut parts of his copy text if he wanted to fit a reasonably coherent version of the play into the available pages. Given the lack of uniformity evident in the dataset I have looked at, it seems to me that establishing (a) will be exceedingly difficult, especially since the Lear Quarto could easily have been more crowded than it already is. But leaving that aside for a moment, what about (b) and (c)?
We have already seen that Okes’s decision to leave a full leaf in the first gathering and the final page blank was neither idiosyncratic nor a necessity: it was a choice he could or could not make, and one he could easily have abandoned if he was running out of room. Vickers seems convinced that Okes could not help but stick to his guns; the printer, having made the irrational choice to leave three pages empty, “had to work even harder to accommodate the nearly 3,000 lines of this play on the 79 pages left, at 38 lines per page.” But no one forced the printer to stick to 38 lines — in fact, on two pages he didn’t. Okes could easily have accommodated 39 lines per page throughout the book, making room for an extra 77 lines. The point, which I made at length in my review, but which clearly needs restating, is this: we have no external evidence that Okes had trouble fitting the text in Shakespeare’s manuscript onto the 10.5 sheets he had allocated to the volume. None whatsoever. The hypothesis that a space problem existed at all needs to be confirmed wholly and entirely with reference to the specific features of that volume as it was printed. It is only from those features that we can deduce what kinds of problems Okes may have run into when turning the manuscript into a printed book.
That is why it is so bizarre for Vickers to write: “wishing to disprove my diagnosis of Okes’s inadequate supply of paper, Syme argues that, since he wasted so much space, he had no shortage — an obvious non sequitur.” Vickers’s “diagnosis,” however, is not an investigation of the Quarto at all. Instead, he proceeds from the assumption that the 1608 text was cut and then seeks an explanation for that thesis, an explanation he finds in the hypothesis that Okes did not have enough paper. But unless there is clear bibliographical evidence for a severe lack of space, we can’t know if this hypothesis is at all tenable. The first question to ask, therefore, is whether the book actually shows symptoms of a paper and/or space shortage: I find it difficult to see how to answer that question without observing how much space the printer left unused.
Vickers suggests that I am “unhappy” that Okes “wasted so much space”; he claims that I try to “second-guess” the printer and write about “the choices [I] thin[k] that Okes should have made”; and concludes that “the fact that Okes did none of” the things I suggest he could have done instead of cutting his copy text “cannot support a claim that he had plenty of print space available.” These are baffling comments, given the context of my remarks. That context is the one provided by The One King Lear: the hypothesis that the manuscript underlying the 1608 volume contained more text than could be accommodated in a quarto of 10.5 sheets. What I argued in my review was this: to test that hypothesis, one needs to look at the Quarto and see if there are clear signs that space was so tight that Okes was forced to cut the text. I take it that Vickers and I concur that omitting copy text would have been a last resort for a printer — if we do not, then, frankly, all bets are off. If we do, however, agree that any stationer would have exhausted other means of making the text fit first, then it is reasonable to consider those other methods by which Okes could have accommodated more text before making cuts, and to ask if he made use of them. As it turns out, such means were readily available to him — and went underutilized.
It seems to me that the only “non sequitur” here is Vickers’s assertion that the indisputable fact that the 1608 Quarto contains much blank space does not warrant the conclusion that Okes “had plenty of print space available.” It is perfectly obvious that he did: at least 77 extra lines and at least one full blank page in the final gathering. That space existed; it could have been used for text; instead Okes chose to leave it blank. Ergo, extra space was available but went unused. To reconcile this fact, Vickers needs to explain why a printer would choose to delete text rather than use unfilled space to print it. Or he may have to accept that a book that displays readily available blank space is unlikely to be a book from which a stationer had to omit significant portions of copy text — a conclusion even more compelling if that book was set seriatim. 
The strangeness of Vickers’s method is perhaps best demonstrated by a specific example. Disputing my suggestion that “the large title and ornament on the first page ‘shows little to no evidence of its printer’s sense that space was at a premium,’” he writes “I don’t know exactly how many early modern printers used an ornament on the first page of text, but many Quartos that I have seen include one, and I suppose it was a standard feature, without which the page might have looked rather bare.” Was Okes following set norms? Vickers doesn’t know, but he “supposes” he did.
One might ask why Vickers is happy to settle for supposition where knowledge is so easily attainable. What the first page of printed plays looked like can be readily ascertained nowadays since, as I mention above, reproductions of virtually all early modern printed playbooks are available online. And one might think that before mounting an argument that relies on assertions about what was and wasn’t standard practice, a scholar would try to establish rather than “suppose” what that practice might have been.
Again, I am not offering to do Vickers’s research for him, but it took less than an hour to look at and analyze the first pages of the 28 quartos in my sample. Eighteen of them feature some sort of ornament on the same page as the first lines of the play (the 10 that do not were all printed by George Eld); all but three of the 28 repeat the play’s title as well. Only one has neither the title nor an ornament: The Travels of the Three English Brothers. But that is only a partial answer to the question, “how unusual was Okes in using as little of the first page for play text as he did?”
The more central question would be, what proportion of those pages is occupied by ornaments and/or titles? On average, 23 percent — from The Phoenix’s 46 percent to the six percent that a large-type “Prologue attired like fame” occupies on sig. A2r of Three English Brothers. At 36 percent, Okes’s Lear is the fifth most generous out of 28 examples, well above the average or the median (both 23 percent).
As with all other aspects of early modern printing, it would be pointless to try to derive general principles from this data (that’s not being “post-truth,” it just means acknowledging the extreme range of variation in the historical evidence). For instance, the compositors who set The Phoenix may have “wasted” a remarkable amount of space on that first page of the text, but they neither included a blank leaf before the title page nor a blank final page — suggesting, if it suggests anything, that they needed all the space they had to accommodate the text. On the other end of the spectrum, The Devil’s Charter has the second-smallest header in the sample but also includes a blank final leaf; its compositors had so much space to spare close to the end of the volume that all the pages of the final two sheets have two fewer lines than the earlier parts of the book.
However, other printed playbooks make for instructive comparison with the King Lear Quarto nonetheless. The printer of The Devil’s Charter may have approached the project with the assumption that space would be tight: accordingly, he even used the verso of the title page for a dedication. (Only two other books in the sample do not have a blank title-page verso.) As the printing progressed, it evidently turned out that the text did not require as much space as the compositor thought. The book has one of the lowest words-per-page counts of the sample, partly because it is mostly in verse, partly because its compositor saw no need to set much if any of that verse as prose; and in the final stages of the production, as mentioned, the printer even reduced the line count per page to spread out the text more evenly between the final pages. The Isle of Gulls tells a different story: it has a small header too, with a narrow ornament (these take up 17 percent of the first page, below average in the sample); it too prints the title on the first page of the first sheet, and also launches into the text of the play on the second page. But unlike the Devil’s Charter printer, this one remained subject to space constraints throughout: as we saw, he set verse as prose; scenes are barely divided on the page; and its text runs right to the end of the final verso, leaving only eight lines for a “FINIS” and a one-line erratum notice correcting an error on sheet B (a reminder that even a printer who crowds his pages may still care about textual fidelity).
What we can observe in these two volumes are stationers approaching a new project with the assumption that space will be tight; one had the luxury of easing up as the printing went on, the other had to economize to the end. And neither did the kinds of things we find Nicholas Okes doing in King Lear.
In fact, none of the printers in my sample quite acted like Okes. Westward Ho, a volume with a rather higher words-per-page count than King Lear features a similarly generous header and ornament, but its printer left no blank leaves or pages on either end of the book. The Puritan is a prose-heavy play with a words-per-page count nearly identical to King Lear’s, and it also leaves the first leaf and the final page blank. But its printer made do with a modest header and no ornament at all. Okes is the only stationer in the sample who manages to combine a high words-per-page count with a very generous header and front-and-end blanks.
My point is not that I find Okes’s mixed messages dismaying. I am not “unhappy,” as Vickers would have it, with what he did. I think his choices are remarkable, though, and defy expectations. The comparison with a more “predictable” quarto is illuminating: the header in the Second Quarto of Hamlet is even more generous than Okes’s — and that volume likewise features an empty first leaf and a protective empty final page. But nobody is trying to argue that James Roberts was fighting a battle for space in that book: the words-per-page figure for Hamlet is average for the sample, and there are almost no instances of verse set as prose. In other words, the book Roberts produced looks, from its first page to its last, like the work of a printer who was not consistently worried about running out of room. Okes’s Lear, on the other hand, looks perfectly relaxed in its first few pages, and unconcerned about space on the final page. But it is quite clear, from the many instances of continuous printing and other expedient uses of space, that the compositors were struggling to fit everything in before reaching the end of the penultimate page.
Saying “Okes’s compositors were struggling,” however, is a very far cry from saying “Okes ran out of paper and had to slice 100 lines out of his copy.” If one made that argument about The Isle of Gulls, it might have some credibility, crowded throughout as that volume is, and given that it uses all the available pages for print. The same would be true of The Revenger’s Tragedy, which R. A. Foakes described as “crammed into nine sheets” in his 1966 Revels edition. It has one of the smallest headers in the sample (nine percent). There are frequent instances of space-saving mislineation and some setting of verse as prose, and an extraordinary amount of continuous printing — 62 instances in the last three sheets alone (compared to 31 in the last three sheets of Lear; the totals are 128 for Revenger’s Tragedy and 56 for Lear). The compositor exceeded his standard line-count of 38 four times, once running it up to 40. And the volume ends with a final page of 36 lines of text, squeezing a small italicized FINIS in at the very end.
It is books such as these that one might look at, from a bibliographical perspective, and conclude that their printers were forced to take fairly desperate measures as they came close to running out of room. The fundamental basis for that conclusion is the observation that there is little to no excess space. The same is simply not true for Okes’s Lear Quarto, which has lots of space remaining. The claim that the printer of that volume was under the same kind of pressure as the printer of The Revenger’s Tragedy has no basis in observable bibliographical data.
I made this case, in less detail, in my original review. Vickers’s response dismisses my arguments as “counterfactuals” and “non-sequiturs.” But nothing illustrates the lengths to which he needs to go to maintain the notion that Okes had no choice but to cut Shakespeare’s text better than his response to my suggestion that besides using all the available space, the printer could simply have resorted more frequently to the method of setting verse as prose that he applied elsewhere in the Quarto. Vickers’s reaction to this proposition deserves to be repeated in full:
Syme appears to be blind to the fact that, in a period that saw the wonderful flourishing of verse drama in England, by setting verse as prose Okes was totally “altering the text” (my italics). Other printers took this desperate step occasionally when caught for space, but no other play quarto did so for page after page. As a teacher of English literature, Syme should realize that a poet-dramatist would not be pleased if a printer reduced his carefully formed verse to prose, not just altering the text but ruining it.
Never mind that he misattributes my phrase about “altering the text” to the printing of verse as prose (I invite readers to check the context), Vickers seems to be taking the extraordinary position that changing lineation and capitalization, which is all “setting verse as prose” does, is just as bad as completely leaving lines out of the book. In fact, he appears to think Okes must have believed it would be better to cut 100 lines than to set a larger number of verse lines as prose. Better no text at all than a ruined one, I suppose. How playwrights may have felt about being misprinted is neither here nor there, though I suspect they would probably have preferred having their words printed in longer lines and without capitalization to not having them printed at all. Like Vickers, though, I’m merely guessing there.
On another note, his rebuttal also serves as yet another illustration of his tendency to use exaggerated rhetoric and to ignore readily available evidence. So far was this particular tactic of saving space from being a “desperate step” that there is hardly an early modern play quarto that does not contain some verse set as prose, as even a casual browse through the lineation notes in scholarly editions shows. Worse, the claim that “no other play quarto” used the method “for page after page” has no basis in factual evidence whatsoever. I would submit that Vickers has no idea if that statement is true or not: he simply made it up. There is nothing resembling a full account of the bibliographical specifics of all early modern playbooks, and for most of them, ascertaining how much of their apparent prose was in fact written as verse in the manuscript copy text is impossible. Worse still, I am certain that Vickers knows his assertion that the 1608 Lear is unique in this regard to be false. In 2002, he published a densely researched chapter on Pericles — and the First Quarto of that play is a far worse offender, with nearly a third of its verse lines set as prose (452 out of “about 1,550,” according to Philip Edwards). By Vickers’s own count, Okes only prosifies 145 lines.
Pericles, in fact, should caution us against assuming that mislineated verse is a clear indicator of a lack of space. Shared between two experienced printers of plays, William White and Thomas Creede, the volume shows few signs that either worried about running out of room. White set pages of 37 lines, Creede of a mere 35; the entire final leaf is blank; the header takes up an above-average 25 percent of the first text page. So generously is the volume laid out that when the play was reprinted for the fourth time, in the same collection as the second Lear Quarto, William Jaggard fit the identical text onto 65 pages, compared to the first three quartos’ 68. It is therefore by no means certain that a high degree of mislineated verse necessarily points to a desire to use less paper: in the case of Pericles, the problem, as almost all scholars of the text agree, lies with the copy text. Can we confidently rule out the possibility that the same isn’t also true, at least to some extent, of the Lear Quarto? Again, the Jaggard printing makes for instructive comparison. As Vickers notes, it uses a far more generous layout, and consequently takes up six more pages of type. But how does that generosity manifest itself? Jaggard had his compositors set 37 lines per page (as in his Pericles quarto); that alone required an additional 2.25 pages. Beginning on sheet D, he almost always inserted a space before and often after entrance directions — something Okes literally never does after the opening stage direction. Jaggard also sets most, but not all, of the lines printed continuously in 1608 on separate lines. In other words, the major difference between the two quartos is that the later printing inserts more blank space into the text. Does that mean, though, that Okes, when deciding not to set his volume as freely as Jaggard would later choose to do, was responding to a need for restraint? Or could it simply mean that he was trying to be reasonably economical? Or was he merely following a different aesthetic protocol than Jaggard? As a matter of bibliography, this is a difficult question to answer. All we can really say is that Okes was more regular in his setting of stage directions than Jaggard: once past the opening direction, he never inserted a blank line between dialogue and directions anywhere.
I could go on — there are, after all, a lot of early modern playbooks, and they tell a rich and varied story of the unpredictable habits of stationers in Shakespeare’s time. But I will end here, hoping that what I have offered suffices to show that there is no tangible evidence to support a narrative that depends on the belief that Nicholas Okes was forced to cut over a 100 lines from Shakespeare’s manuscript. Vickers may still be right, coincidentally, that those lines were deleted from Shakespeare’s draft and later found their way back into the text printed in the Folio. But he has no proof at all that Okes or the amount of paper he had on hand had anything to do with that.
This is my final contribution to this debate. I don’t think much can be gained from a protracted back-and-forth between an author and a reviewer.  But what I hope to have suggested in this essay is that many questions about the status of printed playbooks remain unanswered, and that much groundbreaking research remains to be done. Shakespeare’s plays are not necessarily the best place to focus those endeavors — we know, or think we know, too much about them to see them as the run-of-the-mill print products they were in their time. But I do think we can still learn new things about those books if we consider them side by side with the many other early modern playbooks. A fresh analysis of Okes’s Lear would not begin by trying to define what makes it unusual or particular, but by thinking about how it is similar to other printed plays of its time, and what their common features may tell us about the text(s) of King Lear. In a lot of ways, that would be an old-fashioned, traditional kind of scholarly project, continuing work that started well over 100 years ago — but it’s also work that, with the help of our new digital corpora, has only just begun.
 Parenthetically, since Vickers brings up Every Man in His Humour: What we have in that play is an unequivocal example of a playwright revising his own work. And the revised version is dense with precisely the kinds of small, tinkering, seemingly pointless changes that The One King Lear dismisses, over and over again, as “unlikely” to be the kinds of things an author might do to his text. John Kerrigan made a similar point in an essay included in The Division of the Kingdoms; Vickers does not address his arguments. Kerrigan’s basic proposition, that authors are more likely to “tinker” than third-party revisers (who are more likely to provide new scenes or speeches) has recently been extended and given much greater evidentiary heft by Richard Dutton in his Shakespeare, Court Dramatist (Oxford University Press, 2016).
 A brief word on the differences between these figures and those in The One King Lear: my database, built by my research assistant Lawrence Evalyn from the XML-tagged texts of early modern plays in Martin Mueller’s “Shakespeare His Contemporaries” corpus, counts words spoken by characters. It does not include stage directions, act and scene numbers, speech tags, and the like. The overall word count for each play is therefore lower than in Vickers’s tabulation; the “words per page” count is really a “spoken words per page” figure. It isn’t a book historical database, but one designed for theater historical inquiries. What’s more, the counts for the three Shakespeare texts are best guesses: my current dataset includes data drawn from the XML-tagged Folger Digital Texts — excellent editions, but not transcripts of the relevant Quartos. To arrive at the counts included in the table, I had to estimate the word-count difference between those modern editions and the Quarto texts; I erred on the side of caution, so if anything, the word counts for the three Shakespeare texts are probably a little high. (The percentage of prose lines is based on the counts in Martin Wiggins’s British Drama 1533–1642: A Catalogue; Wiggins himself concedes that in distinguishing between prose and verse, “there is much potential for human error and more for questionable judgment.” The ratios should therefore be taken with at least a grain of salt.)
 To add one further data point: 24 lines in the 1608 Quarto are blank but for one word of text. By cutting strategically in those spots, e.g., eliminating a single adjective in the preceding line, Okes could have saved nearly two thirds of a page. That he neglected to do so is difficult to reconcile with the narrative of a desperate space crunch, and even harder to square with the story of a stationer prepared to cut text to save space.
 But because facts matter, I do want to address three specific accusations of error or mendacity. Vickers writes that he “never mentioned ‘audience members,’ nor did Clare — the RSC’s discussions took place in rehearsal.” This is untrue. On page 281, he writes, paraphrasing Clare, “when this scene [4.3] is omitted, as it was in the Royal Shakespeare Company (RSC) productions of 1990 and 1993, the results proved confusing” and goes on to quote Clare as saying that “the audience cannot properly appreciate what Kent is trying to do” without 4.3; and “without 4.3, the audience must wonder why Kent should have abandoned Lear.”
Secondly, Vickers writes that
Syme then summarizes the disadvantage of seriatim setting for a Quarto — that ‘a compositor … would have to set a total of seven pages before anything could be printed,’ due to the fact that a Quarto sheet would contain on one side pages 2, 3, 6, 7 and on the other side pages 1, 4, 5, 8. A few sentences later, however, he writes that ‘the seriatim method made it necessary for the printer to have 11 pages of type available at all times,’ which shows that he has confused Quarto and Folio setting.
This is a rather telling illustration of the limits of Vickers’s knowledge of the mechanics of printing. The confusion is his, not mine. When setting seriatim, having 11 pages of type available was necessary because of the rhythms of work the method called for: since compositors needed to complete seven pages of type before either side of a sheet could be printed, they would need to start on the next sheet while both sides of the previous sheet were still in press (or sit idle). Vickers confuses the number of pages in a sheet that needed to be set before either forme could be printed and the number of pages of type that a printer needed to have available for the seriatim method to work efficiently. His comment about “Folio setting” is perplexing: a Folio in Sixes, which has quires of three sheets (and thus twelve pages), would not normally be set seriatim, because of the very large amount of type required. But if it were, a printer would need at least 12 pages of type on hand — and in reality, more than that unless he was content to let his compositors sit idle.
Finally, with regard to Vickers’s charge that I should be embarrassed about my alleged ignorance of W. W. Greg’s distinction between “accidentals” and “substantive readings,” I am frankly at a loss as to what this has to do with the question of whether Q2 was consulted in the printing of the Folio. Vickers claimed in his book that those who “have argued that Jaggard also referred to the 1619 Quarto” were wrong, since “that suggestion has been disproved.” Knowles disproves no such thing. What he shows is that it is extremely unlikely that a “marked-up Q2 … was printer’s copy for F1” (144). If that were the case, one would expect to find many of Q2’s substantive readings in F, and if that were the matter under debate, Vickers’s point — that Knowles notes that F rejects around 1,300 substantive Q2 readings — would indeed matter. But it is simply false to assert that Knowles “disproved” that Q2 was not “referred to” in the typesetting of F. In fact, the opposite is true: he explicitly argues that the 1619 quarto was referred to by both F compositors, that it influenced the punctuation, spelling, and elisions in the Folio — and that it may have supplied 90 substantive readings. What precisely the compositors got out of Q2 is immaterial, though: what matters is the fact that they did use the 1619 text as a reference in making the 1623 one. Far from “disproving” such a use of Q2, Knowles in fact reaffirms that it occurred. I am well aware of Greg’s distinction between accidentals and substantive readings; I am also aware of its limitations, and of how the very distinction between those two categories has been questioned by some of the most important bibliographers of our time (from D. F. McKenzie to Randall McLeod to Joseph Dane). What I don’t know is what Vickers thinks this topic has to do with the issue at hand.